The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
U. S. DEPARTMENT OF LABOR
JAMES J. DAVIS, Secretary
BUREAU OF LABOR STATISTICS
ETHELBERT ST EW A R T , Comm issioner
BULLETIN OF THE UNITED STATES )
i IVY
BUREAU OF LA B O R S T A T IS T IC S j * ................ l i l O #
W H O L E S A L E
P R I C E S
S E R I E S
INDEX NUMBERS OF WHOLESALE
PRICES IN THE UNITED STATES
AND FOREIGN COUNTRIES
REVISION OF BULLETIN No. 173
O CTOBER, 1921
W ASH IN G TO N
GOVER NM EN T PR INTING OFFICE
1921
OO A
£04
CONTENTS.
-------------------------
Page.
Introduction..................................................... ..........................................................................
5
5, 6
The aim and scope of the present bulletin......................................................................
Part I.— The making and using of index numbers:
I. The history of index numbers............................................... . .................................... 7-10
II. The difficulties of measuring changes in the level of prices.............................10,11
[II. The characteristics of price fluctuations.................................................................. 11-23
IY . Varieties of methods used in making index numbers........................................ 23-93
1. The relations between methods and uses................................................ 23-25
2. Collecting and publishing the original quotations................ . ............ 25-27
3. Market prices, contract prices, institution prices, and importexport values................................................................................................ 27-31
4. Kelative versus actual prices...................................................................... 31-33
5. The numbers and kinds of commodities included...............................33-59
6. Problems of weighting................................................................................... 59-68
7. Averages and aggregates.......... ................................................ . ...................68-81
8. Base periods, chain index numbers, and fixed base series.......... . 81-91
9. The “ ideal ’ ’ formula.................................................................................... 91-93
Y . A comparison of the leading American index numbers for the years 1890
to 1918........................................... ................................................................................ 94-112
1. Analysis of the similarities and differences, by years, 1890 to 1918 94-105
2. Comparison of four leading American index numbers, by months,
July, 1914, to November, 1918............................................................ 105-108
3. Critical evaluation of the Bureau of Labor Statistics’ , Bradstreet’s,
and Dun’s index n u m bers............................................................. .... 108-112
VI. Conclusions.................................................................................................................... 112-114
LIST OF C H AR TS.
Chart 1.— Conspectus of yearly changes in prices, 1891-1918................. (Facing)
Chart 2.— Distribution of 5,578 price variations (percentages of rise or fall
from prices of preceding year)................................................................. . .......................
Chart 3.— Distribution of the price variations of 241 commodities in 1913
(percentages of rise or fall in price)................................................................... ^...........
Chart 4.— Index numbers made from the market prices and from the import
and export values of identical lists of commodities.
England, 1871-1902.
(Based on Table 5 .) ............................................................ ..................................................
Chart 5.— General-purpose index numbers including 25, 50, and 242 com
modities. (Based on Table 6 .) ........................................................................................
Chart 6.— Index numbers of the prices of 20 raw materials and of 20 products
manufactured from them. (Based on Table 7 . ) .......................................................
Chart 7.— Index numbers of the prices of wool, cotton, hides, wheat, and pig
iron in their raw, partially manufactured, and finished forms.
(Based on
Table 7 .) ............................................... ..................................................................................
Chart 8.— Index numbers of the prices of 19 mineral products and of 18 farm
crops. (Based on Table 8 .)...............................................................................................
Chart 9.— Index numbers of the prices of manufactured goods used for family
consumption and for industrial purposes. (Based on Table 9 .).........................
Chart 10.— Index numbers of the prices of 25 food products and of 25 miscel
laneous commodities. (Based on Table 13 .)..............................................................
Chart 11.— A comparison of medians and arithmetic means of 145 commod
ities. (Based on Table 16.)...............................................................................................
Chart 12.— Index numbers of the Bureau of Labor Statistics, Dun, and Bradstreet, 1890 to 1918. (Based on Table 1 9 .) ................................................ (Facing)
Chart 13.— Secular trends of index numbers of Bureau of Labor Statistics,
Dun, and Bradstreet, 1896 to 1914. (Based on Table 21.)....................................
Chart 14.— Index numbers of Bradstreet, compared with their secular trend,
1896 to 1914. (Based on Table 2 1 .) ...............................................................................
Chart 15.— Index numbers of Bureau of Labor Statistics compared with their
secular trend, 1896 to 1914. (Based on Table 21 .)...................................................
Chart 16.— Index numbers of Dun, compared with their secular trend, 1896
to 1914. (Based on Table 2 1 .)............................................ #...........................................
Chart 17.— Yearly deviations from secular trend of index numbers of the
Bureau of Labor Statistics, Dun, and Bradstreet, 1896 to 1914. (Based on
Table 2 1 .).............................................................................................................................
Chart 18.— Index numbers of the Bureau of Labor Statistics, Dun, Bradstreet,
and the War Industries Board, July, 1914, to December, 1918. (Average
prices July, 1913, to June, 1 9 1 4 = 1 0 0 .)..........................................................................
3
15
19
20
30
38
44
45
47
49
55
77
96
98
99
100
101
103
10T
Part n . — Index numbers of wholesale prices in the United States and foreign countries:
Page.
Introduction........................................................................................................................................
115
U n ited States:
In dex numbers of the U nited States Bureau of Labor Statistics.................. 115-133
Index numbers of the Federal Reserve Board ..................................................... 133-137
Index numbers of the W ar Industries Board......................................................... 137-145
In dex numbers of the U nited States Food A dm inistration............................ 145-149
In dex numbers of the U nited States Senate Committee on Finance.......... 149-159
In d ex numbers of th e A n n a list.................................................................................... 159-161
In d ex numbers of Bradstreet........................................................................................ 161-168
In d e x numbers of D u n ............................................... ..................................................... 168-171
In dex numbers of G ibson................................................................................................ 172-175
Australia:
Index numbers of the Commonwealth Bureau of Census and Statistics. 175-184
In dex numbers of the Bureau of Statistics of New South W a les................ 1 84,185
Austria:
In dex numbers of Dr. B ela i^on Jankovich............................................................ 185,186
B elgium :
* In dex numbers of Hector D en is................................................................................... 186-189
Canada:
In dex numbers of the Departm ent of Labor.......................................................... 189-201
Denm ark:
In dex numbers of the State Statistical Bureau.................................................... 201, 202
France:
Index numbers of the Annuaire Statistique de la France............................... 202-205
In dex numbers of the Statistique Generale de la France................................ 206-209
In d ex numbers of L a Reforme E conom ique.......................................................... 209-213
In dex numbers of fim ile Levasseur........................................................................... 213-216
Germ any:
In dex numbers of the former Im perial Statistical Office............................... 217-227
In d ex numbers of the Jahrbucher fur National okonomie und S ta t is tik .. 228-245
In d ex numbers of Otto S ch m itz.................................................................................. 246-254
In d ex numbers of A dolf Soetbeer............................................................................... 255-259
Great B rita in :
In d e x numbers of the Board of T rade....................................................................... 259-267
In d e x numbers of the E conom ist................................................................................ 267-274
In d ex numbers of Augustus Sauerbeck (S tatist).................................................. 275-281
In d ia :
In dex numbers of the Departm ent of Statistics................................................... 281-288
Ita ly :
In d ex numbers of the Annuario Statistico Italia n o............................................ 288-290
In dex numbers of Prof. Riccardo B a ch i.................................................................. 290-294
In d ex numbers of A ch ille N ecco................................................................................. 294-297
In dex numbers of Mario A lb e r ti.................................................................................. 297-301
Japan:
In dex numbers of the Departm ent of Agriculture and Commerce.............. 301-306
In d ex numbers of the Bank of Japan............................................... ........................ 306-308
N etherlands:
In d ex numbers of the Netherlands Statistical Office......................................... 308,309
New Zealand:
In d ex numbers of the Census and Statistics Office............................................. 309-315
In d ex numbers of James W . M cllr a ith ..................................................................... 315-320
N orw ay:
In d e x numbers of E inar R u u d ..................................................................................... 320-324
R u ssia:
In d ex numbers of the former M inistry of Commerce and Industry, Petrograd................................................................................................................................... 324-328
South Africa:
In dex numbers of the Office of Census and Statistics........................................ 328-332
S p a in :
In d ex numbers of Francisco B ernis............................................................................ 332-336
Select bibliography of additional in d e x num bers....................................................... 336-343
BULLETIN OF THE
U. S. BUREAU OF LABOR STATISTICS.
NO. 284.
W ASH INGTON
OCTOBER, 1921.
INDEX NUMBERS OF WHOLESALE PRICES IN THE UNITED STATES AND
FOREIGN COUNTRIES.
INTRODUCTION.
The first edition of this bulletin was published in July, 1915/ on
the eve of the world-wide revolution in prices caused by the war.
Until victory was achieved this revolution was overshadowed by
military issues; but it bred problems that forced themselves on
every one’s attention soon after the armistice was signed on November
11, 1918. One result of the interest thus excited was a lively demand
for this brief treatise on index numbers, exhausting the first edition
and calling for a new issue.
In the present edition Part I, “ The making and using of index num
b ers/7 has been revised, and Part II, “ Index numbers of wholesale
prices in the United States and foreign countries/7 has been brought
down to the latest date for which complete information is obtainable.
A fundamental change in the method of compiling the Bureau of
Labor Statistics7index numbers was under way in 1914-15. But the
new results were not ready in time to be utilized in the first edition
of this bulletin. Accordingly, free use was made of the bureau’s old
averages of relative prices on the 1890-1899 base. In revising Part
I of the present edition this old material has been kept wherever it
was employed merely for illustration. In other sections, notably in
the comparison of the leading American index numbers, the latest
form of the bureau’s index has been substituted for the old series.
THE AIM AND SCOPE OF THE PRESENT BULLETIN.
The aim of this bulletin is to make the index numbers of wholesale
prices currently published in the United States and foreign countries
more accessible, more intelligible, and therefore more useful.
To this end the leading series of index numbers compiled by
official bureaus, financial journals, and private investigators both at
home and abroad are described in detail. The history of each series,
the source from which quotations are taken, the number and descrip
tion of the commodities included, the methods of averaging, the
statistical results obtained, etc., are set forth as far as the facts
1 Bulletin No. 173 of the United States Bureau of Labor Statistics.
5
6
IJSTDEX LU M BER S OF WHOLESALE PKICES.
could be learned, so that anyone wishing to use the figures in question
may know how they are derived and what they mean. This sys
tematic description of the series now being published is preceded by a
critical analysis of the various methods employed to measure changes
in the level of prices—an analysis which shows the advantages and
the defects of these methods, the purposes for which the different
index numbers may properly be employed, the reasons for the dis
crepancies which usually appear when two or more series for the same
time and country are put side by side, the safeguards which are neces
sary in making comparisons between different series, and the confi
dence which index numbers merit as measures of price fluctuations.
PART I.—THE MAKING AND USING OF INDEX
NUMBERS.
BY
W ESLEY
C . M I T C H E L L .1
I.—THE HISTORY OF INDEX NUMBERS.
The honor of inventing the device now commonly used to measure
changes in the level of prices probably belongs to an Italian, G. R.
Carli. In an investigation into the effect of the discovery of America
upon the purchasing power of money, he reduced the prices paid for
grain, wine, and oil m 1750 to percentages of change from their prices
m 1500, added the percentages together, and divided the sum by
three, thus making an exceedingly simple index number. Since his
book was first published in 1764, index numbers are over 150 years
old.2
It was in England, however, where practically the same device had
been hit upon by Sir George Schuckburg-Evelyn in 1798,3 that the
theory ana practice of index numbers were chiefly developed. The
generation that created the classical political economy was deeply
interested in the violent price fluctuations that accompanied the
Napoleonic wars and the use of an irredeemable paper currency from
1797 to 1821. Several attempts were made to measure these fluc
tuations, and in 1833 G. Poulett Scrope suggested the establishment
of a “ tabular standard of value.’ 74
Interest in the study of price fluctuations lagged somewhat in the
forties; but the great rise of prices after the Californian and Aus
tralian gold discoveries started fresh investigations. W. S. Jevons
in England and Adolf Soetbeer in Germany gave a powerful impetus
to the theoretical discussion and the practical computation of index
numbers. The problem changed somewhat in form but received even
more attention after 1873, when a prolonged fall of prices began. In the
sixties the chief aim of investigation had been to discover the relations
between the rise of prices and the increased production of gold; in the
seventies and eighties the chief aim was to find the relations between
the fall of prices and the restrictions placed upon the free coinage of
1 The writer has received generous help from Prof. Irving Fisher, Prof. Allyn A. Young, Dr. Royal
Meeker, and Mr. C. H . Verrill, all of whom read the first draft of this paper and made many effective criti
cisms. In revising the paper the writer has made free use of the criticisms of the first edition published
by Prof. F. Y . Edgeworth, Economic Journal, June, 1818, Vol. X X V I I I , pp. 176-197, and by Prof. Fred
rick R . Macaulay, American Economic Review, March, 1916, Vol. V I, pp. 203-209. He is indebted once
more to Dr. Royal Meeker for critical and constructive suggestions, and to Prof. W . F. Ogburn for super
vising certain computations and for reading the manuscript. Prof. Macaulay has considered the theoretical
sections with care and suggested numerous improvements in both text and tables.
2 Del Valore e della Proporzione de’ Metalli Monetati con i generiin Italia prima delle Scoperte dell’ Indie
col confronto del Valore e della Proporzione de’ Tempi nostri. Republished by Custodi in his Scrittori
ttaliani de Economia Politiea. Parte Moderna, Vol. X III, pp. 297-368, especially pp. 335-351.
8 “ An account of some endeavors to ascertain a standard of weight and measure, ” Philosophical Trans
actions of the Royal Society of London, 1798, Part I, Art V III, pp. 133-182, especially pp. 175 and 176.
4 Principles of Political Economy, London, 1833, pp. 405-408. It is interesting to note, however,
that neither David Ricardo, who wrote several pamphlets on currency and prices during the “ bank re
striction,” nor Thomas Tooke, who published an elaborate History of Prices in 1793-1847, made use of
index numbers.
7
8
THE M A K IN G AND USING OF INDEX NUMBERS.
silver. The weightiest theoretical contributions of this period were
made by Prof. F. Y . Edgeworth, who served as secretary of a com
mittee appointed by the British Association for the Advancement of
Science for the purpose of investigating the best methods of ascer
taining and measuring variations in the value of the monetary stand
ard.7,5
The problem of price fluctuations entered upon another phase when
the world-wide rise of prices which began in 1896-97 had been under
way for several years. After 1900, and more insistently after 1910,
complaints about the rising cost of living became common in all
civilized countries. Efforts to measure this increase as well as
efforts to explain it multiplied.
Index numbers are both troublesome and expensive to compile,
yet now in the United States not less than seven wholesale-price
series are currently maintained, four of them by financial papers.
In England there are four important series; in France one; in Ger
many, before the beginning of the World War, there were three;
while the Governments of Canada, Australia, South Africa, India,
Netherlands, and New Zealand now publish official index numbers,
and private investigators have made series for Italy, Japan, Belgium,
Denmark, Norway, Austria, Spain, and Sweden, although not all of
these were kept up during the war period. This list may well be
incomplete at pjesent, and is almost certain to require additions
within a short time.
Most of the series just mentioned have been established but recently.
The oldest— that of the London Economist—was begun in 1869.6
Sauerbeck’s English series dates from 1886, Conrad’s German series
from 1887 (though in a sense it continues investigations made by
Laspeyres in 1864), and Bradstreet’s American series from 1897.
Of the remaining index numbers re g u la rly published at present, all
date from years since 1899, and the majority from years since 1909.7
With this increase in numbers there has come an improvement in
quality. The early index numbers were made by private investi
gators, at irregular intervals, from such price quotations as chance
had preserved. As public appreciation of the importance of meas
uring changes in price levels has developed, the w^ork has more and
more been assumed by financial journals and Government bureaus.
This shift has produced a greater measure of continuity in the series,
as well as greater frequency, regularity, and promptness in the pub
lication of the results. Even more important is the improvement
in the character and the scope of the price quotations from which
the index numbers are made Whereas the individual investigator
had to take what he could get in the way of data, financial journals
and Government bureaus can collect those current prices that are
best adapted for statistical treatment, and can give better assurance
of the representative value of their quotations and the uniform
quality of the commodities included in successive years.
e For the reports of this committee, see the Reports of the British Association, 1887, pp. 247-254; 1888,
pp. 181-188; 1889, p. 133; 1890, pp. 485-488, See particularly the memoranda by Prof. Edgeworth subjoined
to these reports.
e From 1864 to 1869 the Economist published the relative prices of commodities, but such separate figures
without a sum or an average do not constitute an index number proper.
7 The years mentioned are the dates of first publication, not the earliest dates 'for which relative prices
are shown. In most cases the computers carried their investigations back into the past, frequently for a
decade or more.
HISTORY OF INDEX NUM BERS.
9
This improvement in the quantity and quality of index numbers
is as marked in the United States as elsewhere. Price quotations
had been published with more or less care and system by various
newspapers and periodicals for many years before the first effort to
compile an average of price variations was made. In 1881, Mr. H. C.
Burchard, Director of the Mint, made an index number covering the
years 1825 to 1880 from quotations that had been printed in certain
reports of the Secretary of the Treasury, supplemented by quotations;
from a New York newspaper. But his data were of uncertain quality!
and his series was allowed to lapse after 1884.8 After an interval
of eight years, the Senate Committee on Finance authorized a more
ambitious effort. Under the direction of Dr. Roland P. Falkner,
the statistician of this committee, the (then) Department of Labor
made a huge collection of price quotations, running back as far as
1840, and compiled an index number including more than 200 com
modities for the years 1860 to 1891, and 85 commodities for 1840 to
1891.9 But this also was a single investigation, and the United States
did not have an index number regularly maintained yeav after year
until the establishment of Bradstreet’s series in 1897. A quasi
continuation of the Senate Finance Committee’s work, covering
the years 1890-1899, was prepared by Dr. R . P. Falkner, and pub
lished by the Department of Labor in March, 1900.10 Another short
lived series was begun by Prof. John R . Commons and Dr. N. I. Stone
in the Quarterly Bulletin of the Bureau of Economic Research later
in the same year.11 In January* 1901, the second continuous Ameri
can series was started by Dun’s Review and gradually carried back
to 1860; the third, covering the years 1890 to date, was added by
the Federal Department of Labor in March, 1902. Other series of
this type were begun by Thomas Gibson’s weekly market letters in
1910, b y the New York Times Annalist in 1913, and by the Federal
Reserve Board in 1918.
This activity in the making of index numbers was accompanied by
a rapid growth of the literature of the subject. Among the later
contributions dealing with theoretical issues, the first place belongs to
the work of an American scholar, Mr. C. M. Walsh. His great
treatise upon the Measurement of General Exchange-Value, published
in 1901, is still the most comprehensive book upon the subject. But
the bibliographies that aim to cover the field now include hundreds
of items, and to them must go the student who wishes a guide to
further reading.12
Some of the more important new series known to have been estab
lished since the war are the series compiled by the Price Section of
the War Industries Board and published in its “ History of Prices
s See Finance Reports, 1881, pp. 312-321; 1882, pp. 252-254; 1883, pp. 316-318; Report of the Director of
the Mint on the Production of the Precious Metals, 1884, pp. 497-502. Compare the criticism of this series
by Prof. J. Laurence Laughlin, Quarterly Journal of Economics, April, 1887, pp. 397 and 398.
9 See the description given on pp. 149-159.
10 See Bulletin No. 27 of the Department of Labor, March, 1900.
11 See the issues for July and October, 1900.
13 For such bibliographies see Walsh, The Measurement of General Exchange-Value, pp. 553-574, and
J. L . Laughlin, Principles of Money, pp. 221-224. The most important contributions of later date than
Laughlin’s entries are Prof. Irving Fisher’s Purchasing Power of Money, pp. 385-429, Mr. C. M. Walsh’s
“ The Problem of Estimation/’ Prof. Irving Fisher’s paper on “ The Best Form of Index Numbers,”
with discussion, in the Quarterly Publication of the American Statistical Association, March, 1921,and
Mr. A. W . Flux’s paper on “ The Measurement of Price Changes,” with discussion, in the Journal of the
Royal Statistical Society, March, 1921.
10
THE M A K IN G AND USING OF IN DEX NUMBERS.
During the W a r /’ the series compiled by the Federal Reserve Board
from data gathered by the United States Bureau of Labor Statistics,
the series designed by the same board for making international com
parisons, the series published by the United States Food Administra
tion in 1918 in a pamphlet entitled “ General Index Numbers of Food
Prices on a Nutritive Value Base,” the series established by the
London Times for Great Britain and by the Handelstidning for
Sweden, the series for Italy compiled by "Prof. Riccardo Bachi, the
series compiled by the Bank of Japan, and those published by the
Governments of South Africa and New Zealand.
II.—THE DIFFICULTIES OF MEASURING CHANGES IN THE
LEVEL OF PRICES.
It is a curious fact that men did not attempt to measure changes
in the level of prices until after they had learned to measure such
subtle things as the weight of the atmosphere, the velocity of
sound, fluctuations of temperature, and the precession of the equi
noxes. Their tardiness in attacking that problem is the more strange
because price changes had frequently been a subject of acrimonious
debate among publicists and a cause of popular agitation. Long
before the high development of the credit system and a class of
permanent wage earners practical issues of grave importance were
raised by the instability of prices, as the disturbances created in
sixteenth-century Europe by the inflow of American silver and gold
abundantly show. Perhaps disinclination on the part of “ natural
philosophers” to soil their hands with such vulgar subjects as the
prices of provisions was partly responsible for the d elay;13 but after
all a number of eminently “ respectable” men wrote upon economic
topics in every generation after the days of Columbus— to go no
further back. Nor can the technical difficulties of the problem
explain this tardiness; for the mathematical intricacy of index num
bers, and even the necessity of allowing for changes in the pure silver
content of coins, are obstacles far less formidable than those sur
mounted long before in other fields of research.
Probably the chief cause of delay was that averages of price fluctu
ations did not promise to command much confidence after they had
been made, The quotations available for use by the early investi
gators were few in number and often of doubtful accuracy. Carli,
for example, dealt with only three commodities; Shuckburg-Evelyn
with 12. About the vastly greater number of unrecorded price
fluctuations the one firmly established fact was that they exhibited
bewildering diversity. Under these circumstances, could an average
made from a few samples be accepted as a reliable measure of changes
in the general level of prices ? And if averages could not be trusted,
why trouble to devise a plan of making them? So writers upon
is One of the early British writers on prices, Bishop Fleetwood, remarked: “ * * * as the World now
goes, the greatest (tho’ I will not think the best) Part of Readers will be rather apt to despise than to com
mend the Pains that are taken in making Collections of so mean Things as the price of Wheat & Oats, of
Poultry, and such like Provisions * *
— Chronicon Preciosum, 1707, 2d ed., 1745, p. 6. Sir G. Shuckburg-Evelyn,in the paper referred to above, also felt himself on the defensive in presenting the first English
index number: “ * * * However, I may appear to descend below the dignity of philosophy, in such
economical researches, I trust I shall find favour with the historian, at least, and the antiquary.” Shuckburg-Evelyn’s discussion of index numbers, indeed, was merely a minor appendix to his discussion of
standards of weights and measures. But it has become his chief claim to remembrance.
DIFFICULTIES
OF
MEASURING
CHANGES
IN
PRICES.
H
prices long contented themselves with statements about the fluc
tuations of particular commodities, and with indefinite assertions that
the purchasing power of money had changed little or changed much.
So, also, when certain bold investigators did finally venture to make
index numbers, no one was particularly impressed by the significance
of their achievement.
This lack of faith in the validity of averages of price variations was
overcome rather slowly, partly in consequence of improvements in
business organization. The multiplication of commercial news
papers and the more systematic keeping of private and public records
provided a larger and more accurate body of quotations. Ifnproved
means of transportation made wholesale prices in the larger cities
basic for many local markets. The grading and standardizing of
commodities increased the number o f articles which could be ac
cepted as substantially uniform in quality from one year to the
next. More important still was the discovery by statisticians that
social phenomena of most kinds, though seeming to result from the
uncontrolled choice of individuals, yet reveal a striking regularity
when studied in large numbers.14 The demonstration that a formerly
unsuspected regularity lay hidden in one set of numerical data after
another encouraged economists to believe that the known price varia
tions might after all be fair samples of the more numerous unknown
variations. The general similarity of the results reached by different
investigators using dissimilar data confirmed this faith. Thus em
boldened, economic statisticians devoted much time to extending the
scope and improving the technique of index numbers. And their
growing confidence in the trustworthiness of their series was gradually
imparted to the public.
To-day few, if any, competent judges doubt the validity of index
numbers or the substantial accuracy of the results they show when
properly constructed from carefully collected data. Indeed the
danger at present is rather that the figures published will be taken
too absolutely as a complete representation of the facts about price
fluctuations. It is therefore well to begin a study of index numbers,
not by analyzing the finished series, but by inspecting the actual
changes in prices from which they are made, and which they purport
to summarize. In no other way, indeed, can the value and the
limitations of index numbers be learned.
III.—THE CHARACTERISTICS OF PRICE FLUCTUATIONS.
An excellent collection of materials for the study of changes in
wholesale prices is found in Bulletin No. 149 of the Bureau of Labor
Statistics, published in 1914. Here are given the average annual
prices at wholesale of more than 230 commodities for a period of al
most a quarter of a century. Comparison of the changes in these
actual prices is facilitated by the publication of two series of relative
prices for each commodity. One series reduces the quotations in
dollars and cents to percentages of the average actual prices in the
decade 1890-1899. The second series, which may be called u chain
relatives/7 shows the percentage by which each article rose or fell in
14 The Belgian statistician, Adolphe Quetelet, and Thomas Henry Buckle, author of the History of
Civilization in England, 1857 and 1861, were perhaps the most effective demonstrators of this fact.
THE M A K IN G AND USING OF INDEX NUMBERS.
12
price each year as compared with the year before.15 Since this sec
tion is concerned wholly with problems of method which have no
connection with any given period of time, there is no reason for
bringing all the illustrative materials down to date.
A survey of these relative figures for the 230 commodities throws
the diversity of price fluctuations into high relief. (1) During the
24 years 1890-1913 no two of the commodities quoted underwent the
same changes in price. Some articles rose rather steadily in price and
fluctuated on a much higher level in 1913 than in 1890; for example,
rosin and crude petroleum. Other articles fell much more than they
rose ancf fluctuated on a much lower level at the end than at the be
ginning; for example, soda and wood alcohol. Some articles were
steady in price, seldom changing from one year to the next; for ex
ample, bread and certain kinds of tools. Other articles changed in
price every year; for example, cotton and pig iron. (2) In every
year a considerable proportion of the commodities rose in price, a
considerable proportion fell, and a somewhat smaller proportion
remained unchanged. (3) The range covered even by the fluctua
tions from one year to the next was very wide. For example, in 1896
potatoes fell 54.6 per cent, while coke rose 41.5 per cent; in 1899
wheat flour fell 20.2 per cent, while steel billets rose 103.3 per cent;
in 1913 onions fell 38.5 per cent, while cabbage rose 58.5 per cen;t.16
Such extreme diversities as have been cited, however, give a mis
leading impression of chaos among the fluctuations. A just impres
sion can be had only from some scheme of presentation which takes
account of all the commodities at once. Table 1 is a first rough
approximation toward this end.17 It shows for each year how many
of the commodities quoted rose, remained unchanged, or fell in price,
and divides those which rose and those which fell into six groups,
according to the magnitude of their fluctuations.
15 The reader may follow the discussion more easily if he runs over the following sample of the figures
referred to.
Cotton, upland, middling.
Year.
Per cent
of in
crease^)
or de
Average
Relative
crease (—)
price per
price.
compared
pound.
with pre
ceding
year.
Year.
Average, 1890-1899. $0.07762
.11089
1890...........................
.08606
1891............................
1892............................
.07686
.08319
1893............................
.07002
1894............................
100.0
142.9
110.8
99.0
107.2
90.2
1895............................
1896...........................
1897...........................
1898...........................
1899...........................
.07298
. 07918
.07153
.05972
.06578
94.0
102.0
92.2
76.9
84.7
1905...........................
1906...........................
1907...........................
-2 2 .4
-1 0 .7
1908...........................
+ 8.2 I 1909...........................
-1 5 .8
1910...........................
1911...........................
+ 4.2
1912...........................
+ 8.5
- 9.7
1913...........................
-1 6 .5
+10.1
1900...........................
1901...........................
1902...........................
1903...........................
1904...........................
.09609
.08627
.08932
.11235
.12100
123.8
111.1
115.1
144.7
155. 9
+46.1
-1 0 .2
+ 3.5
+25.8
+ 7.7
Per cent
of in
crease ( + )
Average
or de
Relative
price per
crease (—)
price. compared
pound.
with pre
ceding
year.
SO.09553
. 11025
. 11879
.10463
.12107
123.1
142.0
153.0
134.8
156.0
-2 1 .0
+15.4
+ 7.7
-1 1 .9
+15.7
.15118
.13037
.11503
.12792
194.8
168.0
148.2
164.8
+24.9
-1 3 .8
-1 1 .8
+11.2
16 All of these figures show percentages of rise or fall from the average prices of the commodities in question
in the preceding year.
17 The figuresln this table have been brought down to 1918 to harmonize with the material in Section V ,
on “ A comparison of the leading American index numbers for the years 1890 to 1918.”
T a b le
1.—CONSPECTUS OF TH E CHANGES IN W H O L E SA L E PRICES IN T H E U N IT E D STATE S, B Y Y E A R S , 1891 TO 1918.
[Based upon the percentages of increase or decrease in price from one year to the next, computed from Table 9 of Bulletin of the United States Bureau of Labor Statistics, No. 209,
May, 1920.]
Number
of com
modities
that fell
in price.
1896...
1897...
1898...
1899...
1900...
240
241
242
242
242
133
118
73
46
38
1901...
1902...
1903...
1904...
1905...
242
242
242
242
242
1906...
1907...
1908...
1909...
1910...
modities
that
did not
change
in price.
Number
of com
modities
Less
2.0 to
5.0 to 10.0 to 20.0 to 50.0 per
that
rose
than 4.9 per 9.9 per
19.9 per 49.9 per cent or
2.0 per cent.
in price.
cent.
cent.
cent.
more.
cent.
13
11
5
29
10
26
47
25
70
35
30
39
44
44
41
19
27
25
34
40
18
16
15
15
12
44
37
42
25
22
17
19
15
4
15
17
9
17
6
18
15
13
21
3
17
16
12
10
3
12
16
2
14
3
13
82
55
78
19
77
22
9
2
1
3
35
22
16
7
4
22
35
21
12
13
30
27
21
16
9
23
24
13
10
9
34
31
34
27
20
15
12
15
17
7
16
20
22
28
25
18
30
34
45
59
18
11
40
39
57
6
17
22
26
33
73
92
135
169
184
128
61
92
106
10
6
9
12
3
40
13
23
24
13
32
14
21
22
26
35
12
22
28
24
11
16
17
20
23
25
38
22
23
22
19
20
16
27
22
22
31
29
32
26
16
35
44
28
37
21
29
29
10
31
9
27
10
16
15
143
128
113
131
242
242
242
253
253
47
48
155
98
81
5
2
25
2
3
10
5
50
15
10
9
9
32
21
20
13
18
30
28
22
10
14
18
32
26
28
32
32
31
26
13
27
14
24
24
31
25
14
24
33
52
43
12
40
42
52
45
11
17
25
19
20
4
16
22
167
162
55
124
146
1911...
1912...
1913...
1914...
1915...
253
253
252
329
342
147
80
84
152
137
18
6
12
12
9
31
12
14
35
35
43
18
16
44
29
30
25
23
28
27
25
19
19
33
35
31
36
35
59
36
19
21
28
30
22
16
27
39
21
38
14
34
30
35
39
13
35
27
23
38
10
20
7
8
28
1916...
1917...
1918...
342
337
348
12
9
56
5
2
11
1
2
16
3
3
12
3
2
2
37
15
27
88
30
73
115
172
136
1
1
2
1
14
10
2
1
11
2
9
27
7
18
2
1
4
42
100
28
75
137
133
118
169
FLUCTUATIONS.
106
140
114
192
138
Less
2.0 to
50.0 per 20.0 to 10.0 to 5.0 to
than
cent or 49.9 per 19.9 per 9.9 per 4.9 per
2.0
per
cent.
cent.
cent.
cent.
more.
cent.
Number of commodities that rose in price by-
PRICE
232
232
234
236
237
Number
OF
1891...
1892....
1893...
1894....
1895...
Number of commodities that fell in price by-
OHAEACTEEISTICS
Year.
Total
number
of com
modities
quoted
each year.
320
326
291
00
14
THE M A K IN G AND USING OF INDEX NUMBERS.
A more significant presentation of the same set of price fluctuations
is given by Table 2. To make this table a tally sheet was drawn
up for each year from 1891 to 1918, on which the changes from
prices in the preceding year were entered in the order of their
magnitude, beginning with the greatest percentage o£ fall and run
ning up through “ no change” to the greatest percentage of rise.
Then the whole number of recorded fluctuations for each year was
divided into 10 numerically equal groups, again beginning with the
case of greatest fall and counting upward. Finally the nine dividing
points between these 10 equal groups were marked off in the percent
age scale of fall, uno change,” or rise. For example, the tally sheet
for 1913 showed how the average prices of 252 commodities in that
year differed from their average prices in 1912. One-tenth of these
252 commodities showed a fall of prices ranging between 38.5 per
cent and 10.4 per cent, the second tenth ranged between a fall of
10.4 per cent and one of 3.7 per cent; the third tenth ranged between
a fall of 3.7 per cent and one of 1 per cent; the fourth tenth between
a fall of 1 per cent and “ no change;” the fifth tenth between “ no
change” and a rise of 0.5 per cent, and so on. These dividing points
( —10.4 per cent, —3.7 per cent. —1 per cent, ± 0 per cent, +0.5
per cent, etc.) between the successive tenths into which the data
were split are called “ decils.” The midmost decil, which of course
divides the whole number of observations into two equal groups, is
called the “ median.” Table 2 presents the results drawn from the
tally sheets— that is, the nine decils for each year, together with the
percentages of greatest fall and of greatest rise from prices in the
year before.
T
able
2 .—CHAIN IN D E X NUMBERS OF PRICES A T W H O L ESAL E IN THE UNITED
STATES, B Y Y E A R S , 1891 TO 1918.
[The decils are those points in the percentage scale of rise or fall in price which divide the whole number
of price changes recorded each year into 10 equal groups. Based upon the percentages of increase
or decrease in price from one year to the next, computed from Table 9 of Bulletin of the United States
Bureau of Labor Statistics, No. 269, May, 1920. ]
(—indicates a fall; + indicates a rise; ± 0 indicates “ no change.” )
'fear.
1891..............
1892..............
1893..............
1894..............
1895..............
1896.............
1897..............
1898..............
1899..............
1900..............
1901..............
1902..............
1903..............
1904..............
1905.............
1906..............
1907..............
1908..............
1909..............
1910..............
1911..............
1912..............
1913..............
1914..............
1915..............
1916..............
1917..............
1918..............
Average..
i
Great1st
• est 1 decil.
fall.
2d
decil.
3d
decil.
4th
decil.
Per ct.
-1 3 .2
-1 6 .0
-1 1 .9
-2 1 .4
—14.0
-1 7 .8
-1 1 .5
- 7.0
- 3.8
- 3.6
-1 5 .0
- 7.4
-1 2 .6
-1 5 .0
- 7.6
- 4.8
- 3.2
-2 1 .3
- 7.7
- 6.1
-1 5 .1
- 6.8
-1 0 .4
-1 2 .0
-1 2 .0
+ 2.1
+ 8.7
- 6.0
-1 0 .1
Per ct.
- 8.0
-1 1 .2
- 8.0
-1 5 .8
- 9.6
-1 1 .3
- 7.2
- 3.3
± o
± o
-1 0 .2
- 1.6
- 5.3
- 7.6
- 3.9
± o
± 0
-1 6 .0
- 3.7
- 2.4
- 9.8
- 2.9
- 3.7
- 7.4
- 5.9
+ 6.7
+ 19.4
+ 2.0
- 5.0
Per ct..
- 4.8
- 8.5
— 5.5
-1 3 .4
- 6.5
- 7.5
- 4.4
.4
± o
+ 3.2
- 6.1
± o
- 2.1
- 3.5
- 1.0
± o
± 0
-1 0 .8
- 1.1
.4
- 7.0
.5
- 1.0
- 4.1
- 1.9
+ 10.5
+25.1
+ 8.6
- 2.9
Per ct.
- 1.4
- 5.4
- 2.4
-1 0 .8
— 4.1
- 3.0
- 1.7
± 0
+ 2.6
+ 5.1
- 3.7
± o
± o
.6
± o
+ 2.8
+ 1.2
- 5.8
± o
± o.
- 4.2
± o
± o
- 1.3
.1
+ 14.4
+28.6
+ 14.8
Per ct.
-3 0 .5
-4 1 .2
-2 7 .5
-4 4 .3
-3 8 .0
-5 4 .6
-5 0 .9
-2 1 .9
-2 0 .2
-2 9 .2
-4 2 .6
-4 0 .6
-3 3 .7
-4 3 .8
-4 4 .9
-3 9 .1
-4 3 .0
-3 9 .5
-2 9 .8
-3 7 .7
-4 7 .4
-3 6 .1
-3 8 .5
-3 7 .3
-6 0 .4
-1 9 .1
-3 4 .1
-5 1 .0
-3 1 .9
+
1
Median.
Per ct.
-J- 0
— 3.1
0
i| ±
— 7.1
| — 2.4
j -- 1.2
± 0
+ 1.8
+ 5.5
+ 7.5
— 1.5
+ 2.2
+ 1.3
± 0
+ .7
+ 5.1
+ 3.9
-- 3.8
± 0
+ 1.5
— .9
+ 1.0
+ .5
± 0
± 0
+ 18.6
+ 34.8
+18.5
.9 + 3.0
6th
decil.
7th
decil.
8th
decil.
9th
decil.
Great
est
rise.
Per ct.
± 0
— 0.5
± 0
— 5.0
± 0
± 0
± 0
+ 5.0
+ 7.6
+ 9.6
± 0
+ 4.7
+ 3.7
1.3
+ 3.2
+ 6.4
+ 6.6
—
.9
+ 1.7
+ 3.6
± 0
+ 3.6
+ 2.4
± 0
+ 2.7
+ 24.0
+42.1
+22.1
+ 5.1
Per ct.
+ 1.5
± 0
+ 1.1
— 3.3
+ .7
+ .3
+ 2.9
+ 8.3
+ 10.6
+ 12.7
+ 1.3
+ 7.1
+ 5.3
+ 3.0
+ 5.9
+ 9.7
+ 8.9
± 0
+ 5.0
+ 6.3
± 0
+ 6.7
+ 4.5
+ 1.5
+ 6.0
+30.1
+ 49.3
+28.6
+ 7.3
Per ct.
+ 5.0
+ 1.1
+ 4.8
- 1.3
+ 4.2
+ 4.3
+ 6.2
+ 13.3
+ 16.4
+ 17.4
+ 4.9
+ 12.1
+ 8.3
+ 5.9
+ 9.6
+ 14.5
+ 12.3
+ .8
+ 8.1
+ 9.2
+ 2.9
+ 11.0
+ 7.5
+ 5.0
+ 10.1
+38.7
+57.5
+36.1
P er ct.
+ 15.3
+ 5.5
+ 11.0
± 0
+ 12.1
+ 10.2
+ 12.7
+ 19.8
+30.8
+25.6
+ 13.2
+20.4
+ 14.1
+ 11.7
+ 15.9
+ 18.9
+ 17.6
+ 6.2
+ 16.0
+ 18.6
+ 11.0
+ 17.7
+ 12.7
+ 9.1
+ 18.7
+53.4
+ 69.3
+46.3
+ 19.0
Per. ct.
+ 53.0
+ 28.0
+ 59.1
+ 31.1
+ 61.9
+ 41.5
+ 101.6
+ 60.4
+ 103.3
+ 72.8
+ 53.0
+ 58.9
+ 37.4
+ 39.9
+ 46.0
+ 40.7
+ 67.8
+ 44.9
+ 70.1
+ 49.5
+ 86.1
+ 46.2
+ 58.5
+ 76.4
+ 172.9
+ 155.1
+ 154.2
+ 118.0
+ 71.0
+ 11.5
Chart 1.—CONSPECTUS OF Y E A R L Y CHANGES IN PRICES, 1891-1918.
(Based on Table 2.)
33220°—21.
(To face page 15.)
CHARACTERISTICS OF PRICE FLUCTUATIONS,
15
Chart 1, based upon Table 2 and drawn to a logarithmic scale, gives
a more vivid idea of these price fluctuations. It shows for each year
the whole range covered by the recorded changes from prices in the
preceding year b y vertical lines, which connect the points of greatest
rise with the points of greatest fall. These lines differ considerably
in length, which indicates that price changes cover a wider range in
some years than in others. The heavy dots upon the vertical lines
show the positions of the decils. One-tenth of the commodities
quoted in any given year rose above their prices of the year before by
percentages scattered between the top of the line for that year and the
highest of the dots. Another tenth fell in price by percentages scat
tered between the bottom of the line and the lowest of the dots. The
fluctuations of the remaining eight-tenths of the commodities were
concentrated within the much narrower range between the lowest
and the highest dots. The dots grow closer together toward the
central dot, which is the median. This concentration indicates, of
course, that the number of commodities showing fluctuations of
relatively slight extent was much larger than the number showing
the wide fluctuations falling outside the highest and lowest decils,
or even between these decils and the decils next inside them.
The middle dots or medians in successive years are connected by a
heavy black line, which represents the general upward or downward
drift of the whole set of fluctuations. To make this drift clear the
median of each year is taken as the starting point from which the
upward or downward movements in the following year are meas
ured. Hence the chart has no fixed base line. But in this respect
it represents faithfully the figures from which it is made; since these
figures are percentages of prices in the preceding year, a price fluc
tuation in any year establishes a new base for computing the percent
age of change in the following year. The fact that prices in the
preceding year are the units from which all the changes proceed is
further emphasized by connecting the nine decils, as well as the
points of greatest rise and fall, with the median of the year before by
light diagonal lines. The chart suggests a series of bursting bomb
shells, the bombs being represented by the median dots of the years
before and the scattering o? their fragments by the lines which radiate
to the decils and the points of greatest rise and fall.18
Time is well spent in studying this chart, because it is capable of
giving a truer impression of the characteristics of price changes than
can be given by any other device. 'T he marked diversity of the fluctu
ations of different commodities in the same year— some rising, some
falling, some remaining unchanged— the wide range covered by these
fluctuations, the erratic occurrence of extremely large changes, and
the fact that the greatest percentages of rise far surpass the greatest
percentages of fall are strikingly shown; but so also are the much
greater frequency of rather small variations, the dense concentration
near the center of the field, the existence of a general drift in the whole
complex of changes, and the frequent alterations in the direction and
the degree of this drift. But if the chart is effective in giving these
impressions, it leaves them rather vague* To render certain of them
18 Owing to the constant shifting of the base line, no fixed scale of relative prices can be shown on the
margin of the chart. Because of its intricacy, the chart had to be reproduced on a larger scale than in the
other cases, but of course that fact does not alter the slant of the lines, and this slant is the matter of
importance.
THE M A K IN G AND USING OF INDEX NUMBEES.
16
more definite, recourse must be had to the figures from which the
chart was drawn.
These figures, already given in Table 2, enable us to measure the
concentration of the mass of fluctuations about the center of the field.
One way to measure this concentration is to compute the differences
between the successive decils; that is, to find the range within which
successive tenths of the fluctuations fall. This “ range” is, of course,
a certain number of points in the percentage scale of change from
prices in the year before. When this computation is made for the
whole period covered by the table, we get the results presented in
Table 3. As heretofore, the successive tenths of the fluctuations
represented are reckoned by starting with cases of greatest fall in price
and counting upward to cases of greatest rise. The central division
of the table shows that the average range covered by the fluctuations
diminishes rapidly as we pass from the cases of greatest fall toward
the cases of little change, and then increases still more rapidly as we
go onward to the cases of greatest rise. The right-hand group of
columns shows how the range increases if we start with the two
middle tenths, take in the two-tenths just outside them, then the twotenths outside the latter, and so on until we have included the whole
body of fluctuations. The left-hand group of columns, on the other
hand, combines in succession the two-tentns on the outer boundaries,
then the two-tenths immediately inside them, and so on until we get
back again to the two central tenths. Perhaps the most striking sin
gle result brought out by this table is that eight-tenths of all the fluc
tuations are concentrated within a range (29.1 per cent) slightly wider
than that covered by the single tenth that represents the heaviest de
clines (21.8 per cent), and much narrower than that covered by the
single tenth that represents the greatest advances (52 per cent).
T
able
3 .—AVER AG E CONCENTRATION OF PRICE FLUCTUATIONS AROUND TH E ME
D IAN , 1891 TO 1918.
[Based upon Table 2.
The fluctuations represent percentage changes from average prices in the preceding
year.]
Average range covered by the—
1st and
10th
tenths
of the
price
fluctu
ations.
73. 8<
2d and
9th
tenths
of the
price
fluctu
ations.
3d and
8th
tenths
of the
price
fluctu
ations.
6.3<
4th and 5th and
7th
6th
tenths tenths
of the of the
price
price
fluctu fluctu
ations. ations.
Central
two
Successive tenths
tenths
of the price fluctu of the
ations.
price
fluctu
ations.
1st
2d
3d
4th
5th
6th
7th
8th
9th
10th
tenth,
tenth,
tenth,
tenth,
tenth,
tenth,
tenth,
tenth,
tenth,
tenth,
21.8
5.1
2.1
3.8
2.1
2.1
2.2
4.2
7.5
52.0
Central
four
tenths
of the
price
fluctu
ations.
10.2
Central
six
tenths
of the
price
fluctu
ations.
Central
eight
tenths
of the
price
fluctu
ations.
Whole
number
of the
price
fluctu
ations.
102.9
CHARACTERISTICS OF PRICE FLUCTUATIONS.
17
Such results as these gain greatly in significance by being put
beside corresponding results for other groups of statistical data. The
best comparison to make, however, is one between the actual distri
bution of our price fluctuations about their average and a “ normal”
distribution of the same data— that is, a distribution according per
fectly with the so-called “ normal law of error. ” This law shows how
phenomena are distributed about their average when the number of
phenomena observed is very large, and when each phenomenon is the
resultant of numerous independent factors, none of which is of pre
ponderating importance. It has been found that many kinds of phe
nomena tend to conform rather closely to this normal distribution; for
example, human heights, errors of observation, shots at a target, wage
rates in different occupations, etc.19 When it can be shown that phe
nomena are distributed approximately in this fashion, their average
can safely be accepted as a significant measure of the whole set of
variations, since even the deviations from the average are then
grouped in a tolerably definite and symmetrical fashion about the
average.
With such a comparison in view we may treat each recorded per
centage of rise or fall in price as an observation of the degree and
direction in which prices vary from one year to the next. Taking
all the commodities and all the years up to 1913 covered by the bu
reau’s chain relatives, we have 5,578 observations for analysis. Table
4 shows how these cases are distributed along a percentage scale of
rise or fall in prices which jumps two points at a time. The columns
headed “ number of cases” show how many price variations of the
given magnitude and directions occur, and the columns headed
“ proportion of cases” show the same numbers in the rather clearer
form of percentages of their sum (5,578).
Such is the actual distribution of the phenomena under analysis.
To compare it with the “ normal” distribution, we put these figures
on a chart, which presents the facts clearly to the eye. Here the
horizontal scale represents percentages of rise or fall in price, and
the vertical scale represents the number of times each percentage of
change is observed. The dotted line shows how our 5,578 cases
would have been distributed had they followed strictly the normal
law of error. The areas included by the unbroken line and the dotted
line are equal.20
19 See, for example, Prof. F. Y . Edgeworth’s article “ Probability,” Part II, Encyclopaedia Britannica,
11th ed., and the references there given.
20 Table 4 and Chart 2 might be improved by a change in form.
If the “ price variations’ 1 in each
year were computed as percentage deviations from their geometric mean in that year, the distribution
of their variations would doubtless be more symmetrical than is the dist ribution nere shown.
33226°— 21— Bull. 284------- 2
THE M A K IN G AND USING OF INDEX NUMBERS.
18
T
4o—DISTRIBUTION OF 5,578 CASES OF CHANGE IN THE W HO LESALE PRICES
OF COMMODITIES FROM ONE Y E A R TO THE N E X T , ACCORDING TO THE M AG N I
TUDE AND DIRECTION OF THE CHANGES.
able
[Based upon the chain relatives in Table II of Bulletin No. 149 of the Bureau of Labor Statistics.]
Rising prices.
Per cent of
change from
the average
price of the
preceding
year.
Num
ber of
cases.
Propor
tion of
cases.
1
1
0.018
.018
1
1
1
1
.018
.018
.018
.018
1
4
1
3
4
.018
.072
.018
.054
.072
4
6
1
3
4
1
5
.072
.108
.018
.054
.072
.018
.090
102-103.9
100-101.9
98- 99.9
96- 97.9
94- 95.9
92- 93.9
90- 91.9
88- 89.9
86- 87.9
84- 85.9
82- 83.9
80- 81.9
78- 79.9
76- 77.9
74- 75.9
72- 73.9
70- 71.9
68- 69.9
66- 67.9
64- 65.9
62- 63.9
60- 61.9
58- 59.9
56- 57.9
54- 55.9
52- 53- 9
50- 51.9
48- 49.9
Falling prices.
Per cent of
change from
the average
price of the
preceding
year,
Num
ber of
cases.
Proportion
of cases.
46-47.9
44-45.9
42-43.9
40-41.9
38-39.9
36-37.9
34-35.9
32-33-9
30-31.9
28-29.9
26-27.9
24-25.9
22-23.9
20-21.9
18-19.9
16-17.9
14-15.9
12-13-9
10-11.9
8- 9.9
6- 7.9
4- 5.9
2- 3.9
Under 2.
11
10
6
14
17
11
18
17
22
30
29
47
45
65
73
i 102
106
115
167
i 237
261
1356
355
i 410
0.197
.179
.108
.251
.305
. 197
. 323
.305
.394
. 538
.520
.843
.807
1.165
1.308
1.828
1.900
2.062
2.994
4.249
4.679
6-382
6.364
7.350
No change.
i 697
12.494
Per cent of
change from
the average
price of the
preceding
year.
Under 2.
2- 3-9
4- 5.9
6 - 7.9
8- 9.9
10-11.9
12-13.9
14-15.9
16-17.9
18-19.9
20-21.9
22-23- 9
24-25.9
26-27.9
28-29.9
30-31.9
32-33.9
34-35.9
36-37.9
38-39.9
40-41.9
42-43.9
44-45.9
46-47.9
48-49.9
50-51.9
52-53-9
54-55.9
Propor
tion of
cases,
Number
of cases.
i 405
i 375
329
1 238
200
173
i 120
107
76
71
45
39
32
17
*27
16
7
10
7
5
5
4
2
1
1
J
7.261
6.723
5.89ft
4.267
3.585
3 .1Q1
2.151
1.918
1.362
1.273
.807
.699
.574
.305
.484
.287
.125
.179
.125
.090
.090
.072
.036
.018
.018
.018
1
.018
S u m m a ry.
*
Number
of cases.
Proportion
of oases.
Rising prices...........................................................................................................................
No change................................................................................................................................
Falling prices............................................................................................................. *...........
2,567
697
2,314
46.021
12. 494
41.485
Total..............................................................................................................................
5,578
100.000
1 Location of the decils.
There are several points to notice here. While the actual and thei
“ normal” distributions look much alike, they are not, strictly speak
ing, of the same type. The actual distribution is much more pointed
than the other, and has a much higher “ mode,” or point of greatest
density. On the other hand, the actual distribution drops away
rapidly on either side of this mode, so that the curve representing it
falls below the curve representing the “ normal” distribution. The
actual distribution is “ skewed” instead of being perfectly symmetri
cal. The outlying cases of a “ normal” distribution extend precisely
the same distance from the central tendency in both directions,
whereas in the actual distribution the outlying cases run about tv/ice
as far to the right (in the direction of a rise of prices) as to the left
(in the direction of a fall). This fact suggests that the actual distri
CHARACTERISTICS OF PRICE FLUCTUATIONS.
19
bution would be more symmetrical if it were plotted on a logarithmic
scale, one which represents the doubling of one price by the same
distance from zero as the halving of another price. Another aspect
of the difference in symmetry is that the central tendency about
which the variations group themselves is free from ambiguity in one
case but not in the other. In the “ normal” distribution this ten
dency may be expressed indifferently by the median, the arithmetic
mean, or the mode; for these three averages coincide. In the actual
distribution, on the contrary, these averages differ slightly; the
median and the “ crude” mode stand at ± 0 , while the arithmetic
mean is +1.36 per cent.21 These departures of the actual distribu
tion from perfect symmetry possess significance; bat the fact remains
that year-to-year price fluctuations are highly concentrated about
their central tendency.
This study of the actual distribution of price fluctuations from one
year to the next will be found to throw light upon several problems
presently to be faced in discussing the methods of making index
21 That the arithmetic mean is slightly above zero arises partly from the fact that there are 33 percentages
of rise greater than any percentage of fall. But it also arises partly from the fact that our data come from
a period (1890-1913) when the trend of year-to-year fluctuations was more often upward than downward;
there were 2,567 cases of advance in price against 2,314 cases of fall. The median is kept from rising above
zero because the cases of “ no change, ” 697 in number, more than offset the difference between the numbers
of advances and of declines in price.
to
o
MAKING
AND
USING
OF
INDEX
NUMBERS.
THE
Chart 3.—DISTRIBUTION OF THE PRICE VARIATIONS OF 241 COMMODITIES IN 1913 (PERCENTAGES OF RISE OR F A L L IN PRICES).
OHAEACTEEISTICS OF PEICE FLUCTUATIONS.
21
numbers. For the moment we have use primarily for the demonstra
tion that these fluctuations are highly concentrated about a central
tendency. This conclusion strengthens the hope that we may make
measurements of price fluctuations that fairly represent the net
resultant of all the changes, miscellaneous as they seem to be. For
properly constructed averages have clearly a better chance of being
representative and significant when the phenomena for which they
stand have a strongly marked central tendency about which devia
tions are grouped than when the phenomena are irregularly scattered
over their range.
But it must be remembered, and with the reminder doubt reenters,
that the variations just analyzed are percentages of increase or de
crease from the prices of the year before. Most index numbers,
however, attempt to measure price fluctuations, not with reference to
the preceding year, but with reference to a period considerably more
remote. For example, in its old series, here used for illustration, the
Bureau of Labor Statistics measured prices in 1913 in terms of aver
age prices in the decade 1890-1899. Are price variations computed
in this manner highly concentrated around their central tendency
.like the price variations with which we have been dealing ?
Chart 3 answers this question emphatically in the negative. It
represents the distribution of the price variations of 241 commodities
quoted by the Bureau of Labor Statistics for the year 1913.23 These
variations are computed in two ways: (1) As percentages of rise or
fall from the prices of 1912; (2) as percentages of rise .or fall from
the average prices of 1890-1899. Of course the first set of varia
tions corresponds in character to the variations represented above in
Chart 2. The distribution of these variations, shown by the area in
closed by the unbroken line, is similar in type to the actual distri
bution in Chart 2; although it is less regular— a difference to be
expected, since the number of observations is only 241 here as against
5,578 there. But the distribution of the second set of variations
(percentages of change from the average prices of 1890-1899) as repre
sented by the area inclosed within the dotted line has no obvious
central tendency; it shows no high degree of concentration around
the arithmetic mean ( + 30.4 per cent) or median ( + 26 per cent)
and it has a range between the greatest fall (52.2 per cent) and greatest
rise (234.5 per cent) so extreme that two of the cases could not be
represented on the chart.23
Price variations, then, become dispersed over a wider range and less
concentrated about their mean as the time covered by the variations
increases. The cause is simple: With some commodities the trend
of successive price changes continues distinctly upward for years at a
time; with other commodities there is a consistent downward trend;
with still others no definite long-period trend appears. In any large
collection of price quotations covering many years each.of these
types, in moderate and extreme form, and all sorts of crossings among
them, are likely to occur. As the years pass by the commodities that
22 The bureau quoted 252 commodities in 1913; but 11 could not be included in the present comparison
because no quotations are given for them in 1890-1899.
23 In commenting on this chart Prof. Edgeworth has shown that, despite appearances, the distribution
of the price variations from the 1890-1899 base, may conform to the normal distribution as closely as the
variation from the preceding year base. For, under the condition presented by prices, the quantity ob
served may move either up or down at each successive interval (here a year), and with a number of obser
vations such as here used, an ideal distribution would appear more or less oblong (as does the dotted line
in Chart 3) rather than bell shaped.—Economic Journal, June, 1918, Vol. X X V I II, pp. 183-185.
22
THE M A K IN G AND USING OF INDEX NUMBERS,
have a consistent trend gradually climb far above or subside far below
their earliei levels, while the other commodities are scattered between
these extremes. Thus the percentages of variation for any given
year gradually get strung out in a long, thin, and irregular line, with
out a marked degree of concentration about any single point. Another
factor in scattering the percentage variations is probably that the
degree of scatter is a function of the degree of variation, and of course
variations are likely to be larger between years far apart than between
years close together.
The consequence is that the measurement of price fluctuations
becomes difficult in proportion to the length of time during which the
variations to be measured have continued. In other words, the
farther apart are the dates for which prices are compared, the wider
is the margin of error to which index numbers are subject, the greater
the discrepancies likely to appear between index numbers made by
different investigators, the wider the divergencies between the aver
ages and the individual variations from which they are computed, and
the larger the body of data required to give confidence in the repre
sentative value of the results.
From this preliminary survey of the characteristics of price fluc
tuations it appears (1) that year-to-year changes in the price level
can be measured with good prospects of success, because such varia
tions show a marked degree o f concentration about their central
tendency, but (2) that measurements of variations between years
far apart have a more problematical value. The practical question
whether the index numbers in current use can be trusted, then, may
have two answers. Perhaps they give results that are reliable as
between successive years, and at the same time doubtful for dates
between which 50, 20, or even 10 years have intervened.
The best way to test the reassuring conclusion about index num
bers for successive years and to resolve the disturbing doubt about
index numbers covering long periods is to compare different series
of index numbers that purport to measure price changes in the same
country during the same time. If the results turn out to be con
sistent with one another, our faith will be confirmed. If the results
are not consistent, we must find a valid reason for the discrepancies,
or become skeptical about the present methods of measuring changes
in the price level.
When this test is applied, the first impression is unfavorable.
For example, the five currently published American index numbers
show the following results for 1912 and 1913:
Year.
1912..............................................................
1913..............................................................
Changes......................................................
Percentage changes................................
Bureau of
Labor Sta Bradstreet’s
index
tistics’ index
number
number.
(old series).
133.6
135.2
+1. 6
+ 1 .2
$9.1867
9.2076
+ .0209
+ .2
Annalist
index
number.
143.25
139.98
- 3 . 27
- 2 .3
Gibson’s
index
number.
62.6
58.1
—4. 5
- 7 .2
Dun’s index
number.
$124. 44
120. 89
—3.55
- 2 .9
Here no two of the series are as closely consistent with each other
as one could wish. On the contrary, the five series disagree not only
as to the degree but also as to the direction of the change in prices.
And this is a comparison between the same successive years, where
measurements should be especially accurate.
METHODS USED IN M A K IN G INDEX NUMBERS.
23
Such offhand comparisons as the above, however, are not fair,
and the conclusion they suggest as to the unreliability of index num
bers can not be accepted without further study, for these various
index numbers mean different things. They do not all undertake
to measure the same quantity, hence they do not all employ the same
methods, and hence the discrepancies among their results may reveal
no real inconsistency. No valid comparison of index numbers can
be made, indeed, without a careful examination of what is measured
and how the measurement is made. Such an examination accord
ingly we must make before we can satisfy our minds upon the question
whether index numbers yield trustworthy results.
IV.—VARIETIES OF METHODS USED IN MAKING INDEX
NUMBERS.
Making an index number involves several distinct operations: (1)
Defining the purpose for which the final results are to be used; (2) de
ciding the numbers and kinds of commodities to be included; (3) de
termining whether these commodities shall all be treated alike or
whether they shall be “ weighted ” according to their relative impor
tance; (4) collecting the actual prices of the commodities chosen,
and, in case a weighted series is to be made, collecting also data
regarding their relative importance; (5) deciding whether the form
of the index number shall be one showing the average variations of
prices or the variations of a sum of actual prices; (6) in case average
variations are to be shown, choosing the base upon which relative
prices shall be computed; and (7) settling upon the form of average
to be struck, if averages are to be used.
A t each one of these successive steps choice must be made among
alternatives that range in number from two to thousands. The pos
sible combinations among the alternatives chosen are indefinitely
numerous. Hence there is no assignable limit to the possible varie
ties of index numbers, and in practice no two of the known series are
exactly alike in construction. To canvass even the important
variations of method actually in use is not a simple task.
1. THE RELATIONS BETWEEN METHODS AND USES.
The first step, framing a clear idea of the ultimate use of the results,
is most important, since it affords the clue to guide the compiler
through the labyrinth of subsequent choices. It is, however, the
step most frequently omitted.
Mr. C. M. Walsh and Prof. Irving Fisher, indeed, hold that “ an
index number is itself a purpose.” “ In averaging price variations,”
Mr. Walsh explains, “ the purpose or object is given: It is to measure
variations in the exchange value or purchasing power of money.”
Hence they logically contend that there is one “ best form of index
number.” 24 But this position is untenable. (1) As a statistical
device, index numbers have found a wide range of application outside
the field of prices. To deny the term index numbers to series which
show average variations in municipal water supply, rainfall, railroad
traffic, and the like conflicts with established usage. (2) Within the
field of prices index numbers are needed which do not aim to measure
24 See Walsh’s The Problem of Estimation, p. 116, and Fisher’s “ Rejoinder" in Quarterly Publication
of the American Statistical Association, March ,1921, p. 547. The merits of the formula which they consider
“ the best^ are discussed below, in section 9, pp. 91-93.
24
THE M A K IN G AND USING OF INDEX NUMBERS.
the purchasing power of money. For example, some one should com
pile a special series for forecasting changes in business conditions.
The compiler might select those commodities whose prices in the
past have given the earliest and most regular indications of changes
that subsequently occurred in the larger index numbers, he might
weight these series in accordance with their past reliability as price
“ barom eters/7 and he might use whatever method of averaging the
fluctuations gave the best results for his purpose. Such a series
probably would not be a reliable measure 01 variations in “ the pur
chasing power of m on ey/7 but it probably would be better adapted
to its special purpose than a series made by the formula which Prof.
Fisher and Mr. Walsh advocate as “ the best.77 (3) To “ measure
variations in the exchange value or purchasing power of m oney77 is
not a clearly defined aim. For example, in explaining his new form
of the British Board of Trade index number to the Royal Statistical
Society Prof. A. W. Flux pointed out that he might have aimed
either to find the change in the money cost of the things people buy,
or to find the net effect of the general economic situation, and espe
cially of currency and credit, on prices. In discussing this paper
Prof. G. Udney lu le added a third aim, “ To find the effect of pricechanges on currency and credit.7725 These three aims, which at first
sight seem much the same, turn out 011 closer scrutiny to differ and
to call for the use of dissimilar formulas, as Prof. Flux and Prof.
Yule argued. Nor is their list of aims in measuring the purchasing
power of money exhaustive. (4) What does “ the purchasing power
of m oney77 include ? Merely the standardized wares of the wholesale
markets which are sampled with varying thoroughness in the current
index numbers ? Or does it include also commodities at retail,
stocks, bonds, labor of all sorts, farm lands and town lots, loans, trans
portation, insurance, advertising space, and all the other classes of
goods that are bought and sold ? As Mr. W. T. Layton remarked in
discussing Prof. Flux7s paper, “ The wholesale price index number is
not a measure of the general purchasing power of money, though all
the wholesale price index numbers are constantly quoted as such. 7726
In fine, the problem of measuring the purchasing power of money
has not yet been thoroughly explored. To insist that this problem
has but one meaning and therefore one “ best77 solution obstructs
progress. It is wiser to exploit all the significant interpretations of
the problem and to consider what solution is appropriate to each.
And in addition to this general problem we should devise “ specialpurpose77 index numbers to solve particular problems with a view to
learning all we can about the fluctuations of economic quantities,
physical as well as pecuniary. The making of index numbers is still
in the experimental stage, and it will progress by the differentiation
of many types of series, each with its clearly defined uses.
The most systematic plan of treating the subject, then, would be
to begin with the different uses of index numbers and to consider
the methods appropriate to each. But that plan can not be fol
lowed in an interpretative study of the currently published series,
because most of the wholesale price index numbers are “ generalpurpose77 series designed with no aim more definite than that of
“ measuring changes in the price level.77 The only plan feasible
» Journal of the Royal Statistical Society, March, 1921, pp. 175-179 and 200.
26
Idem, p. 206.
METHODS USED IN M A K IN G INDEX NUMBERS.
25
for such a study at present is to invert the problem. Instead of
studying methods in the light of uses, we must study uses in the
light of methods. That is, we must analyze the effect of the different
methods followed in practice and so determine what the resulting
figures mean and the uses to which they may properly be put.
The following discussion proceeds upon this plan. It deals prima
rily with the popular general-purpose series and endeavors to show
how the various methods used in constructing these index numbers
determine the uses to which they are severally adapted.
2. COLLECTING AND PUBLISHING THE ORIGINAL QUOTATIONS.
The reliability of an index number obviously depends upon the
judgment and the accuracy with which the original price quotations
were collected. This field work is not only fundamental, it is also
laborious, expensive, and perplexing beyond any other part of the
whole investigation. Only those who have tried to gather from the
original sources quotations for many commodities over a long series
of years appreciate the difficulties besetting the task. THe men who
deal with data already published are prone to regard all this prelimi
nary work as a clerical compilation requiring much industry but
little skill. To judge from the literature about index numbers, one
would think that the difficult and important problems concern meth
ods of weighting and averaging. But those who are practically
concerned with the whole process of making an index number from
start to finish rate this office work lightly in comparison with the
field work of getting the original data.
We commonly speak of the wholesale price of articles like pig iron,
cotton, or beef as if there were only one unambiguous price for
any one thing on a given day, however this price may vary from
oi}e day to another. In fact there are many different prices for
every great staple on every day it is dealt in, and most of these differ
ences are of the sort that tend to maintain themselves even when
markets are highly organized and competition is keen. Of course
varying grades command varying prices, and so as a rule do large
lots and small lots; for the same grade in the same quantities, differ
ent prices are paid by the manufacturer, jobber, and local buyer; in
different localities the prices paid by these various dealers are not the
same; even in the same locality different dealers of the same class
do not all pay the same price to everyone from whom they bay the
same grade in the same quantity on the same day. To find what
really was the price of cotton, for example, on February 1, 1920,
would require an elaborate investigation, and would result in show
ing a multitude of different prices covering a considerable range.
Now the field worker collecting data for an index number must
select from among all these different prices for each of his commodi
ties the one or the few series of quotations that make the most repre
sentative sample of the whole. He must find the most reliable
source of information, the most representative market, the most
typical brands or grades, and the class of dealers who stand in the
most influential position. He must have sufficient technical knowl
edge to be sure that his quotations are for uniform qualities, or to
make the necessary adjustments if changes in quality have occurred
in the markets and require recognition in the statistical office. He
26
TH E M A K IN G AND USING OF INDEX NUM BERS.
must be able to recognize anything suspicious in the data offered
him and to get at the facts. He must know how commodities are
made and must seek comparable information concerning the prices of
raw materials and their manufactured products, concerning articles
that are substituted for one another, used in connection with one
another, or turned out as joint products of the same process. He
must guard against the pitfalls of cash discounts, premiums, rebates,
deferred payments, and allowances of all sorts. And he must know
whether his quotations for different articles are all on the same basis,
or whether concealed factors must be allowed for in comparing the
prices of different articles on a given date.
Difficult as it is to secure satisfactory price quotations, it is still
more difficult to secure satisfactory statistics concerning the relative,
importance of the various commodities quoted. What is wanted
is an accurate census of the quantities of the important staples*
at least, that are annually produced, exchanged, or consumed. To
take such a census is altogether beyond the power of the private
investigators or even of the Government bureaus now engaged in
making index numbers. Hence the compilers are forced to confine
themselves for the most part to extracting such information as they
can from statistics already gathered by other hands and for other
purposes than theirs. In the United States, for example, estimates
of production, consumption, or exchange come from most miscella
neous sources: The Department of Agriculture, the Census Office,
the Treasury Department, the Bureau of Mines, the Geological
Survey, the Internal Revenue Office, the Mint, associations of manu
facturers or dealers, trade papers, produce exchanges, traffic records
of canals and railways, etc. The man who assembles and compares
estimates made by these various organizations finds among them
many glaring discrepancies for which it is difficult to account. Such
conflict of evidence when two or more independent estimates of the
same quantity are available throws doubt also upon the seemingly
plausible figures coming from a single source for other articles. T o
extract acceptable results from this mass of heterogeneous data
requires intimate familiarity with the statistical methods by which
they were made, endless patience, and critical judgment of a high
order, not to speak of tactful diplomacy in dealing with the authori
ties whose figures are questioned. The keenest investigator, after
long labor, can seldom attain more than a rough approximation to
the facts. Yet it is only by critical use of the data now available
that current index numbers can be weighted, and the best hope of
improving weights in the future lies in demonstrating not only the
imperfections of our present statistics of production, consumption,
and exchange, but also the importance of making them better.
When all this preliminary work has been done, the original quota
tions and the weights should be published at length. Unfortunately,
many compilers of index numbers publish only the final results of
their computations, upon the ground of expense or lack of interest
in the detailed information. But much is sacrificed by taking this
easy course. First, the reputation of the index number itself is
compromised, and deservedly. No one can really test whether a
series is accurately compiled from representative quotations unless
the data and their sources are given in full. Second, and more
important, the publication of actual quotations greatly extends the
METHODS USED IN M A K IN G INDEX NUM BERS.
27
usefulness of an investigation into prices. Men with quite other
ends in view than those of the original compilers can make index
numbers of their own adapted to their peculiar purposes if provided
with the original data.
Nor is the importance of such unplanned uses to be rated lightly.
If we are ever to make the money economy under which we live
highly efficient in promoting social welfare we must learn how to
control its workings. What wares our business enterprises produce
and what goods our families consume are largely determined by
existing prices, and the production and consumption of goods aro
altered by every price fluctuation. What we waste and what we
save, how we divide the burden of labor and how we distribute its
rewards, whether business enjoys prosperity or suffers depression,
whether debts of long standing become easier or harder to pay— all
these and many other issues turn in no small measure upon what
things are cheap and what are dear, upon the maintenance of a due
balance within the system of prices, upon the upward or downward
trend of the price changes that are always taking place. But if the
prices of yesterday are powerful factors in determining what we
shall do and how w^e shall fare to-day, what we do and how we fare
to-day are powerful factors in determining what prices shall be
to-morrow. If prices control us, we also control them. To control
them so that they shall react favorably upon our economic fortunes
we need more insight than we have at present. It is, then, one of
the great tasks of the future to master the complicated system of
prices which we have gradually developed— to find how prices are
interconnected, how and why they change, and what consequences
each change entails. For when men have learned these things* they
will be vastly more skillful in mending what they find amiss in
economic life, and in reenforcing what they find good. As yet our
knowledge is fragmentary and uncertain. But of all the efforts
being made to extend it none is more certain to prove fruitful than
the effort to record the actual prices at which large numbers of com
modities are bought and sold. For such data are the materials with
which all investigators must deal, and without which no bits of
insight can be tested. Indeed, it is probable that long after the best
index numbers we can make to-day have been superseded, the data
from which they were compiled will be among the sources from
which men will be extracting knowledge which we do not know
enough to find.
3. MARKET PRICES, CONTRACT PRICES, INSTITUTION PRICES, AND
IMPORT-EXPORT VALUES.
Most American index numbers are made from “ market prices.”
These prices are usually obtained directly from manufacturers,
selling agents, or wholesale merchants; from the records of produce
exchanges and the like; or from trade journals and newspapers
which make a specialty of market reporting in their respective
fields.
Several of the important foreign index numbers are made wholly
or partly from “ import and export values7' ; that is, from the average
prices of important articles of merchandise as officially declared by
the importing or exporting firms, or as determined by governmental
commissions. For example, Soetbeer’s celebrated German series,
28
THE M A K IN G AND USING OF INDEX NUMBERS.
and (until 1921) the British Board of Trade’s official series were made
mainly from such material, and the official French series was made
wholly from import values until 1911.
A fourth source of quotations often drawn upon in Europe is the
prices paid for supplies by such institutions as hospitals, normal
schools, poorhouses, army posts, and the like. The official Italian
series, Alberti's series for Trieste, and Levasseur’s French series are
examples.
These four classes of quotations— market prices, contract prices,
import and export values, and institution prices— usually differ some
what, not only with respect to the prices prevailing on a given date,
but also with respect to the degree of change from time to time.
Accordingly it is desirable to inquire into the several advantages
possessed by each source of quotations.
Institution prices may be set aside promptly, because index num
bers made from them have a limited range of usefulness. Though the
institutions whose records are drawn upon often make purchases on
a considerable scale, yet the common description of their contract
rates as “ semiwholesale” prices points to the peculiar and there
fore unrepresentative character of such data. Moreover, there is
standardized goods which are usually quoted in the market reports.
If the aim of the investigation is to find the average variations m the
cost of supplies to public institutions, doubtless the prices they pay
are the best data to use. But if the aim is to measure the average
variations in the wholesale prices paid by the business world at
large/ then market and contract prices are distinctly the better source.
Indeed, institution prices are seldom used for the latter purpose
except when well-authenticated market quotations can not be had.
So far as is known, the series of index numbers compiled by the
Price Section of the War Industries Board for 1913-1918 is the only
series in which free use has been made of contract prices, and even in
this series contract prices were not obtained for some important articles
handled largely on the contract basis— especially pig iron. Contract
prices, indeed, seem more difficult to ascertain than open-market
prices, and they are really less appropriate data than the latter when
the purpose is primarily to ascertain in what direction prices are tend
ing from one month to the next. But when it is desired to show the
fluctuations in the prices at which the bulk of business is carried on,
it is clear that the index numbers should be made from both contract
and open-market prices and that the two sets of quotations should be
weighted in accordance with the volume of transactions which each
set represents. In the long run there may be little difference between
the fluctuations in the contract and the open-market prices for the
same commodity; but within short periods the difference is sometimes
wide. In 1915-1918, for example, contract prices made at the begin
ning of a year were often far below the level attained by open-market
prices by the end of the year. The collection of contract prices on a
larger scale and the analysis of their relation with open-market prices
are matters to which the makers of index numbers may profitably
direct greater attention.27
METHODS USED IN M A K IN G INDEX NUMBERS.
29
The theory on which import and export values are sometimes pre
ferred to market prices is that the former figures show more nearly
the variations in the prices actually paid or received by a country for
the great staples which it buys and sells than do market quotations
for particular brands or grades of these commodities. For example,
England buys several different kinds of cotton in proportions that
vary from year to year. A price obtained by dividing the total de
clared values of all the cotton consignments imported by their total
weight will show the average cost per pound actually paid b y Eng
lishmen for cotton with more certainty than will Liverpool market
quotations fo ra single grade of cotton like “ Middling American7’—
provided always that the “ declared values77 are trustworthy. Now,
if the aim of the investigation is to find out the variations in the
average prices paid or received for staples— irrespective of minor
changes in their qualities— then the preference for import and export
values is cleanly justified, again granted the trustworthiness of the
returns. But if the aim is to measure just one thing— the average
variation in prices— market prices for uniform grades are clearly bet
ter data. For index numbers made from import and export values
measure the net resultant of two sets of changes, and one can not tell
from the published figures what part of the fluctuations is due to
changes in prices and what part is due to changes in the qualities of
the goods bought and sold.
As might be expected, import and export series generally pursue
a more even course than market-price series. But this difference
may be due less to the sources from which the quotations are obtained
than to differences in the lists of commodities used. Fortunately, we
can arrange a more certain test than any of the common series pro
vide. In 1903 the British Board of Trade published the average
import or export prices of 25 commodities for which Mr. Sauerbeck
has published market prices.28 Index numbers made from these two
23
Wholesale and Retail Prices. Return to an Order of the . . . House of Commons . . . for “ Report
on Wholesale and Retail Prices in the United Kingdom in 1902, with Comparative Statistical Tables for a
Series of Years.” For Sauerbeck’ s figures see his annual articles in the Journal of the Royal Statistical
Society. The list of commodities in question is as follows:
Commodity.
Quotations given by
Board of Trade.
Bacon........................... Average import values
Barley.......................... ........do.............................
Coal............................... Average export values.
Coffee............................ Average import values
Copper.......................... ........do.............................
Cotton.......................... ........do.............................
Flax.............................. ........do.............................
Hides............................
do............................
Iron, pig...................... Average export values.
Jute............................... Average import values
Lead.............................. ........do.............................
Linseed........................ ........do.............................
Maize............................ ........d o . . .........................
Oats.............................. i........do.............................
Oil, olive................................. :........do.............................
Oil, palm..................... i ........do.............................
Petroleum................... ........do.............................
Rice.............................. !........do.............................
Silk............................................... j ........do ............................................
Sugar, refined.................... ........do ............................................
Tea.
................. i . ..d o ...........................................
Tin . .
....................... i ........do ............................................
W heat . .
................. I........do ............................................
Wool.............................
Do.............................. j Average export values.
Brands quoted by Sauerbeck.
Waterford.
English Gazette.
Wallsend, Hetton, in London.
Rio, good channel.
Chile bars.
Middling American.
St. Petersburg.
River Plata, dry.
Scotch pig.
Good medium.
English pig.
Linseed.
American mixed.
English Gazette.
Olive oil.
Palm oil.
Petroleum, refined.
Rangoon, cargoes to arrive.
Tsatlee.
Java, floating cargoes.
Congou, common.
Straits.
English Gazette.
Merino, Adelaide, average grease.
English, Lincoln, half hogs.
30
THE M A K IN G AND USING OF INDEX NUMBERS.
4 .—IN D E X NUMBERS MADE FR OM THE M AR K ET PRICES AND FROM THE IMPORT
AND E X P O R T VALU ES OF ID ENTICAL LISTS OF COMMODITIES. EN G LAN D , 1871-1902.
Ch ar t
(Based on Table 5 .)1
i This and the succeeding charts have been drawn on a logarithmic, instead of an arithmetic, scale in
order that the per cent of change may easily be discerned.
METHODS USED IN M A K IN G INDEX NUM BERS.
31
sets of data for the same commodities for the years 1871 to 1902 are
given in Table 5. The results confirm the expectation: As compared
with the import and. export index number, the market-price index
number starts on a higher level in 1871, falls to a lower point dur
ing the middle nineties, rises to a higher level in 1900, and again
drops to as low a level in 1902. But the differences are not wide.
T
5 .—COMPARISON OF IN D E X NUMBERS MADE FROM IM PORT AND E X P O R T
V AL U ES W IT H IN D E X NUM BERS MADE FROM TH E M A R K E T PRICES OF TH E
SAME COMMODITIES, B Y Y E A R S, 1871 TO 1902.
able
[Data from the British Board of Trade and from Sauerbeck.]
(Arithmetic means of relative prices.
Import
and ex
port
values.
Year.
1871.
1872.
1873.
1874
1875
1876.
1877.
1878.
1879
1880.
1881
1882.
1883
1884,
1885
1886
158
169
170
162
152
149
150
139
128
136
133
129
125
118
110
105 I
Average prices in 1890-1899=100.
Market
prices.
170
185
182
168
155
152
152
138
131
137
130
125
123
116
112
107
25 commodities.)
Year.
Import
and ex
port
values.
i 1887
1888
1889
1890
1891
1892
1893
1 1894
j
| 1896
! 1897
{ 1898
! 1899
| 1900
: 1901
' 1902
104
108
108
109
111
105
103
95
93
94
93
95
101
114
107
104
Market
prices.
107
110
110
111
111
103
104
94
94
93
91
95
105
117
100
104
4. RELATIVE VERSUS ACTUAL PRICES.
In February, 1864, Hunt’s Merchants’ Magazine published the fol
lowing statement to show how rapidly prices rose after the suspension
of specie payments in December, 1861, and the issue of the irredeem
able United States notes.29 These figures are the total prices of 55
articles quoted by their customary commercial units.
Value o f 55 leading articles o f New York commerce.
January, 3862................................................................................
April, 1862.....................................................................................
January, 1863................................................................................
March, 1863....................................................................................
July, 1863......................................................................................
October, 1863.................................................................................
January, 1864................................................................................
$804
844
1,312
1, 524
1,324
1,455
1, 693
For example, in January, 1862, coal oil is entered as 30 cents per
gallon and pig iron as $24 per ton; molasses is entered as 42^ cents
per gallon and whalebone as $69 per ton ; oats is entered as 38 cents
per bushel and corn as $59.25 per hundred bushels, etc.30
Clearly, this simple method of measuring changes in the price level
by casting sums of actual prices is not trustworthy. For a relatively
slight fall in the quotation for whalebone would affect the total, as
Hunt’s Merchants’ Magazine computes it, much more than a rela
tively enormous increase in the price of molasses. The fact that com
29 Vol. 50, p. 132.
30
See vol. 48, p. 129.
32
TH E M A K IN G AND USING OF INDEX NUMBERS.
happens to be quoted by the hundred bushels makes a 1 per cent
change from its price in January, 1862, equal to a 43 per cent change in
the price of wheat and to a 156 per cent change in the price of oats,
both of which are quoted by the bushel.
It was to avoid such patent absurdities that Carli threw his actual
prices of grain, wine, and olives in 1750 into the form of percentages
of rise or fall from their prices in 1500, and then struck the average
of the three percentages. When this operation is performed it makes
no difference whether the commodities are quoted by large or by
small units. The obvious common sense of this precedent has caused
it to be followed or reinvented by most makers of index numbers to
this day— with one slight modification. To avoid the awkwardness
of the plus and minus signs necessary to indicate whether prices have
advanced or receded, it is usual to substitute for percentages of rise
or fall relative prices on the scale of 100. For example, a rise of 10
per cent and a fall of 10 per cent are expressed by relatives of 110
and 90, respectively. Occasionally, however, percentages of rise or
fall are still used as by Carli; as, for instance, in the chain relatives
published by the Bureau of Labor Statistics in Bulletin No. 149 and
averaged in Table 4 of this bulletin. A second unimportant variant,
long practiced by the London Economist, but now seldom used
is to publish as the final result the sums of relative prices, instead
of their averages.31
In recent years a few statisticians have gone back from the use of
relative to the use of actual prices, adopting various devices to avoid
such crude errors as those perpetrated in the figures cited from Hunt's
Merchants7 Magazine. In 1897 Bradstreet’s began reducing all its
original quotations by the gallon, ton, dozen, square yard, etc., to
prices by the pound, and presenting as its index number the aggregate
prices per pound of 98 articles.32 Four years later, Dun’s Review
followed this lead with an important difference. Instead of reducing
actual quotations to quotations by the pound, it multiplied the actual
quotation for each article included by the quantity of that article sup
posed to be consumed in the course of a year by the average indi
vidual. These products were then cast up, and the sums, in dollars
and cents, were presented as an index number purporting to show
the changes in the per capita cost of a year’s supplies.33
Still later (1912), the method practiced by Dun was adopted by
the Commonwealth statistician of Australia as the basis of his official
series. However, after he had calculated the aggregate expenditure
of Australians upon his bill of goods in terms of pounds sterling, he
threw these pecuniary sums back into the form of relative numbers
on the scale of 1,000. In 1914 the United States Bureau of Labor
Statistics dropped its former practice of averaging relative prices on
the 1890-1899 base, and began to use aggregates of actual prices,
weighted by quantities entering into exchange and thrown into the
form of relatives to facilitate comparison.
Accordingly, three types of index numbers are now in general use:
(1) Averages of relative prices or average percentages of change in
31 Gibson's index number is such a sum.
See pp. 172 to 175. - The difference between sums of relative
rices and these sums divided by the number of articles included is, of course, purely formal. Averages
Eave
displaced sums in current use mainly because it is easier to make comparisons on the scale of 100
^ n o n the scale of 2,200, or whatever number is given by the addition of relative prices
32 For a criticism of this method, see p. 110.
33 The confidence merited by this index number is discussed in Section V.
METHODS USED IN M A K IN G INDEX NUM BERS.
33
prices; (2) sums in dollars and cents showing changes in the aggre
gate cost of certain definite quantities of certain commodities; (3)
relative figures made from series of the second sort. The first type
shows average variations, the second type shows the variations, oi an
aggregate, the third type turns these variations of an aggregate into
percentages of the aggregate itself as it stood at some selected time.
The differences between these types, it is true, are differences of
form, not differences of kind. As will later be shown, by using
a certain scheme of weights an aggregate of actual prices can be
made to give precisely the same results when turned into relatives
that will be given by an average of relative prices computed from
the same data. But it will also be shown that the differences of form
are important. The advantages and shortcomings of the several
types will appear as the various problems encountered in making
index numbers are discussed.
5. THE NUMBERS AND KINDS OF COMMODITIES INCLUDED.
Since the earlier makers of index numbers had to use such price
quotations as they could find, the problems how many and what
kinds of commodities to include were practically solved for them.
As Prof. Edgeworth remarks, “ Beggars can not be choosers.”
Paucity of data still hampers contemporary efforts to measure
variations of prices in the past; but the compilers of index numbers
for current years have a wider range of choice. The scope of their
data is limited not by the impossibility but by the expense of col
lecting quotations. And in the case of governmental bureaus or
financial journals the limits set by expense are neither narrow nor
rigid. Such organizations can choose many commodities if they will
or content themselves with few.
One principle of choice is generally recognized. Those commodities
are preferable that are substantially uniform from market to market
and from year to year. Often the form of quotation makes all the
difference between a substantially uniform ana a highly variable com
modity. For example, prices of cattle and hogs are more significant
than prices of horses and mules, because the prices of cattle and hogs
are quoted per pound, while the prices of horses and mules are quoted
per head.
It is often argued that the application of this common-sense prin
ciple rules out almost all manufactured goods, because such Articles
are continually altered in quality to suit the technical exigencies
of new’ industrial processes or the varying tastes of consumers. But
minor changes in quality, provided their occurrence is know~n, do not
necessarily unfit a commodity for inclusion. When the brand for
merly sold is replaced by a variant it is usually possible to get over
lapping quotations for the old and new qualities during the time of
transition. Then the new series may be spliced upon the old by
means of the ratio borne by the price of the new grade to the price of
the old grade in the years when the substitution is made. Statis
ticians willing to take the extra precautions and trouble involved by
such operations can legitimately include not only a large number of
staple raw materials and their simplest products, but also an even
larger number of manufactured goods.
33226°— 21— Bull. 284------- 3
34
TH E M A K IN G AND USING OF INDEX NUMBERS.
Some of the modem index numbers, accordingly, have long lists of
commodities. Dun's index number seems to be built up from about
300 series of quotations, the official Canadian index number includes
271, the Bureau of Labor Statistics7index number for 1919 has 328,
and the index number compiled by the Price Section of the War
Industries Board has 1,366 price series. On the other hand, many
of the best-known index numbers use less than 50 series of quota
tions. Forty-five is a favorite number, largely because of the high
reputation early established by Sauerbeck’s English series. The
British Board of Trade’s series to 1921, the official French series, the
New Zealand series, Von Jankovich’s Austrian series, and Atkinson’s
series for British India all have just 45 commodities, while the new
series of the London Economist and the relative prices published by
the former Imperial Statistical Office of Germany include 44 articles.
Even shorter lists are often used. For example, Schmitz’s German
series has only 29 commodities, the New York Annalist series 25,
and Gibson’s series 22. Private investigators working with limited
resources sometimes confine themselves to a bare dozen commodi
ties, or even less.34
These differences of practice raise important questions of theory.
Does it make any substantial difference in the results whether 25
or 50 or 250 commodities be included— provided always that the
Hsts be well chosen in the three cases ? If differences do appear in
the results, are they merely haphazard, or are they significant differ
ences? If there are significant differences, which set of results is
more valuable, that made from the long or from the short lists?
And what does the proviso that the lists be well chosen mean ? In
short, do the index numbers including hundreds of commodities pos
sess advantages over those including 50 or 25 sufficient to compen
sate for the greater trouble and expense of compiling them?
The best way to answer these questions is to experiment with lame
and small index numbers, made on a strictly uniform plan for the
same country arid the same years. Table 6 presents six such index
numbers which differ only in respect to the number and kind of
commodities included. The first column includes all the commod
ities quoted by the Bureau of Labor Statistics in 1913 except the 11
whose prices do not run back of 1908.35 . Many of the commodities
in this list are merely different varieties of the same article; for exam
ple, there are two kinds of com meal, four kinds of leather, six kinds of
women’s dress goods, eleven kinds of steel tools, etc. The second
column gives an index number in which all such groups are repre
sented by single averages, so that the number of series which enter
directly into the final results is cut down to 145.36 The third column,
which includes 50 commodities, is made up from the list adopted for
3* These statements refer to the number of series of relative prices averaged to get the final resulis as now
presented. Often two or more different varieties of an important article are counted as separate com
modities, and, on the other hand, the relative prices of slightly different articles are sometimes averaged
to make one of the series which enters into the final averages. In view of the diversity of practice in this
respect, a perfectly consistent counting of the number of distinct “ commodities ” included in the general
series is impossible. Moreover, the figures are often published with such imperfect explanations as to
make the counting of the commodities included doubtful or impossible on any interpretation of that term.
In 1921 the number of price series used in the British Board of Trade index was increased to 150.
35 To facilitate comparison, decimals have beep, dropped and the index for each year rounded off to the
nearest whole number. Regarding the changes in the number of commodities included, see Bulletin
No. 149, p. 11. The reader may be reminded once more that this is the Bureau’s old index number, mad©
before the improved method of compilation was introduced.
36 This experimental list of 145 commodities is given below. When the relative prices of closely related
articles are averaged to make a single series, the number of these articles quoted by the Bureau and
included in the group is indicated. Most of the bureau’s series which do not cover the whole period,
METHODS USED IN M A K IN G INDEX NUM BERS.
35
the Gibson index number in its original form.37 The fourth series is
made from the prices of 20 pairs, each commodity being given in two
forms, raw and manufactured, e. g., barley and malt, cattle and beef,
copper ingots and copper wire, etc.38 The last two columns contain
1890-1913, are dropped altogether. As the basis of a general-purpose index number, this revised list is
worse than the bureau's list in certain respects and better in others. See Section V.
Barley.
Cattle, 2.
Corn.
Cotton.
Flaxseed.
Hay.
Hides.
9.
10.
11.
12.
13.
Hops.
Oats.
Rye.
Sheep, 2.
Wheat.
FUEL AND LIGHTING.
1.
2.
3.
4.
5.
6.
7.
Candles.
Goal, anthracite, 4.
Coal, bituminous, 3.
Coke.
Matches.
Petroleum, crude.
Petroleum, refined, 2.
FOOD, ETC.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
Apples, evaporated
Beans.
Bread, crackers, 2.
Bread, loaf, 3.
Butter, 3.
Cheese.
Coffee.
Currants.
Eggs.
Fish, 4.
Flour, buckwheat.
Flour, rye.
Flour, wheat.
Lard.
Meal,corn,2.
Meat, beef, 3.
Meat, pork, 4.
Meat, mutton.
Milk.
Molasses.
Onions.
Potatoes.
Prunes.
Raisins.
Rice.
Salt.
Soda.
Spice, pepper.
Starch, corn.
Sugar, 3.
Tallow.
Tea.
Vinegar.
LUMBER AND BUILDING
MATERIALS.
CLOTHS AND CLOTHING.
FARM PRODUCTS.
1.
2.
3.
4.
5.
6.
7.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Bags.
Blankets, 3.
Boots and shoes, S.
Broadcloths.
Calico.
Carpets, 3.
Cotton flannels, 2.
Cotton thread.
Cott on yarns, 2.
Denims.
Drillings, 2.
Flannels.
Ginghams, 2.
Horse blankets.
Hose.
Leather, 4.
Linen thread.
Overcoatings, 2.
Print cloths.
Sheetings, 7.
Shirtings, 5.
Silk, 2.
Suitings.
Tickings.
Underwear, 2.
Women’s dress gdeds, 6.
W ool, 2.
Worsted yarns, 2.
METALS AND IMPLEMENTS.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Bar iron, 2.
Barb wire.
Builders’ hardware, 3.
Copper, ingot.
Copper, wire.
Lead, pig.
Lead pipe.
Nails, 2.
Pig iron, 4.
Quicksilver.
Silver.
Spelter.
Steel billets.
Steel rails.
Tin, pig.
Tools, 11.
Wood screws.
Zinc.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Brick.
Carbonate of lead.
Cement.
Doors.
Hemlock.
Lime.
Linseed oil.
Maple.
Oak, 2.
Oxide of zinc.
Pine, white, 2.
Pine, yellow.
Plate glass, 2.
Poplar.
Putty.
Rosin.
Shingles, 2.
Spruce.
Tar.
Turpentine.
Window glass, 2.
HOUSE-FURNISHING GOODS.
1.
2.
3.
4.
5.
Earthenware, 3.
Furniture, 4.
Glassware, 3.
Table cutlery, 2:
Wooden ware, 2.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Cottonseed meal.
Cottonseed oil.
Jute.
Malt.
Paper, 2.
Proof spirits.
Rope.
Rubber.
Soap.
Starch, laundry
Tobacco, 2.
MISCELLANEOUS.
DRUGS AND CHEMICALS.
1.
2.
3.
4.
5.
6.
7.
8.
9.
Alcohol, grain.
Alcohol, wood.
Alum.
Brimstone.
Glycerine.
Muriatic acid.
Opium.
Quinine.
Sulphuric acid.
37 The list is as follows: Wheat, wheat flour (two kinds), barley, oats, com, corn meal, potatoes, rye,
sugar 89°, sugar 96°, coffee, tea, steers, fresh beef, salt beef, sheep, mutton, hogs, bacon, hams, butter,
cotton, cotton yarns (two kinds), jute, wool (two kinds), worsted yarns, raw silk (two kinds), pig iron,
bar iron, cement, copper ingots, copper sheets, lead, anthracite coal, bituminous coal (two kinds), hides,
leather, cottonseed oil, linseed oil, petroleum (crude and refined), rubber, spruce lumber, yellow-pine
lumber, and paper. See J. P. Norton, “ A revised index number for measuring the rise in prices,”
Quarterly Journal of Economics, August, 1910, vol. 24, pp. 750-758.
'f The remajining 17 pairs are corn and corn meal, cotton and cotton textiles, flaxseed and linseed oil,
window glass and glassware, hides and leather, hogs and pork, lead (pig) and lead pipe, milk and cheese,
petroleum (crude and refined), pig iron and nails, pine boards and pine doors, rye and rye flour, sheep
and mutton, spelter and zinc, steel billets and steel tools, wheat and wheat flour, wool and woolen textiles.
36
THE M A K IN G AND USING OF INDEX NUMBERS.
index numbers each made from the prices of 25 important articles
selected at random, the two lists having no items in common.39
T
able
6 .—SIX IN D E X NUMBERS FOR THE UNITED STATES MADE FROM QUOTATIONS
FOR D IF F E R E N T NUMBERS OF COMMODITIES, B Y Y E AR S, 1890 TO 1913.
[Data from the Bulletin of the Bureau of Labor Statistics, No. 149.]
(Arithmetic means.
Year.
Average prices in 1890-1899= 100.)
25 com 25 com
242 to 261 145 com 50 com
40 com
modities,
commod modities. modities. modities. modities, second
ities.
first list.
list.
1890......................................................................
1891......................................................................
1892......................................................................
1893......................................................................
1894......................................................................
1895 . .
..................................................
1896......................................................................
1897......................................................................
1898......................................................................
1899......................................................................
1900......................................................................
1901......................................................................
1902......................................................................
1903......................................................................
1904......................................................................
1905......................................................................
1906......................................................................
1907......................................................................
1908......................................................................
1909......................................................................
1910......................................................................
1911.................*...................................................
1912......................................................................
1913....................................................................
113
112
106
106
96
94
90
90
93
102
111
109
113
114
113
116
123
130
122
125
130
126
130
130
114
113
106
105
96
93
89
89
93
103
111
110
114
114
114
116
122
130
121
124
131
130
134
131
114
114
105
105
94
94
87
89
95
103
112
109
116
115
116
118
123
132
125
132
135
129
138
138
113
114
105
101
93
95
88
89
95
108
115
116
122
118
118
122
128
138
129
135
141
135
142
139
115
112
103
103
92
95
88
90
96
107
113
111
116
118
122
123
130
132
124
133
133
129
140
142
113
118
112
107
96
93
85
84
90
103
109
107
117
117
110
115
122
132
122
128
134
131
138
133
Averages 1890-1899 . ..
........................
1900-1909..........................................
1910-1913..........................................
Number of points by which prices rose ( + )
or fell ( - ) i n 1890-1896.........................................................
1896-1907.........................................................
1907-1908.........................................................
1908-1912
.................................................
100
118
129
100
118
132
100
120
135
100
124
139
100
122
136
100
118
134
—23
+40
— 8
+ 8
-2 5
+41
- 9
+13
-2 7
+45
- 7
+ 13
-2 5
+50
- 9
+13
-2 7
+44
- 8
+ 16
-2 8
+47
-1 0
+16
Difference between highest and lowest rel
ative prices....................................................
40
45
51
54
54
Average change from year to year.............
4.0
4.1
4.9
5.5
54
6.2
5.0
39 The first list includes cotton, corn, wheat, hides, cattle, hogs, coffee, wheat flour, salt, sugar, tea,
potatoes, wool, silk, anthracite coal, bituminous coal, erode petroleum, pig iron, steel billets, copper ingots,
lead (pig), brick, average of nine kinds of lumber, jute, and rubber.
The second list includes hay, oats, rye, eggs, sheep, lard, beans, corn meal, butter, rice, milk, prunes,
cotton yarns, worsted yarns, coke, cement (Rosendale 1890-1899, Portland domestic 1900-1913), tallow,
spelter, bar iron, tin (pig), quicksilver, lime, tar, paper, proof spirits.
METHODS USED IN M A K IN G INDEX NUM BERS.
37
N um ber o f poin ts by which the selected index num bers were greater ( + ) or less (—) than the
Bureau o f Labor Statistics* series.
i
25 com
25
com
145 com- i 50 com
40 com
modities,
modities.! modities. modities. modities,
second
first list.
list.
!
Year.
1S90.........................................................................................
1891........................................................................................
1892.........................................................................................
1893.........................................................................................
1894.........................................................................................
1896.........................................................................................
1897.........................................................................................
1898.........................................................................................
1899.........................................................................................
1900...................................................................................
1901.........................................................................................
1902.........................................................................................
1903.........................................................................................
1904.........................................................................................
1905.........................................................................................
1906.........................................................................................
1907................: .......................................................................
1908.........................................................................................
1909.........................................................................................
1910.........................................................................................
1911........................................................................................
1912.........................................................................................
1913............................................. *.........................................
+
+
±
±
db
+
±
+
+
±
+
±
db
+
4+
+
1
1
0
1
0
1
1
1
0
1
0
1
1
0
1
0
1
o
1
1
1
4
4
1
Arithmetic sums..................................................................
Algebraic sums............... ...................*...............................
Average differences computed from the—
Arithmetic sums..........................................................
Algebraic sums.............................................................
23
+ 9
Maximum differences.........................................................
Minimum differences.........................................................
+ 4
± 0
+
1
2
1
1
2
o
3
1
2
1
1
o
3
1
3
2
o
2
3
7
5
3
8
9
± o
4- 2.
- 1
- 5
- 3
4- 1
- 2
- 1
4- 2
4- 6
4- 4
4- 7
+ 9
4- 4
4- 5
4- 6
4- 5
4- 8
4~ 7
4-10
+ 11
4- 9
4-12
4- 8
4- 2
± o
- 3
- 3
- 4
4- 1
- 2
± 0
4- 3
4- 5
4- 2
4- 2
4- 3
4- 4
4- 9
4- 7
4- 7
4- 2
4- 2
4- 8
4- 3
4- 3
4-10
4-12
±
444±
44+
4db
+
4444-
GO
4-44
129
4-105
97
+-73
70
4-22
4-
+
+
±
4+
+
±
+
+
+
4db
44444+
+
1.0
.4
2.5
4- 1.8
5.4
4 -4 .4
4.0
4 -3 .0
4- 8
± o
4-12
db o
4-12
± o
i
0
6
6
1
o
1
5
6
3
1
2
2
4
3
3
1
1
2
o
3
4
5
8
3
2.9
.9
4- 8
± 0
N um ber o f points by which each index n u m bet rose (4 ) or fell (—) in each successive year.
!
Year.
1891......................................................................
1892......................................................................
1893......................................................................
1894......................................................................
1895......................................................................
1896......................................................................
1897......................................................................
1898......................................................................
1899......................................................................
1900......................................................................
1901......................................................................
1902......................................................................
1903......................................................................
1904......................................................................
1905......................................................................
1906.....................................................................
1907......................................................................
1909....................................................................
1910......................................................................
1911......................................................................
1912......................................................................
1913......................................................................■
25 com
242 to 261 1145 com- | 50 com
25 com
commod moditiesJ modities. 40 com modities, modities,
modities. first list.
second
ities.
list.
-
1 i
6
±o
-1 0
- 2
- 4
± o
4- 3
4- 9
4- 9
- 2
4- 4
4* I
- 1
4* 3
4- 7
4- 7
- 8
4- 3
+ 5
- 4
4- 4
± 0
- 1 I
- 7
- 1
- 9
- 3
- 4
± o
4- 4
4-10
4- 8
- 1
4- 4
±o
± o
4- 2
4- 6
4- 8
- 9
4- 3
4- 7
- 1
4- 4
- 3
db 0
- 9
± o
-1 1
± 0
- 7
4- 2
4* 6
4444444444±
8
9
3
7
1
1
2
5
9
7
7
3
6
9
0
4- 1
- 9
- 4
- 8
4- 2
- 7
4- 1
+ 6
+ 13
4- 7
+ 1
+ 6
- 4
± o
+ 4
+ 6
+10
- 9
+ 6
+ 6
- 6
+ 7 1
- 3
I
- 3
— 9
±o
-1 1
+ 3
- 7
+ 2
+ 6
+11
+ 6
_ 2
+ 5
+ 2
+ 4
+ 1
+ 7
+ 2
- 8
+ 9
± o
- 4
+11
+ 2
+ 5
— 6
- 5
-1 1
- 3
_ 8
- 1
+ 6
+ 13
+ 6
- 2
+ 10
±o
- 7
+ 5
+ 7
+ 10
-1 0
+ 6
+ 6
- 3
+ 7
— 5
38
THE M A K IN G AND USING OF INDEX NUM BERS.
Now, these six index numbers, large and small, certainly have a
strong family likeness. The great movements of American prices
from 1890 to 1913 stand out boldly in them all— the heavy fall of
prices in 1890-1896, the distinctly greater rise in 1896-1907, the
sharp decline in 1908, the recovery in 1909, and the wavering course
Chart 5 .— GENERAL-PURPOSE IN D E X NUM BERS, INCLUDING 25, 50, AND 242 COMMOD
ITIES, B Y YE A R S, 1890 TO 1913.
(Based on Table 6.)
in 1910-1913. If index numbers could pretend to nothing more
than to show roughly the trend of price fluctuations, then it would
indeed matter little which of these series were used. Either of the
sets including only 25 commodities would serve that limited purpose
as well as the set containing nearly ten times as many commodities,
though doubtless the longer lists would command more confidence.
METHODS USED IN M A K IN G INDEX NUM BERS.
39
But the very success with which index numbers, even when made
from scanty and dissimilar data, bring out the broader features of
price movements encourages one to hope, from this device, for more
than an indication of the direction and a rough approximation to the
degree of change* Instead o f concluding that an easy compilation,
based on a few series of quotations “ will do,” we may hope that
careful work covering a wide field will enable us to improve upon our
first results and attain measurements that have a narrow margin of
error.
When we make these more exacting demands upon our six index
numbers we attach importance to the fact that their general similarity
does not preclude numerous differences of detail. For example, two
series indicate that prices rose in 1891, one indicates that prices did
not change, and three indicate a fall; three put the lowest point in
1896, one in 1897, and two make the price level the same in these
years; one series shows a rise in 1901, five show a fall; in 1913 again
one series indicates a rise of prices, three indicate a fall, and two indi
cate no change; the general level of prices in the final year is made
to vary between an average rise of 30 per cent and one of 42 per cent
above the level of 1890-1899; there is also a difference in steadiness,
the small series fluctuating through a wider range than the large
ones, etc.
To what are these discrepancies due? Are they discreditable to
the large series, or to the small ones, or to neither set ? Can they be
accounted for except as the results of random differences in sampling ?
If an index number made from the wholesale prices of 25, or 50,
or 250 commodities can measure approximately the changes in all
wholesale prices, it must be because the known fluctuations in the
prices of these selected commodities are fair samples of the unknown
fluctuations in the_ prices of the vastly larger number of other com
modities for which quotations are not collected. Now if (1) the
price fluctuations of each commodity that is bought and sold were
strictly independent of the price fluctuations of every other com
modity, and if (2) each commodity had just the same importance as
an element in the general system o f prices as every other commodity,
then any series of price quotations collected at random would be a
fair sample for determining the average changes in the wholesale
prices o f commodities in general. Of course, the larger the number
of commodities included, the more trustworthy would be the index
number. In Table 6f for example, the first index n amber would be
adjudged the best, and the divergencies between it and its fellows
would be held to result from the scantier material from which the
latter are made.
In fact, however, the situation is by no means so simple, because
neither o f the above-mentioned conditions holds true. Commodities
are far from being all of the same importance as elements in the whole
system of prices. With the complications arising from this fact the
section on the problems of weighting will deal. Neither are the price
fluctuations of different commodities independent of each other. On
the contrary, the price changes of practically every commodity in the
markets o f the whole country are causally related to the changes in
the prices of a few or of many, perhaps in the last resort of all, other
commodities that are bought and sold. Most of these relations are
40
TH E M A K IN G AND USING OF INDEX NUMBERS.
so slight that they can not be traced by statistical methods. But
certain bonds are so close and so strong that they establish definite
groups of related prices which fluctuate m harmony with one another
and which differ in definable ways from the fluctuations of other such
groups. The present task is to show the existence of these groups
and the effects which they exercise upon index numbers.
First, the price fluctuations of a raw material are usually reflected
in the prices of its manufactured products. Hence to quote in some
cases both the raw material .and several of its finished products, and
to quote in other cases the raw material alone, assigns certain groups
of related priced a larger influence upon the results than is assigned
the other groups. When the aim is to secure a set of samples which
fairly represent price fluctuations as a whole, the existence of these
groups must be taken into account. Neglect on this score may give
a misleading twist to the final index numbers. A celebrated case in
point is that of the Economist index number in 1863-1865. Out of
the 22 commodities included in the Economist’s list as then consti
tuted 4 consisted of cotton and its products. Hence when the
blockade of Southern ports during the Civil War raised the price of
cotton, the Economist index numbers grossly exaggerated the aver
age rise in the price level, as appears from the following comparison
between the Economist’s results for 1860-1865 and the corrresponding
English figures compiled by Sauerbeck:40
Year.
1860....................................................................................................................................
1861....................................................................................................................................
1862....................................................................................................................................
1863....................................................................................................................................
1864............................. .*....................................................................................................
1865_______________________ _______________________ __________________________
Economist
Sauerbeck’s
index number index number
(prices in
(prices in
1860=100).
1860=100).
100
102
109
136
145
136
100
100
106
109
112
106
Directly opposing the relations which unite the prices of finished
goods with the prices of their raw materials is a second set of influences
which make the price fluctuations of manufactured goods considered
as a group characteristically different from the price fluctuations
of their raw materials considered as a separate group. Table 7
presents several sets of index numbers designed to throw these
characteristic differences into high relief. The first two columns
compare the relative prices of the 49 raw materials quoted by the
Bureau of Labor Statistics in 1913 and of the 183 to 193 more or less
manufactured commodities in its list.41 The second pair of columns
contains index numbers made from the prices of 20 raw materials and
of 20 products manufactured from these same materials.42 Then
4o To make the comparison as fair as possible, both series are here given, not in their original form, but
recomputed on a common basis. See Wholesale Prices, Wages, and Transportation, report by Mr. Aldrich
from the Committe on Finance, Mar. 3, 1893, 52d Cong., 2d sess., Senate Report No. 1394, Part I, pp. 226
and 255.
See Bulletin No. 149, pp. 13 and 14. The differences between the original figures and those given here
are due (1) to the dropping of decimals, (2) to the exclusion of 11 commodities which the Bureau of Labor
Statistics quotes in the years 1908-1913 only, (3) to the computation of the arithmetic means in these years
by the method applied in 1890-1907 in place of the Bureau’s roundabout method.
42 The articles included here are those from which the index number of 40 commodities in Table 6 was
made. For the list, see p. 35 and note.
METHODS USED IN M AKING INDEX NUM BERS.
41
come three columns giving index numbers made from the prices of
five great staples at three successive stages of manufacture: Wheat,
flour, and bread; cotton, cotton yarns, and cotton textiles; wool,
worsted yarns, and woolen textiles; pig iron, steel billets, and steel
tools; hides, leather, and shoes.43 The later sections of the table give
the data for each of these last-mentioned groups separately. These
several comparisons establish the conclusion that manufactured
goods were steadier in price than raw materials. The manufactured
goods fell less in 1890-1896, rose less in 1896-1907, again fell less in
1907-1908, and rose less in 1908-1913. Further, the manufactured
goods had the narrower extreme range of fluctuations, the smaller
average change from year to year, and the slighter advance in price
from one decade to the next.44 It follows that index numbers made
from the prices of raw materials, or of raw materials and slightly
manufactured products, must be expected to show wider oscillations
than index numbers including a liberal representation of finished
commodities.
« For the list of textiles and of tools, see Bulletin No. 99 of the Bureau of Labor, March, 1912, pp. 554-556
and 682-683.
44 Like most generalizations about price changes, these statements are strictly valid only in the case of
averages covering several commodities, but the exceptions are not numerous, even in the case of single
commodities, as detailed study of the wheat, cotton, wool, iron, and leather groups will show.
THE M A K IN G AND USING OF INDEX NUM BERS.
42
T able
7 .—IN D E X NUMBERS MADE FROM THE PRICES OF R A W M ATER IALS
[Data from Bulletin No. 149 of the
(Arithmetic means.
Year.
183 to
193
49
raw man
ma ufac
teri tured
als. prod
ucts.
Twenty
pairs.
Five triplets.
Average
Wheat group.
. . . .
Raw
ma
teri
als.
Man
ufac
tured
goods.
Inter
Raw medi Fin
ma
Wheat
teri ate ished Wheat. flour. Bread.
prod goods.
als.
ucts.
1
2
2
115
1890..........................................................
116
1891.........................................................
1892..........................................................
108
1893.......................................................... . 104
1894.........................................................
93
92
1895..........................................................
•84
1-896..........................................................
88
1897..........................................................
1898..........................................................
94
1899............................................... : ......... 106
112
1900..........................................................
1901..........................................................
111
1902..........................................................
122
1903..........................................................
123
1904..........................................................
120
121
1905.........................................................
127
1906.........................................................
1907..........................................................
133
124
1908..........................................................
1909..........................................................
131
135
1910................. ........................................
1911..........................................................
135
1912..........................................................
145
139
1913..........................................................
112
111
106
106
97
94
92
90
93
101
110
108
111
112
111
115
122
129
121
123
129
124
127
128
113
114
104
99
91
94
85
88
98
114
118
120
127
122
123
127
135
146
135
143
149
144
151
149
112
114
105
103
94
96
92
89
92
103
111
113
118
114
113
117
120
131
124
127
132
127
132
128
125
117
103
95
79
89
87
94
101
111
120
no
123
125
128
132
136
145
130
149
149
135
141
143
119
116
109
100
86
89
88
90
95
107
110
102
110
114
115
115
119
126
117
126
125
115
119
122
108
107
106
105
98
95
95
94
95
98
105
102
103
106
110
114
121
125
120
121
124
120
124
127
119
128
105
90
74
80
85
106
118
95
94
96
99
105
138
135
106
121
132
160
146
131
140
127
121
126
104
89
78
84
91
no
109
88
88
87
90
97
125
122
97
109
119
139
126
112
122
109
101
101
101
101
101
98
97
101
101
101
101
101
101
101
106
110
no
110
113
116
118
118
122
123
. Averages, 1890-1899..............................
1900-1909..............................
1910-1913..............................
Number of points by which prices
rose ( + ) or fell ( —) in—
100
122
139
100
116
127
100
130
148
100
119
130
100
130
142
100
115
120
100
113
124
100
119
136
100
107
117
100
107
120
1890-1896.........................................
1896-1907.........................................
1907-1908.........................................
1908-1913.........................................
Difference between highest and
lowest relative prices.
-3 1
+ 49
- 9
+ 15
61
-2 0
+37
- 8
+ 7
39
-2 8
+61
-1 1
+ 14
66
-2 0
+39
- 7
+ 4
43
-3 8
+58
-1 5
+ 13
70
-3 1
+38
- 9
+ 5
40
-1 3
+30
- 5
+ 7
33
-3 4
+36
+ 11
- 5
86
-3 0
+ 18
+ 10
-1 0
61
Average change from year to year...
5.5
4.0
6.4
4.9
8.4
5. 5
3.1
13.6
11.6
+
+
+
4
13
3
10
26
1.3
METHODS USED IN M A K IN G INDEX NUM BERS.
43
A ND OF M ANUFAC TU R ED GOODS, B Y YE A R S, 1890 TO 1913.
Bureau of Labor Statistics.]
prices in 1890-1899=100.)
Cotton group.
Raw Cot
ton
cot
ton. yams.
Leather group.
Cot
Wool
Steel
Steel
ton Raw Worst
en
Pig
ed
tex wool.
tex iron. bil tools. Hides. Leather. Shoes.
lets.
yarn.
• i
tiles.
tiles.
24
2
16
4
1
11
1
112
143
111
113
99
117
107
111
90
93
94
92
102
93
92
91
77
91
85
89
124
116
111
98
115
94 i
145
113
156
120
123
106
142
121
153
134
135
109 •
156
119
195
133
168
125
148 i 120
165 : 132
117
112
111
109
98
94
95
90
85
91
103
99
100
105
114
107
117
133
116
117
127
125
122
126
132
126
113
102
79
70
71
89
108
111
118
97
101
110
116
127
121
122
118
127
116
108
111
105
122
123
117
110
91
74
73
83
101
107
118
102
112
118
117
125
129
128
118
130
124
116
119
113
111
112
112
109
96
88
87
90
98
100
111
105
106
111
112
119
125
124
121
122
124
120
123
123
131
116
106
96
83
91
88
78
77
134
140
112
155
141
104
124
145
175
125
127
124
112
118
122
142
118
110
95
77
86
88
70
71
145
116
112
142
130
103
112
128
136
122
114
118
100
104
120
107
106
105
103
99
95
96
95
94
101
112
110
115
118
118
128
134
138
134
129
131
123
124
126
100
102
93
80
68
110
87
106
123
132
127
132
143
125
124
153
165
155
143
176
165
158
188
196
101
101
97
97
92
108
95
96
104
109
113
111
113
112
109
112
120
124
119
127
125
121
129
139
106
104
103
101
99
100
101
96
94
95
98
96
96
96
98
106
119
120
114
121
118
116
127
137
100
136
169
100
113
128
100
111
125
100
116
no
100
120
118
100
116
123
100
135
119
100
122
111
100
124 ,
126 ;
100
144
177
100
116
129
100
106
125
-4 1
4-51
-1 8
+30
118
-1 9
+ 41'
—25,
+23
45
-2 2
+38
-1 7
+ 10
48
-6 1
+ 51
- 4
-1 3
62
-4 9
+ 55
-1 0
- 5
57
-2 4 +37
- 3
+ 2
38
-4 3
+ 87
—50
- 3
98
-5 4
+ 48
-1 4
- 2
75
-1 1
+ 42
- 4
- 8
44
-1 3
+68
-1 2
+ 53
128
- 6
+29
- 5
+20
47
- 5
+ 19
- 6
+23
43
18.1
9.8
6.1
9.1
8.1
3.9
17.5
16.0
3.7
14.7
5.0
3.7
1
.
Iron group.
Wool group.
2
2
i
4
3
Year.
Number of com
modities included.
1890.
1891.
1892.
1893.
1894.
1895.
1896.
1897.
1898.
1899.
1900
1901.
1902.
1903.
1904.
1905.
1906.
1907.
1908.
1909.
1910.
1911.
1912.
1913.
Averages, 1890-1899.
1900-1909.
1910-1913.
Number of points
by which prices
rose ( + ) or fell
( - ) in—
1890-1896.
1896-1907.
1907-1908.
1908-1913.
Difference between
highest and low
est relative prices.
Average
change
from year to year.
44
THE M A K IN G AND USING OF INDEX NUMBERS.
Third, there are characteristic differences among the price fluctua
tions of the groups consisting of mineral products, forest products,
animal products, and farm crops. Table 8 presents index numbers
for these four groups. Fifty-seven commodities are included, all of
them raw materials or slightly manufactured products.45 Here the
Ch a r t
6 .—IN D E X NUM BERS OF THE PRICES OF 20 R A W M ATERIALS AN D 20 PRODUCTS
M ANUFACTURED FROM THEM.
45 The lists of commodities are as follows:
Farm crops: Cotton, flaxseed, barley, corn, oats, rye, wTheat, hay, liops, beans, coffee, rice, pepper, tea,
onions, potatoes, cottonseed meal, and jute—18 articles.
Animal'products: Hides, cattle, hogs, sheep, eggs, lard, milk, tallow', silk, and wool—10 articles.
Forest products: Hemlock, maple, oak, white pine, yellow pine, poplar and spruce lumber, together with
turpentine, tar, and rubber—10 articles.
Mineral products: Salt, anthracite coal, bituminous coal, coke, crude petroleum, copper ingots, lead (pig),
pig iron, bar iron, steel billets, quicksilver, silver bars, tin (pig), spelter, zinc, brick, cement, lime, and brim
stone—19 articles.
METHODS USED 11ST M A K IN G INDEX NUM BERS.
45
striking feature is the capricious behavior of the prices of farm crops
under the influence of good and bad harvests. The sudden upward
jump in their prices in 1891, despite the depressed condition of busi
ness, their advance in the dull year 1904, their fall in the year of
revival 1905, their failure to advance in the midst of the prosperity of
7.—IN D E X NUM BERS OF THE PRICES OF W OO L, COTTON, HIDES, W H E A T , A N D
PIG IRON IN THEIR R A W , P A R T IA L L Y M ANUFACTURED, AND FINISHED FORMS.
Ch a r t
(Based on Table 7.)
46
THE M A K IN G AND USING OF INDEX NUMBERS.
level of fluctuation in 1902-1913 than any of the other groups, a level
on which their fluctuations, when computed as percentages of the
much lower prices of 1890-1899, appear extremely violent. Finally,
the prices o f minerals accord better with alternations of prosperity,
crisis, and depression than any of the other groups. And the anom
alies that do appear— the slight rise in three years (1896, 1903,
and 1913) when the tide of business was receding—would be removed
if the figures were compiled by months. For the trend of mineral
prices was downward in these years, but the fall was not so rapid as
the rise had been in the preceding years, so that the annual aver
ages were left somewhat higher than before.46 An index number
composed largely of quotations for annual crops, then, would be
expected at irregular intervals to contradict capriciously the evidence
of index numbers in which most of the articles were mineral, forest, or
even animal products.
T
able
8 .—IN D E X
NUM BERS MADE FROM PRICES OF M IN ER AL, FOREST, AN IM AL,
AND FAR M PRODUCTS, B Y Y E A R S , 1890 TO 1913.
[Data from the Bulletin of the Bureau of Labor Statistics, No. 149J
(Arithmetic means.
Average prices in 1890-1899=100.)
Year.
Mineral
Forest
Animal
products. products. products.
Farm
crops.
Number of commodities included................................ .....................
19
10
18
16
1890............................................ ...............................................................
1891............................................................................................................
1892.............................................................................................................
1893.............................................................................................................
1894.............................................................................................................
189q.............................................................................................................
1896. .1___ '..................................- ..............................................................
1897.............................................................................................................
1898...........................................................................................................
1899.............................................................................................................
1900.............................................................................................................
1901.............................................................................................................
1902.............................................................................................................
1903.............................................................................................................
1904.............................................................................................................
1905.............................................................................................................
1906.............................................................................................................
1907.............................................................................................................
1908.............................................................................................................
1909.............................................................................................................
1910.............................................................................................................
1911.............................................................................................................
1912............................................................................................................
1913.............................................................................................................
119
111
105
98
87
91
92
88
92
117
120
113
119
124
115
123
135
137
118
121
120
120
132
136
107
105
99
98
95
96
94
95
99
112
121
113
123
137
142
149
163
169
151
164
181
172
168
169
106
108
109
116
94
95
82
88
97
105
111
112
128
117
113
121
128
135
126
144
152
131
146
150
119
126
110
105
101
92
76
83
92
96
105
114
120
116
124
116
116
125
124
130
134
151
158
135
Averages, 1890-1899................................................................................
1900-1909.............. - ........................... ...................................
1910-1913......................................... .....................................
Number of points by which prices rose ( + ) or fell ( —) in—
1890-1896............................... ...........................................................
1896-1907............................................................................................
1907-1908............................................................................................
1908-1913............................................................................................
100
123
127
100
143
173
100
124
145
100
119
145
-2 7
+45
—19
+ 18
—13
+75
—18
+18
—24
+53
— 9
+24
—43
+49
— 1
+ 11
Difference between highest and lowest relative prices..................
50
87
70
82
Average change from year to year......................................................
7.0
:
7.4
8.9
8.2
Fourth, there are characteristic differences between the price fluc
tuations of manufactured commodities bought by consumers for
family use and the price fluctuations of manufactured commodities
bought by business men for industrial or commercial use. Such at
46 Compare the monthly figures compiled by the Bureau of Labor Statistics for its group of “ Metals and
implements,” Bulletin No. 149, p. 18. These figures are largely influenced by the relatively stable prices
of 11 different kinds of tools. Monthly data for the 19 mineral products of Table 8 would probably show
even more Mediae between ifanuary and December in these years.
METHODS USED IN M A K IN G INDEX NUM BERS.
47
least is the story told by Table 9. The data employed here are
quotations for 28 articles from the Bureau of Labor Statistics’ list
that rank distinctly as consumers’ goods and 28 that rank as pro
ducers’ goods.47 Though consisting more largely of the erratically
fluctuating farm products, the consumers’ goods are steadier in
Ch a r t
8.—IN D E X NUM BERS OF THE PRICES OF 19 M IN ER AL PRODUCTS AN D
FARM CROPS.
OF 18
The consumers’ goods are bread, crackers, butter, cheese, salt fish, evaporated apples, prunes, raisins,
beef, mutton, pork, molasses, cornstarch, sugar, vinegar, shoes, cotton textiles, woolen textiles, candles,
matches, quinine, furniture, earthenware, glassware, woodenware, table cutlery, soap, and tobacco. The
producers’ goods are bags, cotton yarns, leather, linen shoe thread, worsted yarns, refined petroleum,
barbed wire, builders’ hardware, copper wire, lead pipe, nails, steel rails, tools, wood screws, pine doors,
plate glass, window glass, carbonate of lead, oxide of zinc, putty, rosin, shingles, muriatic acid, sulphuric
acid, malt, paper, proof spirit, and rope.
It will be noticed that a large proportion of the consumers’ goods are subject to very slight manufacturing
processes, notably the foods. Hence the difference between the two index numbers can scarcely be re
garded as merely a fresh contrast between the fluctuations of finished goods and of intermediate products-
48
THE M A K IN G AND USING OF INDEX NUMBERS.
price than the producers’ goods, because the demand for them is less
influenced by changes in business conditions.
T able
9 .—IN D E X NUMBERS MADE FROM TH E PRICES OF CONSUMERS’ GOODS AND
OF PRODUCERS’ GOODS, B Y Y E A R S , 1890 TO 1913.
[Data from Bulletin of the Bureau of Labor Statistics, No. 149.]
(Arithmetic means.
Average prices in 1890-1899= 100.)
Year.
Consum
1 ers’
goods.
!
1890
.......................................................................................................................
1391 ................................................................................................................................................................................................................................................
1892 .................................................................................................................................................
1893
..................................................................................................................................
1894
......................................................................................................................................
1895
.
..................................................................................................................................
1896
......................................................................................................................................
1897 .................................................................................................................................................
1898
.............................................................................................................................................
1899
........................................................................................................................................
1900
...........................................................................................................................
1901
.
..................................................................................................................................
1902...................................................................................................................................................
1903
............................................................................................................................................
1904
..................................................................................................................................
1905
......................................................................................................................................
190-5
....................................................................................................................................
1907
..................................................................................................................................
1908
..................................................................................................................................
1909
..................................................................................................................................
1911.
............................................................................................„........................................
1912
...........................................................................................................................
1912..................................................................................................................................................1
A \Tara<ras:. 1890-1899_______ _________
1900-1909....................................................................: .................................................
1S10-1913 .
..
....................
........................
Number of points bv which prices, rose ( + ) or fell ( — ) in—
i son-18Q7 .
”
______________________________________ __________________ _
|
1897-1907..............................................
1907-190 8
|
1908-191 3
Difference between highest and lowest relative prices.........................................................
Average change from year to year........
Produc
ers’
goods.
112
109
104
108
100
95
91
90
94
98
106
105
108
105
.103
106
110
114
112
114
118
119
118
121
115
111
107
102
92
91
93
89
93
107
117
113
114
114
114
117
124
133
119
118
126
125
125
123
100
108
119
100
118
125
-2 2
+24
T i
-2 6
+ 14
-1 4
+ 4
31
44
3.4
4.7
Other groups of related prices having specific peculiarities of fluc
tuation doubtless exist,48 but the analysis has been carried far enough
for the present purpose. That purpose is to show how the existence
of groups of prices which fluctuate in harmony with each other and
at variance with other groups affects index numbers in general and
in particular the six index numbers for the United States given in
Table 6. T o apply the knowledge gained from the preceding analysis
to the explanation of the differences among these six index numbers
is not difficult when once the commodities included in each index
number have been classified on the basis of the groups which have
been examined.
First, the list of conpanodities used by the Bureau of Labor Sta
tistics includes 29 quotations for iron and its products, 30 quotations
for cotton and its products, and 18 for wool and its products, besides
8 more quotations for fabrics made of wool and cotton together. On
the other hand, it has but 7 series for wheat and its products, 8 for
coal and its products, 3 for copper and its products, etc. The iron,
48 For example, there is evidence that the products of industries characterized by a large measure of
concentration in business control are steadier in price than products of industries characterized by un
hampered competition.—See W . C. Mitchell, Business Cycles, pp. 462-464.
METHODS USED IN M A K IN G INDEX NUM BERS.
49
cotton and wool groups together make up 85 series out of 242, or
35 per cent of the whole number. The same three groups furnish 36
(or 25 per cent) of the 145*series in the second index number in Table 8.
Chart 9 .—IN D E X NUMBERS OF THE PRICES OF M ANUFACTURED GOODS USED FOR
F AM ILY CONSUMPTION AND FOR IN D U STR IAL PURPOSES.
(Based on Table 9.)
Does this large representation of three staples distort these index
numbers— particularly the bureau’s series where the disproportion is
greatest? Perhaps, but if so the distortion does not arise chiefly
from the undue influence assigned to the price fluctuations of raw
cotton, raw wool, and pig iron. For, contrary to the prevailing
impression, the similarity between the price fluctuations of finished
products and their raw materials is less than the similarity between
3 3 2 2 6 °— 21— B u ll. 284--------4
50
THE M A K IN G AN © W I N G OF INDEX NUMBERS*
the price fluctuations of finished products made from different mate
rials. Such at least is the testimony of Table 7. As babies from
different families are more like one another than they are like their
respective parents, so here the relative prices of cotton textiles,
woolen textiles, steel tools, bread, and shoes differ far less among
themselves than they differ severally from the relative prices of raw
cotton, raw wool, pig iron, wheat, and hides.49 Hence the inclusion
of a large number of articles made from iron, cotton, and wool affects
an index number mainly by increasing the representation allotted
to manufactured goods. What materials those manufactured goods
are made from makes less difference in the index number than the
fact that they are manufactured. To replace iron, cotton, and
woolen products by copper, linen, and rubber products would change
the result somewhat, but a much greater change would come from
replacing the manufactured forms of iron, cotton, and wool by new
varieties of their raw forms.50
This similarity among the price fluctuations of manufactured goods
arises from the fact demonstrated by Table 7 that such articles are
relatively steady in price. Does knowledge of this steadiness assist
in explaining the differences among the six American index numbers
of Table 6? To answer we must find the proportions of raw and
manufactured commodities included in each index number. Classi
fication along this line is rather uncertain in many cases, but the
results shown in the following schedule, if not strictly correct, are at
least uniform in their errors.
T a b l e 1 0 . —N UM BER
AND PER CENT OF R A W AND M AN U FACTU R ED COM M ODITIES
INCLUDED IN T H E SIX IN D E X NUMBERS OF T A B L E 6.
Number of—
Index number.
First.......................................................................................
Second...................................................................................
Third......................................................................................
Fourth...................................................................................
Fifth.......................................................................................
Percentage of—
Total
number
Man u- ; Raw
Manu
Raw
of com
modities. commod factured commod factured
commod
commod
ities.
ities.
ities.
ities.
242
145
50
40
25
25
49
36
26
17
19
10
193
109
24
23
6
15
20
25
52
43
76
40
80
75
48
57
24
60
« A compilation of the differences among the relative prices in question taken seriatim for each of the 24
years 1890-1913 yields the following results:
Average differences between the relative prices of—
Raw cotton and cotton textiles...................................................................................20.7 points.
Raw wool and woolen textiles..................................................................................... 8.9 points.
Pig iron and steel tools................................................................................................. 14.0 points.
Wheat and bread............................................................................................................ 15.0 points.
Hides and shoes.............................................................................................................. 31.6 points.
Average.................................................................................................................. 18.0 points.
Cotton textiles and woolen textiles...............................................................................5.3
Cotton textiles and steel tools..................................................................................... ...7.8
Cotton textiles and bread............................................................................................. ...6.9
Cotton textiles and shoes.................................................................................................6.7
Woolen textiles and steel tools............................................................. ........................6.1
Woolen textiles and bread..............................................................................................7.3
Woolen textiles and shoes............................................................................................ ...8.1
Steel tools and bread.................................................................................................... ...9.4
Steel tools and shoes...................................................................................................... ...9.6
Broad and shoes.................................................................................................................4.7
points.
points.
points.
points.
points.
points.
points.
points.
points.
points
Average.................................................................................................................. 7.2 points.
m While the fluctuations in the prices of manufactured goods are generally slighter than those in the
prices of raw materials, they are nevertheless violent at times, as in the case of cotton yarns and cotton
textiles during the Civil W ar. ( See p . 40.)
METHODS USED IN M A X IN O IKDEX NUM BERS.
51
On this showing the Bureau of Labor Statistics series ought to be
the steadiest, and the second series the next steadiest— and so they are,
as the summaries at the bottom of the columns in Table 6 show.
With the smaller index numbers, however, the rule does not work
well, for the most variable of all— the sixth— has a larger per cent of
manufactured goods than the other three. Moreover, number four
is more variable than number three, though it has relatively more
manufactured goods. But the preceding studies of different groups
throw further light upon the matter.
It has been found that among manufactured commodities those
bought for family consumption are steadier in price than those bought
for business use. To take account of this factor the manufactured
goods in the several series are classified as primarily consumers ’
goods, primarily producers’ goods, or as bought in large measure by
both classes of purchasers.
T able
1I.-O L A S S IF IC A T IO N OF T H E M ANU FACTURED COMMODITIES INCLUDED IN
TH E SIX IN D E X NUMBERS OF TA B L E 6.
Number of—
Index number.
First..............................
Second..........................
Third............................
Fourth..........................
Fifth.............................
Sixth............................
Per cent of—
Both con
Both con
Con
Pro
Con
sumers’
Pro
sumers’
Manu
Manu
ducers’ and pro factured
sumers’
ducers’ and pro
factured sumers’
com
com
com
com
ducers’
articles. modities. modities.
articles. modities. modities. ducers’
com
com
modities.
modities.
193
109
24
23
6
15
108
51
11
10
3
4
73
47
12
12
3
11
12
11
1
1
80
75 :
48
57
24
60
45
35
22
25
12
16
30
32
24
30
12
44
5
8
2
2
Here it does turn out that the two series {numbers four and six)
which are highly variable despite the inclusion of many manufactured
goods have relatively more of those manufactured goods which as
a group are most variable. So far as this factor counts, then, it
counts toward clearing up the contradiction pointed out in the
preceding paragraph. It also brings out a further reason for the
comparative stability of the first two series.
The one remaining form of analysis suggested above seems easy
to apply. In the schedule below, raw and slightly manufactured
commodities like those used in Table 8 are distributed among four
groups according as their constituents come chiefly from mines,
forests, animal sources, or cultivated fields. There is little doubt
about the classification here, but there is much doubt about the
significance of the results as applied to our six index numbers. The
figures in the schedule are either such small percentages of the whole
number of series that they can not exercise much influence upon the
results, or such small numbers that they can not claim to be typical
of their groups. Further, the second part of the schedule shows that
there is less difference among the six index numbers than appears at
first sight in the proportions of the raw and slightly manufactured
commodities which consist of mineral, forest, animal, and farm prod
ucts. Hence it is not surprising that efforts to account for the
divergences in Table 6 by appealing to this schedule and to Table 8
52
THE M A K IN G AND USING OF INDEX NUMBERS.
accomplish little, especially for the smaller index numbers. This
much does appear regarding the first two series: Whenever mineral
products and farm crops move sharply in opposite directions the
Bureau of Labor Statistics' index diverges from its mate in harmony
with mineral products, while the series of 145 commodities bends
toward the agricultural products— which is what should happen
according to the schedule.
12.— FARM, AN IM AL, FOREST, AND M IN ER AL PRODUCTS IN R A W OR SLIG HTLY
M ANUFACTURED FORM, INCLUDED IN TH E SIX IN D E X NUMBERS OF T ABL E 6.
T able
Per cent of the whole number consist
ing of—
Number of—
Index
number.
First............
Second........
Third..........
Fourth........
Fifth...........
Sixth...........
Total
num
ber
Raw
and
of
com slightly Farm
modi manu crops.
ties.
fac
tured
goods.
242
145
50
40
25
25
74
57
30
19
23
18
Ani
mal
prod
ucts.
Forest
prod
ucts.
15
10
8
6
5
5
12
10
3
1
2
1
18
18
10
6
7
5
Raw
and
Min
slightly
eral
manu
prod
fac
ucts.
tured
goods.
Farm
crops.
Ani
mal
prod
ucts.
Forest
prod’
uets.
30
39
60
48
92
72
7
12
20
15
28
20
6
7
16
15
20
20
5
7
6
3
8
4
29
19
9
6
9
7
•
Min
eral
prod
ucts.
12
13
18
15
36
28
Per cent of the raw and slightlymanufactured commodities con
sisting of—
inU.UA llulllUvl •
Farm
crops.
Fifth..................................................................................................................
25
31
33
32
30
28
Animal
prod
ucts.
Forest
prod
ucts.
20
18
27
32
22
28
16
18
10
4
9
5
Mineral
prod
ucts.
39
33
30
32
39
39
Two practical conclusions of moment to both the makers and the
users of index numbers are established by this section. (1) To make
an index number that measures the changes in wholesale prices at
large, samples must be drawn from all the various groups that behave
in peculiar ways. (2) In using an index number made by others, one
must study the list of commodities included critically with these
groups in mind to know what it really does measure.
The first conclusion seems to contradict a rule often practiced and
sometimes preached. Most of the middle-sized index numbers are
confined to raw materials and slightly manufactured goods. Most
of the small index numbers are confined to foods alone. The makers
of both sets argue that their series are more “ sensitive” and therefore
better measures of price changes than the larger series, which are
“ loaded dow n” with a mass of miscellaneous manufactured goods.
And many users of index numbers seem to prefer a series like Sauer
beck's with only 45 commodities, or even one like the Annalist's
with only 25 commodities, to one like that of the Bureau of Labor
Statistics with five or ten times the number.
METHODS USED IK M AKIN G IN D EX NUM BERS.
53
Critics who take this stand usually assume tacitly that the purpose
of an index number is to serve as a “ business barometer,” or to
measure changes in “ the cost of living.” If these aims were always
clearly realized by the critics and clearly stated for their readers the
room left for differences of opinion would be narrow. In Table 6
the index number with 145 commodities shows itself a more sensi
tive and on the whole more faithful barometer of changing business
conditions during the 24-year period from 1890 to 1913 than the
official series with 242 commodities,51 and the preceding analysis shows
that the sluggishness of the larger index number is due chiefly to its
proportion of manufactured goods. For this particular purpose, then,
a series modeled after Sauerbeck’s has strong claims to preference
over one including a larger number of commodities* Indeed, in the
light of the preceding discussion one might carry the process of exclu
sion much further and throw out of the business barometer not only
manufactured goods but also all farm crops, on the ground that their
prices depend on the eccentricities of the weather, and most forest
products, on the ground that their prices in the period covered by
Table 6 were rising so fast as to obscure the effects of bad times, etc.
But clearly such exclusions, while they might make the resulting fig
ures more responsive to changes in business conditions, would also
make the figures less acceptable as a measure of changes in prices as a
whole. The sluggish movements of manufactured goods and of con
sumers’ commodities in particular, the capricious jumping of farm
products, etc., are all part and parcel of the fluctuations which the
price level is actually undergoing. Consequently, an index number
which pretends to measure changes in the general level of prices can
not logically reject authentic quotations from any of these groups.
Every restriction in the scope ol the data implies a limitation in the
significance of the results.
As for the small series made from the prices of foods alone or from
the prices of any single group of commodities, it is clear that however
good for special uses they may be, they are untrustworthy as generalpurpose index numbers. Table 13 shows what differences are likely
to appear at any time between series confined to foods and series
covering a wider field. The general-purpose indexes are taken from
Table 6, two of the food indexes include the commodities quoted by
the Annalist index number and by the Gibson index number as now
constituted; the third food index is the bureau’s old series for foods,
with decimals dropped and new arithmetic means for 1908-1913.
si Compare p. 36.
54
TH E M A K IN G AND USING OF INDEX NUMBERS.
T able 1 3 .—IN D E X NUM BERS OF THE PRICES OF FOODS, AN D GENERAL-PURPOSE
IN D E X NUM BERS, B Y Y E A R S, 1890 TO 1913.
[Data from Bulletin, of the Bureau of Labor Statistics, No. 149.]
(Arithmetic means.
Average prices in 1890-1899=100.)
General-purpose index ,
number from Table 6.
Year.
242 to 261
com
modities.
25 com
modities,
first list.
Index numbers of the prices of
foods.
25 com
modities,
Annalist,
list.
22 com
modities,
Gibson
list.
48 com
modities,
Bureau of
Labor Sta
tistics list.
1890......................................................................
1891......................... .............................................
1892.......................................................................
1893.......................................................................
1894......................................................................
1895......................................................................
1896.......................................................................
1897.......................................................................
1898.......................................................................
1899 ....................................................................
1900 ....................................................................
1901.......................................................................
1902.......................................................................
1903 ....................................................................
1904.......... ............................................................
1905.......................................................................
1906.......................................................................
1907 ....................................................................
1908 ..................................................................
1909................. .....................................................
1910.......................................................................
1911.......................................................................
1912......................................................................
1913.......................................................................
113
112
106
106
96
94
90
90
93
102
111
109
113
114
113
116
123
130
122
125
130
126
130
130
115
112
103
103
92
95
80
90
96
107
113
111
116
118
122
123
130
132
124
133
133
129
140
142
109
119
108
116
102
95
81
84
92
93
99
105
117
107
109
110
115
120
126
134
137
131
143
139
109
121
108
110
98
94
81
87
96
96
100
106
118
107
115
114
111
121
128
127
137
134
147
139
112
116
104
110
100
95
84
88
94
98
104
106
111
107
107
109
113
118
122
125
129
127
135
131
Averages, 1890-1899..........................................
1900-1909..........................................
1910-1913..........................................
Number of points by which prices rose (4- )
or fell ( —)in—
1890-1896......................................................
1896-1907......................................................
1907-1908......................................................
1908-1912 ..................................................
1912-1913......................................................
100
118
129
100
122
136
100
114
138
100
115
139
100
112
131
Difference between highest and lowest
relative prices.................................................
Average change from year to year................
—
44±
23
40
8
8
0
40
4.0
+
—
+
4-
27
44
8
16
2
54
5.0
—
4+
4—
29
40
6
17
4
63
7.1
—
+
+
4—
28
40
7
19
8
66
7.3
—
444—
28
34
4
17
4
51
5.0
The three index numbers for foods agree better than might have
been expected in view of the dissimilarity of the lists of commodities
which they quote and the brevity of two of the lists.52 The bureau
52 Of the 56 articles included altogether, only 11 are common to all three lists. The Gibson list has 8
commodities and the Annalist list has 4 commodities classified by the bureau with farm products instead
of with foods, while the bureau has 34 foods not quoted by Gibson and 27 not quoted by the Annalist.
Even the two short lists have only 15 articles in common, while Gibson has 7 articles not quoted by the
Annalist, and the Annalist has 10 articles not quoted by Gibson.
For the Bureau’s list see Bulletin No. 149, pp. 90-107.
The Annalist list runs—oats, cattle, fresh beef, salt beef, hogs, bacon, salt pork, lard, sheep, mutton,
butter (two kinds), cheese, coffee, sugar, wheat flour (two kinds), rye flour, corn meal, rice, beans, potatoes,
prunes, evaporated apples, and codfish.
The Gibson list is—barley, corn, oats, rye, wheat, cattle, hogs, sheep, butter, coffee, wheat flour (two
kinds), com meal, bacon, fresh beef, salt beef, hams, mutton, sugar (two kinds), tea, and potatoes.
M E IH P O S USED IN M A K IN G INDEX NUM BERS.
55
series is rather steadier than the others, because of the larger propor
tion of manufactured products included in it; but this series and that
of the Annalist invariably agree about the direction in which prices
Ch a r t
10.—IN D E X NUM BERS OF THE PRICES OF 25 FOOD PRODUCTS AND OF 25 MIS.
CELLANEOUS COMMODITIES.
(Based on Table 13.)
are moving,53 and the Gibson figures agree with the other two series
in 19 years out of the 24. On tne other hand, the three food indexes
53 Even in 1903-4 the bureau’s figures record a slight advance of prices in harmony with the Annalist
figures, though this advance is confined to the decimal columns and disappears when the decimals are
rounded off.
56
THE M A K IN G AND USING OF INDEX NUMBERS.
often contradict the evidence of the two general-purpose index num
bers in a striking fashion. Such contradictions occur in 1890-1891,
1892-1893,1900-1901,1902-1903, 1907-1908, and 1912-1913. These
differences are due chiefly to a contrast in the years mentioned between
business conditions and harvest conditions. They parallel the differ
ences in Table 8 between the index numbers of mineral products and
those of fta n crops, or farm crops and animal products taken together;
for the food indexes are made up almost wholly from the pieces of vege
table crops, food animals, and their derivatives.54 A food index num
ber, then, is likely at any time to give a wrong impression regarding
the shifting of prices in general and is especially treacherous as a busi
ness barometer. Nor can such an index when made from wholesale
prices be trusted to show changes in the “ cost of living” ; for living
expenses are made up of retail prices, and fluctuations in retail prices
do not always follow closely those in the wholesale markets.
But while it is clear that an index number intended to measure
fluctuations in “ the general level of prices” should grant due repre
sentation to the various groups of prices that behave in specific ways,
it is not possible to give a definitive list of these groups. For our
knowledge of the interrelations among prices even in the recent past
is very limited. Moreover, a change in the social conditions under
which business is done may at any time produce new groupings of
commodities important to the maker of index numbers, or may cause
old groups to fluctuate in novel ways. For example, the distinction
between commodities over which the Government assumed some
form of price control and commodities whose prices were left unre
stricted became of first importance in the summer of 1917. After
July the controlled prices dropped,* and while they advanced again in
the latter part of 1918, they did not again attain the high level at
which they stood when the price control began. Uncontrolled prices,
on the contrary, which stood lower than the other group in July, 1917,
advanced month by month until the armistice was signed.55 Forest
products in 1915-1918 illustrate the way in which a group may change
its characteristic price behavior. The demand for lumber has been
declining jerkily in the United States since 1909, primarily because of
the increased use of cement for building. Further, the terms on
which many large lumber holdings are financed compel the owners to
cut a,nd market their timber as fast as possible. Finally, in 1917-18
the War Industries Board discouraged the construction of buildings
that were not called for by the military program. Under these cir
cumstances, the price of forest products lagged behind most classes
64 The exceptions are salt and soda, and of these articles the Annalist and Gibson quote neither.
55 See the tables in Government Control Over Prices, by Paul W . Garrett, War Industries Board Price
Bulletin, No. 3.
The following index numbers, while not covering the whole ground, bring out the main point. One
series shows the fluctuations of 5S6 commodities that were subjected to price control at some time during
American participation in the war; the second series shows the fluctuations of 780 commodities that were
left uncontrolled in price. Since the practice of “ fixing” prices did not begin until several months after
the declaration of war (April, 1917), and was extended gradually month by month until the signing of the
armistice (November, 1918), the “ controlled” list contains many articles that remained uncontrolled until
latein 1918. The two series therefore minimize rather than exaggerate the differences between the behavior
of prices that were controlled earlier in the war and prices that were left to find their own levels. That
this understatement is not more serious arises from the fact that the Government naturally took the most
important (and therefore most heavily weighted) commodities under control at an early date. It may be
METHODS USED IN M A K IN G INDEX NUM BERS.
57
of commodities in the wartime rise.56 To give another illustration,
rubber is rapidly passing from the group of forest products to the
group of cultivated crops. These cases give force to the warning that
the groupings with which the economic statistician deals do not
always rest on permanent foundations. It would be as unwarranted
to draw up a list of groups that should be represented in index num
bers for all periods as to draw up a list of groups to be represented for
all purposes. In every case in which an investigator plans to measure
changes in the general level of prices he should canvass his particular
field to see whether there are not hitherto unrecognized groups of
commodities that fluctuate in similar ways, and then try to represent
each group in the due measure of its importance. Such investigations
may add much, not only to the accuracy of index numbers but also
to our knowledge of the interrelations among price fluctuations.
In most large index numbers the commodities quoted are divided
into several classes; but these classes seldom have economic signifi
cance or even logical consistency. Among the nine groups recognized
by the Bureau of Labor Statistics, for example, one group, “ Farm
products,” emphasizes the place of production; four groups, “ Food,
etc.,” ftFuel and lighting,” “ Lumber and building materials,” and
“ House-furnishing goods,” emphasize the use to which commodities
are put; three groups apply a double criterion, use and physical
character of the goods, namely, “ Cloths and clothing,” “ Metals and
implements,” and “ Drugs and chemicals” ; the remaining group is
frankly styled “ Miscellaneous.” Such a classification is not without
usefulness, for there doubtless are readers especially interested in the
prices of, say, all things that are raised on farms, and others who care
especially about the prices of things used to furnish houses, or things
that can be classed together as drugs and chemicals whether they are
used chiefly as medicines or to make farm fertilizers. But if a classi
fication of this empirical character is maintained, it might with
advantage be accompanied by a classification that throws more light
upon the workings of the complex system of prices.
pointed out also that the commodities early brought under control were articles that, as a group, had risen
more than the average in price before we entered the war.
Index numbers of commodities that were and of commodities that were not subjected to price control by the
Government during the war with Germany.
[From War Industries Board Price Bulletin No. 3.]
(Relatives made from weighted aggregates of actual prices.
Year and month.
Uncon
Controlled
commod trolled com
modities.
ities.
1917
Tannarir
February
............. 1.....................
Mn.rr*h
__ 1 __________
183
April......................................
192
May
.................................
201
June. .
............................
209
July
204
A u g u st.................................
205
September............................
198
October.................................
200
N ovem ber
.
193
December
....
146
149
]52
160
162
163
167
172
174
Average prices in July, 1913, to June, 1914=100.)
Year and month.
1918
January...............................
February..............................
March.....................................
April......................................
M ay........................................
June.......................................
July........................................
August....................; ............
September............................
October.................................
N ovember.............................
December..............................
Controlled
Uncon
commod trolled com
ities.
modities.
195
198
197
196
192
189
195
199
204
201
200
204
178
180
182
187
189
191
194
195
199
201
200
197
5CSee R. B. Bryant, The Prices of Lumber, War Industries Board Price Bulletin, No. 43, and Homer
Hoyt, The Prices of Building Materials, War Industries Board Price Bulletin, No. 6.
•
58
TH E M A K IN G AND USING OF INDEX NUM BEBS.
Another interesting experiment has recently been made by the
Price Section of the War Industries Board. This section was able
to collect quotations for so large a number of price series (1366 in
form to be used in the index number) that it attempted to classify
its commodities according to industries by which they are manu
factured. The advantage of this arrangement is that many users
of index numbers desire to follow the fluctuations of the prices that
are paid for materials and received for products in different lines of
business and to compare fluctuations in one line with those in others.
There are many industries in which the plan works well, because the
demarcation between industries follows, at least roughly, commodity
lines; for example, in the cotton, woolen, silk ana leather trades.
But many commodities are used in such a variety of industries, and
many industries use such a variety of commodities, that the classifier
is forced to resort at times to other criteria, such as the physical
characteristics of commodities, their uses, or their sources of supply.
Probably the most illuminating way of presenting an index num
ber that aspires to cover the whole field of prices at wholesale would
be to publish separate results for the groups that have characteristic
differences of price fluctuations, and then to publish also a grand total
including all the groups. The groups to be recognized and the distri
bution of commodities among them is a difficult matter to decide.
But, as matters stand, the most significant arrangement seems to be
(1) a division of all commodities into raw and manufactured prod
ucts; (2) the subdivision of raw commodities into farm crops and
animal, forest, and mineral products; (3) the subdivision of manu
factured products according as they are bought mainly for personal
consumption, mainly for business use, or largely for both purposes.57
This classification is based upon differences among the factors affect
ing the supply of and the demand for commodities that belong to the
several groups— that is, upon differences among the factors which
determine prices. If we wish our index numbers to help toward an
understanding of changes in the price level, a classification along
such causal lines seems to be the most promising line of progress.
Where means permit, it is desirable to supplement this general
scheme by a series of special indexes for classes of commodities that
possess interest for whatever reason. These supplementary indexes
would not rest on classifications which include all the commodities,
and they might, therefore, employ many different criteria and employ
each one only in those cases in which it was significant. Some
commodities might appear in several of the special indexes, and others
might appear in none. There need, then, be no artificial forcing of a
criterion upon facts which it does not fit,, and no hesitation about
presenting any classes that merit separate attention.
Large index numbers are more trustworthy for general purposes
than small ones, not only in so far as they include more groups of
related prices, but also in so far as they contain more numerous
samples from each group. What is characteristic in the behavior of
the prices of farm crops, of mineral products, of manufactured wares,
of consumers’ goods, etc.— what is characteristic in the behavior of
any group of prices— is more likely to be brought out and to exercise
its due effects upon the final results when the group is represented by
57
Since the first edition of this bulletin appeared, the Federal Reserve Board has adopted this suggestion
with interesting results. In its monthly bulletin the board publishes the index number compiled by the
Bureau of Labor Statistics recast into tne six groups mentioned.
METHODS USED IN M A K IN G INDEX NUMBERS
59
10 or 20 sets of quotations than when it is represented by only one or
two sets. The basis of this contention is simple: In every group that
has been studied there are certain commodities whose prices seldom
behave in the typical way, and no commodities whose prices can be
trusted always to behave typically. Consequently, no care to in
clude commodities belonging to all the important groups can
guarantee accurate results, unless care is also taken to get numerous
representatives of each group.
Even here the matter does not end. The different groups that have
been discussed, the other groups that might have been discussed,
and the commodities that are included within the several groups
differ widely in importance as elements in the system of prices. To
these differences, and to the methods of making them count in index
numbers, we must now turn.
6. PROBLEMS OF WEIGHTING.
It is customary to distinguish sharply between “ simple” and
“ weighted” index numbers. When an effort is made to ascertain
the relative importance of the various commodities included, and to
apply some plan by which each commodity shall exercise an influence
upon the final results proportionate to its relative importance, the
index number is said to be weighted. When, on the contrary, no such
effort is made, but every commodity is supposedly allowed just the
same chance to influence the result as every other commodity, the
index number is said to be unweighted, or simple.
This expression, however, that “ every commodity has just the
same chance to influence the result as every other commodity”
conveys no clear meaning. It is better to think of all index numbers
as weighted, for so they are whether their maker knows it or not, and
to ask whethei\the scheme of weights is good or bad. For example,
in Bradstreet’s index the influence of every article upon the result
varies as its price per pound happens to be large or small.58 Again,
the decisive objection to making index numbers by merely adding
the ordinary commercial quotations for different articles is that these
nominally simple series are in fact viciously weighted series.59 Nor
does the substitution of relative prices for actual prices assure an
“ equal chance” to every article. For instance, in its famous report
of 1893, the Senate Committee on Finance presented three wholesaleprice index numbers— one simple and two weighted; but in the simple
series it included relative prices for 25 different kinds o*f pocketknives,
giving this trifling article more than eight times as many chances to
influence the results as they gave wheat, corn, and coal put together.
Finally, even if one series of relative prices, and only one, be accorded
each commodity, it does not follow that equal percentages of change
in the price of every article will always exercise equal influence upon
the results. For when relative prices are computed upon a fixed base
and averaged by the use of arithmetic means, those commodities that
have a long period upward trend in price will presently for outweigh
in influence those commodities whose prices are declining.
Lack of attention to weighting, then, does not automatically secure
a fair field and no favor to every commodity; on the contrary, it
58 For details, see pp. 161-168.
» See p. 31.
60
TH E M A K IN G AND USING OF INDEX NUM BERS.
results in what Walsh happily termed haphazard weighting.60 Per
haps “ unconscious weighting” would be an even better expression.
The real problem for the maker of index numbers is whether he shall
have weighting to chance or seek to rationalize it.
There are two excuses for neglect of weighting. First, as has been
shown in another connection, to collect satisfactory statistics showing
the relative importance of different commodities is extremely labori
ous and extremely difficult.61 Second, there are high authorities who
hold that the results turn out much the same whether or not formal
weights are used.62 Certainly 11the weights are of * * * less impor
tance in determining an index number of prices than the prices
themselves.” 63 But whether their importance is negligible is a ques
tion best answered by a study of actual cases such as are shown in the
next table.64
The discrepancies here revealed between the averages with hap
hazard and with systematic weights seldom amount to 10 per cent of
the results, except under the chaotic price conditions created by the
greenback standard in 1862-1873. In many kinds of statistics a 10
per cent margin of error is not accounted large. But in making wholesale-price index numbers for current years we may reasonably try to
get not two, but three, significant figures; and the third figure is
60 C. M. Walsh, The Measurement of General Exchange-Value, pp. 81 and 82. Haphazard weighting is
not necessarily the worst weighting; indeed, it may be better than the weighting which results from some
systematic calculations. For example, Bradstreet’s plan of using actual prices per pound is certainly
systematic, but the weighting which this system involves is probably less defensible than the haphazard
weighting involved in most averages of the relative prices of commodities selected at random. See p. 78.
61 See p. 26. When the (then) Department of Labor started its former index number it canvassed the
subject of weighting, but decided to use a simple average, because of the “ impossibility of securing even
approximately accurate figures for annual consumption in the United States of the commodities included.”
(Bulletin No. 39, of the Department Of Labor, p. 234, March, 1902.) It did, however, allot two or more series
to certain commodities, and thus introduced a rough system of weights. Unfortunately the number of
series allotted to each commodity seems to have been determined quite as much by the" ease of securing
quotations as by the importance of the articles. For criticism of the weighting whicn resulted, see pp. 48
and 49.
62 Compare A. L. Bowley, Elements of Statistics, 2d ed., pp. 113 and 220-224.
63 Irving Fisher, The Purchasing Power of Money, revised edition, p. 406. For further details see the
papers by Edgeworth, to which Fisher refers in his footnote.
64 Details concerning the first three sets of simple and weighted averages can be found in the documents
referred to in the table. But the fourth set of comparisons is based upon hitherto unpublished data and
requires description.
The “ unweighted” index numbers in this set are arithmetic means of the relative prices given in the
bulletins of the Bureau of Labor Statistics for the commodities listed below. But where two or more series
of relative prices are shown in the bulletins for different grades of the same articles, as in the case of cattle,
hogs, bacon, butter, corn meal, pig iron, etc., they were replaced by a single average series for the article in
question before the arithmetic means of the group were computed.
The “ weighted” index numbers were made from these same relative prices in the following way: (1)
For each commodity included the Bureau of Labor Statistics made a careful estimate, based upon a critical
study of the best available sources of information, of the physical quantity of it entering into exchange in
the year 1909. B y ‘ 1
‘ quantity entering into exchange” is meant the quantity bought and sold, irrespective
of the number of times it changed hands. (See pp. 63 and 64.) (2) These physical quantities were multi
plied by the average prices in 1909 of the respective commodities. (3) The resulting sums of money
were used as weights to multiply the relative prices of the respective commodities on the 1890-1899 base.
(4) The sums of the products were cast up for each year, and finally these sums were divided by the sums
of the weights, i.e ., the value in exchange for 1909.
The average prices of the commodities in 1909 may be found in any of the recent wholesale-price bulletins,
e. g., No. 149. The commodities included, and the estimated physical quantity of each entering into
exchange in 1909, are as follows:
Farm 'products: Barley, 75,300,538 bu.; cattle, 124,346.349 cwt.; corn, 460,778,251 bu.; cotton, 5,409,760,011
lbs.; flaxseed, 20,106,433 bu.; hay, 10,685,804 tons; hides, 922,243,894 lbs.; hogs, 76,438,923 cwt.; hops,
48,076,921 lbs.; oats, 267,859,660 bu.; rye, 29,520,508 bu.; sheep, 11,498,090 cwt.; wheat, 683,416,528 bu.
Food, etc.: Beans, 8,468,385 cwt.; butter, 1,042,709,708 lbs.; cheese, 353,641,892 lbs.; coffee, 1,038,439,285
lbs.; eggs, 926,690,112 doz.; codfish, 684,692 cwt.; herring, 428,804 bbls.; mackerel, 190,565,bbls.; salmon,
18,431,003 doz. cans; buckwheat flour, 2,009,599 cwt.; rye flour, 1,594,346 bbls.; wheat flour, 107,306,408 bbls.;
currants, 32,163,998 lbs.; prunes, 138,795,607 lbs.; raisins, 12,438,044 boxes; glucose, 7,701,223 cwt.; lard
1,243,572,129 lbs.; corn meal, 53,353,466 cwt.; bacon, 741,354,500 lbs.; beef, fresh, 4,209,196,748 lbs.; beef,
salt, 632,388 bbls.; hams, 789,861,744 lbs.; mutton, 495,458,067 lbs.;- pork, salt, 4,760,690 bbls.; milk,
7,749,070,256 qts.; molasses, 55,689,983 gals.; rice, 1,042,538,693 lbs.; salt, 22,136,489 bbls.; soda, bicarbonate,
165,600,000 lbs.; pepper, 36,241,462 lbs.; sugar, raw, 6,316,033,669 lbs.; sugar, granulated, 7,366,818,210 lbs.;
tallow, 203,209,103 lbs.; vinegar, 98,403,927 gals.; potatoes, 397,491,062 bu.; onions, 4,972,947 cwt.; tea,
113,547,647 lbs.
Metals and implements: Bar iron, 2,166,529,067 lbs.; barbed wire, 6,471,300 cwt.; copper, ingot, 1,312,437,919
lbs.; copper wire, 278,964,000 lbs.; lead, pig, 732,152,538 lbs.; lead pipe, 1,058,280 cwt.; nails, wire, 13,916,097
kegs; pig iron, 9,896,248 tons; tin (pig), 94,248,471 lbs.; silver, 151,969,144 ozs.; spelter, 464,903,059 lbs.; Steel
billets, 4,972,179 tons; steel rails, 3,025,009 tons; tin plate, 12,968,174 cwt.
METHODS USED IN M A K IN G INDEX NUM BERS.
61
usually altered in appreciable degree by the substitution of systematic
for haphazard weights. Even the large Canadian series, with its 272
commodities, is shifted 9.5 points, or more than 7 per cent, in 1912 by
weighting.
T able
1 4 .—COMPARISONS OF W E IG H T E D AND U N W E IG H T E D IN D E X NUM BERS.
[1. From the report of the Senate Committee on Finance, Mar. 3, 1893.
(Arithmetic means.
All articles
averaged
according
Simple
to
impor
arithmetic
cer
means, all tance,
tain ex
articles.
penditures
being
uniform.
Year.
1860................................................
1861................................................
1862................................................
1863................................................
1864................................................
1865................................................
1866................................................
1867................................................
1868........... : ..................................
1869................................................
1 870.............................................
1 8 7 1 ............................................
1872................................................
1873................................................
1874................................................
1875................................................
1876................................................
1877................................................
1878................................................
1879...............................................
1880................................................
1881................................................
1882................................................
1883................................................
1884................................................
1885................................................
1886................................................
1887................................................
1888................................................
1889................................................
1890................................................
1891................................................
100.0
100.6
117.8
148.6
190. 5
216.8
191.0
172.2
160.5
153.5
142.3
136.0
138.8
137.5
133.0
127.6
118.2
110.9
101.3
96.6
106.9
105. 7
108.5
106.0
99.4
93.0
91.9
92.6
94.2
94.2
92.3
92.2
By years, 1860 to 1891.]
Prices in 1860= 100.)
100.0
95.9
102.8
122.1
149.4
190.7
160.2
145.2
150.7
135.9
130.4
124.8
122.2
119.9
120.5
119.8
115.5
109.4
103.1
96.6
103.4
105.8
106.3
104.5
101.8
95.4
95.5
96.2
97.4
99.0
95.7
96.2
All articles
averaged Difference Difference Difference
between
according
between
between
to impor simple and simple and first and
tance: 68.6
first
second
second
weighted
per cent of weighted
weighted
averages.
averages.
total ex
averages.
penditure.
,
100.0
94.1
104.1
132.2
172.1
232.2
187.7
165.8
173.9
152. 3
144.4
136.1
132.4
129.0
129.9
128.9
122.6
113.6
104.6
95.0
104.9
108.4
109.1
106.6
102.6
93.3
93.4
94.5
96.2
98.5
93.7
94.4
4.7
15.0
26.5
41.1
26.1
30.8
27.0
9.8
17.6
11.9
11.2
16.6
17.6
12.5
7.8
2.7
1.5
1.8
3.5
.1
2.2
1.5
2.4
2.4
3.6
3.6
3.2
4.8
3.4
4.0
6.5
13.7
16.4
18.4
15.4
3.3
6.4
13.4
1.2
2.1
.1
6.4
8.5
3.1
1.3
4.4
2.7
3.3
1.6
2.0
2.7
.6
.6
3.2
.3
1.5
1.9
2.0
4.3
1.4
2.2
1.8
1.3
10.1
22.7
41.5
27.5
20.6
23.2
16.4
14.0
11.3
10.2
9.1
9.4
9.1
7.1
4.2
1.5
1.6
1.5
2.6
2.8
2.1
.8
2.1
2.1
1.7
1.2
.5
2.0
1.8
[2. From Bulletin of the Department of Labor, No. 27, March, 1900. January of the years, 1890 to 1899.]
(Arithmetic means.
Averages of 9 quarterly quotations, January, 1890, to January, 1892=100.)
Year and month.
1890, Januarv..............................
1891, January..............................
1892, Januarv..............................
1893, January..............................
1894, January..............................
1895, January..............................
1896, January..............................
1897, January..............................
1898, January..............................
1899, January..............................
All articles All articles
averaged
averaged
according according Difference Difference Difference
to
impor
to
impor
between
between
between
All articles tance, cer tance,
com
simple
simple
first and
simply
tain ex
prising
and
first
and
second
second
averaged. penditures
68.6 per
weighted
weighted
weighted
being con
cent of
averages.
averages.
averages.
total ex
sidered
1
uniform. penditure.
102.0
100.6
96.5
97.2
89.6
84.7
85.2
82.0
83.3
86.5
100.1
102.2
100.0
103.4
97.5
93.5
92.8
90.3
91.0
91.0
100.2
103.2
100.1
105.0
96.4
90.5
89.5
85.9
86.8
86.8
1.9
1.6
3.5
6.2
7.9
8.8
7.6
8.3
7.7
4.5
1.8
2.6
3.6
7.8
6.8
5.8
4.3
3.9
3.5
.3
0.1
1.0
.1
1.6
1.1
3.0
3.3
4.4
4.2
4.2
THE M A K m G AND USI^TO OF INDEX LUMBERS.
T jjsle .W ,—COMPARISONS OF W E IG H T E D AN D U N W E IG H T E D IN D E X NU M BER S—Con.
{3. From Wholesale Prices, Canada, 1915.
Report by R. H . Coats,
B y years, 1890 to 1913.J
(Arithmetic means. Average prices in 1890-1899=100.)
Year.
Un
Weighted weighted
index
index
number. number.
112.0
111.3
104.9
103.9
97.2
95.6
90.6
89.9
95.5
99.0
105. 8
106.0
1890.,
895.
896..
897.,
900.
901.
Differ
ences.
110.3
108.5
102.8
102.3
97.2
95.6
92.5
92.2
96.1
1.7
2.8
2.1
1.4
1 .9
2.3
.6
1.1
1.0
100.1
108.2
107.0
2.4
Year.
Un
Weighted
weighted
index
index
number. number.
1902.
1903.
109.6
109.7
1904.
110.6
1905.
1996.
1907.
1908.
1909.
191®.
1911.
19121913.
113.8
120.1
129.2
125.1
126.3
128.0
131.1
143.9
139.6
Differ
ences.
109.0
110.5
111.4
113.8
120.0
126.2
120.8
121.2
124.2
127.4
134.4
135. 5
0.6
.1
3.0
4.3
5.1
3.8
3.7
9.5
4.1
[4. From computations by the Bureau of Labor Statistics.1]
(Arithmetic means.
Average prices in 1830-1899=100.)
37 food products.
13 farm products.
Year.
14 metallic products.
Weighted
Weighted
Weighted
by esti
by esti
byesti- ,
mated ex Dif
mated ex-, Dif
mated ex
Dif
Un
pendiUn
pendi
Un
pendi
fer- ’
fer
fer
tures
weighted.
tures
weighted.
tures
ences.- weighted.
upon each
upon each ences.
upon each, ences.
article in ;
article in
article in ;
1909.
s'
1909.
1909.
1890.......................
1891.......................
1892.......................
1893.......................
1894.......................
1835.......................
1896.......................
1897.......................
1898.......................
1899.......................
1900.......................
1901.......................
1902.......................
1903.......................
1904.......................
1905.......................
1906.......................
1907.......................
1 9 0 8 ....................
1909....................
1 9 1 0 ................ .
1911...................
1912.......................
1913.......................
113
124
112
1Q6
96
93
78
84
97
99
109
117
130
120
m
125
122
139
135
150
161
166
173
152
109
117
105
107
94
95
86
93
97
98 •
109
115
129
120
128 ,
123
124
136
135
154
165
150
164
161
4
7
7
1
2
2
8
9
0
1
0
2
1
0
2
2
2
3
0
4
4‘
16
9
9
114
116
105
112
99
95
83
87
93
98
198
110
114
110
113
110
115
120
122
124
129 ;
128
137
133
114
114
103 ;
111 1
97
94
86
90
96
95
100
102
108
104
110
109
106
112
119
126
127
125
137
127
0
2
2•
1
2
1
3
3
3
2
8•
8
6
6
3.
1
9.
8
3
2
2
3
0
6
128
118
110
102
88
88
93
82 •
83
124
124
114
114
114
105
116
131 :
138
103
109
111
111
120
119
131 .
11-6 ■
107
98 •
84 .
88 r
91
80
81
124
123
113
114
113
102113
130140
198
107
108
103
114
115
3
2
3
4
4
0
2
2
2
0
1
1
0
1
3
3
1
2
5
2
3
8
6
4
1 See explanations in footnote, p. CO.
If rational weighting is worth striving after, then by what method
shall the weights of the different commodities be arrived at? That
depends upon the object of the investigation. If, for example, the
aim be to measure changes in the cost of living, and the data be
retail quotations of consumers7 commodities, then the proportionate
expenditures upon the different articles as represented by collections
of family budgets make appropriate weights. If the aim be to study
changes in the money incomes of farmers, then the data should be
METHODS USED IN MAKING INDEX NUMBERS.
63
“ farm prices,” the list of commodities should be limited to farm
products,' and the weights s1 111
' *onate to the total money
receipts from the several
aim be to construct a
it i
i
j
11 ii
“ business barometer,” the
4 ices from the most repre
sentative wholesale markets, the list should be confined to com
modities whose prices are most sensitive to changes in business pros
pects and least liable to change from other causes, and the weights
may logically be adjusted to the relative faithfulness with which the
quotations included reflect business conditions. If the aim be merely
to find the differences of price fluctuation characteristic of dissimilar
groups of commodities, or to study the influence of gold production
or the issue of irredeemable paper money upon the way in which
prices change, it may be appropriate to strike a simple arithmetic
average of relative prices. If, on the other hand, the ann be to make
a general-purpose index number of wholesale prices, the question is
less easy to answer.
One proposition, however, is clear. The practice of weighting
wholesale-price index numbers by figures drawn from family budgets
is to be deprecated; for family budgets do not show the importance
of wheat and cotton, of petroleum and spelter, of tar and lime, of pig
iron and hides, of brick and lumber; indeed, to apply budget weights
to half or more of the articles in any wholesale list is nonsensical.
And to pretend that wholesale-price index numbers when weighted
on the basis of family expenditures show fluctuations in the cost of
living is to overtax the credulity of those who know and to abuse the
confidence of those who do not.
Allied to the family-budget method of weighting and yet vastly
better for wholesale-price index numbers is the “ aggregate expend
iture” method.05 Here an attempt is made to ascertain the aggre
gate sums of money laid out by the people of a whole country upon
the articles quoted and to adjust the weights upon this basis. Of
course the country as a whole buys raw materials, as single families
do not, and of course consumers’ commodities can be taken at their
aggregate values in wholesale markets. Similar in net effect is the
weighting on the basis of consumption practiced by the British
Board of Trade. For “ consumption is taken to mean any process
by which the commodity is substantially changed in character. In
other words, consumption in manufacture is recognized as well as
consumption by an individual.” 66 Somewhat different weights would
result i f quantities or values produced were taken in place of quanti
ties or values consumed. Mr. Walsh thinks it best to combine these
two criteria— that is, to take “ either the total product or the total
consumption according as the one or the other is the greater.” 67
Prof. Irving Fisher prefers “ an index number in which every article
or service is weighted according to the value of it exchanged at base
prices in the year whose level of prices it is desired to find.” 68 On
this system the weight assigned to each article would be affected by
the number of times it changed hands on its way from producer to
65 See G. H. Knibbs, Prices, Price Indexes, and Cost of Living in Australia. Commonwealth Bureau of
Census and Statistics, Labour and Industrial Branch, Report No. 1, pp. 11-14.
66 Report on Wholesale and Retail Prices in the United Kingdom in 1902. London, 1903, p. 441. The
accuracy of the statistics upon which the Australian and British index numbers are based may be open
to question. Not the data, but the method is of interest here.
67 C. M. Walsh, The Measurement of General Exchange-Value. New York, 1901, p. 95.
68 Irving Fisher, The Purchasing Power of Money, revised edition. New York, 1913, pp. 217 and 218.
64
THE M A K IN G AND USING OF INDEX NUMBERS.
final consumer. A variation of his plan is therefore represented by
the proposal to weight each article according to the quantity of it
which enters into the country's commerce, irrespective of the fre
quency with which it changes hands.
The practical consequences of adopting these different systems of
weighting may be illustrated by considering their application to
cotton, corn, and coffee in the United States. Production weights
would give cotton much greater importance than consumption or
aggregate-expenditure weights, because so large a part of the Ameri
can crop is exported and consumed abroad. Exchange weights would
be practically equivalent to production weights, because practically all
the cotton grown is sold b y the planters and enters into the commerce
of the country, and relatively little cotton is imported. On Prof.
Fisher's plan, however, the exchange weights would be some multiple
of the production weights, depending upon the average number of
American hands through which the cotton passed. In the case of
corn, production and consumption weights would substantially agree,
for w^e import little corn ana export but a small percentage of the
>roduction. On the other hand, exchange weights would be much
ess than either production or consumption weights, because a large
part of the corn crop is never sold, but is consumed on the farms
where it is grown. In the case of coffee, production weights would
be zero, while consumption and exchange weights.would correspond
closely.
We are helped toward a choice among these rivals by common
agreement upon a slightly different point. In arranging any system
of weights except Prof. Fisher's, double counting is to be avoided so
far as possible. For example, if cotton is counted at its full impor
tance as a raw material, then cotton yarns and later cotton fabrics
made of the yarn can not be counted at their full importance with
out assigning triple weight to the raw cotton which is represented at
these two successive stages of manufacture. Now, if this sensible
observation be applied to cases like those of corn, hay, etc., it casts
the die in favor of exchange weights. For if these articles, which
are used largely by the original producers in making things quite
different from corn and hay (for instance, pork and beef) are counted
at the full amount produced or consumed, and if their products (the
pork and beef) are also counted at the full amount produced or con
sumed, there will be a great deal of double counting. Not all but
much of this duplication can be eliminated by counting only the
amount of corn and hay sold by the producers and letting the rest of
these articles produced and consumed get their proper representation
under the caption of pork, beef, etc.69
If for this reason exchange appears a rather better criterion of
importance than production, consumption, or a combination of the
two, it remains only to decide whether the number of times a thing
is exchanged should be recognized. Prof. Irving Fisher had good
cause to propose multiple counting, for he wanted an index number
of prices for constructing the “ equation of exchange," a mathematical
expression of the necessary equivalence between the total volume of
(
69 Of course, this same end might be attained without surrendering the production or consumption
basis if the rule against double counting of raw materials and products were made broad enough to include
corn, for example, as the raw material of pork; but needless to say there is little liklihood that the common
meaning of terms will be stretched to such an extent.
METHODS XJSED IK M A K IN G INDEX NUM BERS,
65
business done in a country and the total volume of payments effected
by me.ans of money and credit instruments. Of course the oftener
an article is sold and paid for the more important it is as a factor in
this equation. But it does not follow that the economic importance
of an article is greatly changed by reorganizing the chain of business
enterprises that deal in it. “ Integration of industry,” as expressed
in our trusts, does not make pig iron less significant as an item in the
country’s economic life, except #in the sense that it reduces the
average number of transfers of ownership. T^Jie quantity of the
article that enters into exchange, then, irrespective of the number
of turnovers, is probably the most satisfactory gauge of importance
to apply in making general-purpose index numbers. Anyone experi
enced in the search for statistical information will need no warning
that in the working out of weights along this line many puzzling
cases will arise in which consistency will be difficult to maintain, to
say nothing of the wide gaps and weak places that will be revealed
among the available data. That this system of weighting* is feasible
in practice as well as desirable in theory, however, was proved by the
Bureau of Labor Statistics in 1914, when it gave up averaging
relative prices and began multiplying actual prices by the quantities
of commodities that entered into trade in the base year 1909.70
Three interesting questions remain: Should the weights be sums
of money or physical quantities ? Should the weights be changed
from year to year or kept constant ? Should the weights be adjusted
to the importance of the commodities as such, or should there be
taken into account also the importance of the commodities as repre
senting certain types of price fluctuations ?
When relative prices are being used the weights should be reduced
to a common denominator. As multipliers, of course, weights may
be regarded as merely abstract numbers; but in studying the weights
themselves it is necessary to have some common standard by which
the relative importance assigned to various commodities can be
accurately compared. The only common denominator for all com
modities that is significant for economic ends and capable of quanti
tative expression is money value. But it is ill advised to weight by
money values when actual prices are being used, for the common
denominator is already present in the quotations themselves. These
price quotations are best multiplied by the physical quantities, of the
goods produced, exchanged, or consumed, as the case may be.
Like most of the issues on which authorities differ, the question
whether it is desirable to change weights at frequent intervals depends
upon the precise end in view. Most makers of index numbers have
wished to isolate the price factor from other changes in the economic
complex. Hence they have preferred to keep their weights as nearly
constant as possible. For when the weights are altered the index
number becomes a measure of two sets of changes, and no one can
tell what part of the net results is due to variations in prices and what
to variations in weights.71 Yet it is clear that a system of fixed
weights applied over a long period is certain to become inaccurate
for most o f the years, however carefully it is adjusted to conditions
prevailing at some base period. Practically, then, a compiler who
wishes to ascertain how prices have changed must choose between
70 For details see Bulletin No. 181 of the Bureau of Labor Statistics.
71 See the criticism of index numbers made from import-export values, pp. 29-31.
33226°— 21— Bull. 284------ 5
66
M A K IN G AND USING OF INDEX NUMBERS.
two evils— inaccurate weights and ambiguous price measures. Some
times he cam minimize the first evil b y collecting data showing the
average importance of his commodities over a period of yearsr for
these averages are less likely to go awry than figures for a single
year. In other cases the least objectionable compromise is probably
to revise the scheme of weights, say, once a decade, and to show the
effect of this change by competing overlapping results for a few years
with both the old and new weight^.72 A further practical reason in
favor of this compromise is found in the heavy expense in time and
labor required for frequent revisions of the weights.
Writers like Mr. Walsh, Prof. Pigou, and Prof. Fisher, who urge
the adoption of a formula in which the weights are changed every
year, put another aim in the foreground. Their primary purpose is
to secure the utmost possible nicety in measuring the rise or fall of
prices in each pair of years treated. Of course an index number
made with these changing weights 11measures neither the varying
cost of a Constant amount of goods nor the varying amount of goods
which a dollar will buy.” 73 But, since the importance of price fluc
tuations depends largely upon the accompanying changes in the
quantities of goods bought, there is use for index numbers that do
not attempt to measure the price factor in isolation. By changing
weights each year it is possible to make these constantly occurring
changes in quantities bought influence the price index, and therefore
to secure results better fitted for certain uses than the results of an
unambiguous measure of fluctuations in prices.74
To the third question, whether weights should be adjusted to the
importance of the commodities as such, or whether there should also
be taken into account the importance of these commodities as repre
sentatives of certain types 01 price fluctuations, little attention has
been paid. But the preceding section shows that this neglected
problem is both important and difficult. The prices of raw materials
behave differently from the prices of manufactured goods; among
the raw materials the prices of farm crops, of forest, animal, and
mineral products behave differently; there are also differences of
behavior between the prices of manufactured goods bought by pro
ducers and by consumers, etc. Is an accurate measure of changes
in the level of all wholesale prices obtained unless all of the different
types of fluctuation, doubtless including types not vet definitely rec
ognized, are represented in accordance with the relative importance
o f the commodities exhibiting each type ?
How can such representation be attained ? If all the commodities
bought and sold could be included on a strictly uniform basis in the
index number, it would suffice to weight each by the criterion of its
own individual importance. Since that is out of the question, it is
desirable to draw from each part of the whole system of prices
samples sufficient to determine its characteristic fluctuations, and then
to make sure that each part of the whole system counts for the proper
amount in determining the final result. On this plan commodities
would be weighted simply as commodities in making the subtotals
72 Compare G. H . Knibbs, Prices, Price Indexes, and Cost of Living in Australia. Commonwealth
Bureau of Census and Statistics, Labour and Industrial Branch, Report No. 1, pp. xxiv and xlix.
73 prof. Warren M. Persons: wFisher’s formula for index numbers,” Review of Economic Statistics,
May, 1921, p. 115, note.
74 See the discussion of the ccideals formula in section 9, p. 91.
METHODS
V&El>
IN M A K IN G INDEX NUM BERS.
67
for each recognized -^group, and these subtotals would be weighted
again in making up the grand totals.
Such a plan was adopted by the Price Section of the War Industries
Board in making their index number of prices in 1913-1918. As
noted above, the subdivisions used by the Price Section were 50
classes of commodities based, so far as possible, on the organization
of industries. Within each class, raw materials were weighted
according to quantities used by the industry represented, and
products were weighted according to the quantities produced. A
separate index number was made in this way for each of the 50
classes. These indexes and the materials from which they were
made, both price quotations and weights, seemed fairly satisfactory
as such matters go; but before the aggregates of the commodity
prices times commodity weights for these 50 classes were assembled
to make aggregates for “ all commodities,” it was clear that there
would be wide differences in the fullness with which prices in the
various industries had been covered. In some industries 75 to 90
per cent of the value of the transactions was represented by the
prices multiplied by the weights; in other industries the percentage
sank below 25. Again, there were industries in which it had been
possible to quote commodities at three stages—raw materials, partly
manufactured goods, and finished products— while in other industries
the available data represented only raw materials or only finished *
products. That is, while .the weights within each class had been
systematized, and while the plan of systematizing the w&ights was
uniform in all classes systematized, the weights as between different
classes were haphazard to a degree. To overcome the difficulty, the
Price Section prepared a second set of weights. It estimated the
value of the products sold b y each industry represented, divided these
estimates by the aggregate of commodity prices times commodity
weights, and so obtained a set of factors which when applied to the
class aggregates give each class an influence upon the index for “ all
commodities>} proportioned to its estimated importance.75
Professor Edgeworth has pointed out a yet further desideratum in
weighting. Most index numbers are made from samples of the data
which logically fall within the field investigated; and the task is to
make from these samples the best approximation to a measure of the
unknown whole. Now “ the theory of errors-of-observation shows
that in the combination of the given observations, 4less weight should
be attached to observations belonging to a class which are subject to
a wider deviation from the mean. Such would be prices of articles
which, exclusive of the common price movement of all the selected
articles, are liable to peculiarly large 'proper fluctuations.' ” 76
Perhaps it is a counsel of perfection to urge such refinements in
systems of weighting. Certainly the difficulties to be encountered
are very great. Statistical knowledge is not complete enough to sup
ply accurate data for weighting all the different parts of the system of
prices that are known to have characteristic peculiarities of fluctua
tion. Nor have these different types and the commodities exhibiting
See History of Prices During the War, Summary, War Industries Board Price Bulletin, No. 1.
™ Economic Journal, June, 1918 (Vol. X X V I I I . p. 188). The quotation within the quotation is from*
the British Association Memorandum^ 1887 (p. 36). To make his point clearer, Prof. Edgeworth adds in a
footnote this remark from the corresponding memorandum of 1889 (p. 157): “ If more weight attaches to a
change of price in one article rather than another, it is not on account of the importance of that article to the
consumer or the shopkeeper, but on account ©f the importance to the calculators of probabilities as affording
an observation which is peculiarly likely to be correct.”
68
TH E M A K IN G AND USING OF INDEX NUMBERS.
each been adequately studied. And puzzling difficulties are raised
by overlapping among the types— there are commodities that belong
in two places at once. But here is certainly a promising lead for
future efforts to improve present measurements of changes in the
price level. Even now it might be feasible by taking pains to secure
rough justice as between raw and manufactured commodities, and as
between raw vegetable, animal, forest, and mineral products. One
modest step in the right direction can readily be taken by any com
piler of index numbers: He can make clear that his results do not
measure changes in the general level of wholesale prices accurately
when they are obtained without an effort to represent each part of the
field according to its due importance.
7. AVERAGES AND AGGREGATES.
Among all the problems involved in the making of index numbers
the one that has been most discussed is the best form of average
to strike. Most of these discussions have come from men inter
ested in the mathematical side of statistics rather than in the
problem of ascertaining what changes have actually occurred in
rices. The practical makers of index numbers, on the contrary,
ave seldom troubled themselves greatly about theoretical con
siderations. Indeed, the two problems of finding out how prices
have actually changed, and finding the best method of measuring
changes, appeal to two types of interest, which are seldom
strongly developed in the same mind. The mathematical statistician
is likely to know little and care less about the field work of collecting
price quotations. To the practical statistician this field work is of
overshadowing importance, and the subsequent manipulation of his
data is a matter of secondary interest. Hence, a study of index
numbers as they are made need not carry one into long mathematical
flights.77
First, it should be recalled that certain compilers of index numbers
do not strike averages at all. The old form of the Economist index
and Gibson’s present index, for example, are sums of relative prices.
More important are the series which dispense with the use of relative
prices for each commodity, and give results in the form of sums of
actual prices, or such sums thrown back into a series of relative num
bers. These cases are still exceptional, however, and most index
numbers are made by finding some sort of average from the relative
prices of the commodities included.
The sort of average struck is usually the arithmetic mean— that
is, the sum of the relative prices divided by their number. Occasion
ally medians are used— that is, the midmost relative prices which
divide the whole number of cases into two equal groups, half above
the median and half below. In one famous investigation,78 geometric
means were employed— that is, all the relative prices for a given date
were multiplied together and the nth roots of the products were
extracted, n standing for the number of commodities included. But
Jevons has had few imitators, though Mr. A. W. Flux has just adopted
E
• ” The best systematic discussions of averaging for the purpose in hand are to be found in Prof. Edge
worth's papers referred to in the footnote on p. 8; Irving Fisher’s The Purchasing Power of Money,
revised edition, 1913, pp. 385-429; and C. M. Walsh’s The Measurement of General Exchange-Value, 1901,
and his new treatise, The Problem of Estimation, 1921.
78 W . S. Jevons, “ A serious fall in the value of gold ascertained/' 1863. Reprinted in his Investigation
in Currency and Finance, 1884, pp. 13-118.
METHODS USED IN M A K IN G INDEX NUM BERS.
69
the geometric mean for the new form of the British Board of Trade
index number. The other standard forms of averages— the mode
and the harmonic mean— have been discussed frequently, but so
far as is known they have never been consciously used in making
index numbers.79
For the geometric mean two great merits are claimed. First, unlike
the arithmetic mean, it is not in danger of distortion from the asym
metrical distribution of price fluctuations. Chart 2 shows that in a
large collection of percentage variations from the prices of the pre
ceding year, the extreme cases of rise run about twice as far up the
scale as the extreme cases of fall run down. Such a distribution is
characteristic of relative prices in general. Indeed, the case cited is
distinctly moderate; most collections *of variations covering many
years would show a greater difference. There is indeed no limit to
the possible percentage of rise in prices, while the possible percentage
of fall can not exceed 100.80 The cases of extraordinary advance,
accordingly, tend to raise the arithmetic mean more than the cases
of extraordinary decline tend to depress it. If, for example, one
commodity rose tenfold in price and another commodity fell to onetenth of the old price, the arithmetic mean would show an average
rise of 505 per cent (1,000 + 10-^-2), while the geometric mean would
show no change in the average, since Vl>000x 10 = 100.
This favorite imaginary case of *10 and 1,000 seems extreme, but
contrasts approximately as violent as that actually occurred in the
recent war. The Price Section of the War Industries Board has
computed the relative prices of 1,437 commodities in 1918 on the
basis of their average prices in the twelve months, July, 1913, to
June, 1914. These figures are reproduced in somewhat condensed
form in Table 15. Here the array of relative prices is far more elon
gated in one direction than in the other, and the highest relative
price is upwards of 100 times as great as the lowest relative price.81
Accordingly, the arithmetic mean (217) stands high above the geo
metric mean (194) and median (191).82
79 Concerning the properties of these averages see, for example, F. Zizek, Statistical Averages (translated
by W . M. Persons), and O. U. Yule, Introduction to the Theory of Statistics, pp. 120-123, 128-129. The
“ crude mode” is that relative price which occurs most frequently in the data under examination, e. g., in
Chart 2 it is “ no change.” The true mode is “ the value of the variable corresponding to the maximum of
the ideal frequency-curve which gives the closest possible fit to the actual distribution.” “ The harmonic
mean of a series of quantities is the reciprocal ofthe arithmetic mean of their reciprocals.”
80 Negative prices are conceivable of course; but do they ever occur in the sources which the maker of
index numbers uses? Suppose that some kind of factory waste, which usually commands a low price,
should fail ofits market, and accumulate so as to become a nuisance. The factory jnanager might logically
set it down at a negative price; but he is much more likely to offer a positive price for another commodity—
the removal ofthe waste.
81 This ratio of 100 to 1 was indeed surpassed in some months. The highest relative price found was
5,081 (acetiphenetidin, November, 1916).— See History of Prices during the War (War Industries Board
Pricp Bulletin, No. 54, p. 18).
82 From the skewed distribution characteristic of relative prices when arranged on the ordinary arith
metic scale, Prof. Frederick R. Macaulay has developed an ingenious argument in favor of the geometric
mean.
#
He puts the matter in this way: “ What is the most probable value for the general percentage movement?
If the1errors* (variation due to the influence of particular commodity factors) were distributed arithmetically
according to the normal law, the arithmetic mean—least mean square deviation—would certainly seem
indicated. But if the logarithms of the percentages and not the percentages themselves follow more closely
the curve of error, is not the geometric mean indicated? From that point the curve of the squares of the
logarithms of the percentage deviations will be a minimum; and is not this what sound theory should de
mand?” American Economic Review, March, 1916, Vol. V I, p. 207.
The answer to Prof. Macaulay’s final question is that what sound theory demands depends upon the
precise magnitude one desires to measure. Itis argued hereafter in the text that if the purpose be to measure
the average ratio of change in prices, the geometric mean is in strictnessthe only proper average to employ.
Those who can utilize measures of average change for their purposes will be gratified to know that the arrays
from which their averages are made usually conform better to the normal law of distribution than the arrays
from which arithmetic means of relative prices are derived. As Prof. Edgeworth humorously admits, “ it
is a merit in a statistical group to conform to the normal law.” (Economic Journal, June, 1918, Vol. X X V I II,
p. 182). But, of course, the symmetry of the distribution of data from which different averages are derived
is but one, and generally a subordinate, consideration in the choice of averages.
10
THE MAKJNG AND USING OF INDEX NUM BEBS.
T able
1 5 .—DISTRIBUTION OF THE R EL ATIV E PRICES OF 1,437 COMMODITIES IN 1918.
(Average prices in July, 1M.3, to June, 1914=* 100.)
Relative prices.
36............................
49............................
50-69......................
70-89......................
90-109....................
11CM29..................
130-149..................
150-169..................
170-189
...........
190-209.................
210-229..................
230-249..................
Num
ber of
cases.
1
1
4
17
61
64
130
212
219
164
135
104
Relative prices.
250-269...................
270-289...................
290-309...................
310-329...................
330-349...................
350-369...................
370-389...................
390-409...................
410-429..................
430-449...................
450-469..................
470-489...................
Num
ber of
cases.
76
54
42
30
31
16
13
7
7
8
4
4
Relative prices.
490-509.................
510-529.................
530-549.................
550-569............. .
587........................
627........................
727........................
739........... -...........
743........................
761........................
784........................
826........................
Num
ber of
cases.
4
5
3
4
1
1
1
1
1
1
1
1
Relative prices.
Num
ber of
cases.
848........................
900........................
1,165.................
1,355.................
1, 585...................
1,764....................
2,049....................
2,863....................
3,009.............
1
1
1
1
1
1
1
1
1
The second merit claimed for geometric means is that they can be
shifted from one base period to another without producing results
that seem to be inconsistent. Suppose, for example, that the price
of wheat falls from II per bushel in 1913 to 50 cents in 1914, while
the price of corn remains unchanged at 40 cents. Then the relative
prices are—
(1) On the basis, prices in 1913 = 100:
1913
W heat...........................................................
Corn...............................................................
100
100
1914
50
100
(2) On the basis, prices in 1914 = 100:
1913
Wheat...........................................................
Com .............................................................
200
100
1914
100
100
The arithmetic and geometric means of these relative prices are—
(1) On the basis of prices in 1913:
1913....................
1914....................
Arithmetic means.
Geometric means.
(100+100)-5-2= 100
(50+100)^-2= 75
VlOGX 100= 100.00
V 50X100= 70.71+
(2) On the basis of prices in 1914:
Arithmetic means.
1913....................
1914....................
(200+100)-r-2= 150
(100+100)-2= 100
Geometric means.
:
V200X100=141.42V 100X100* 100.00
METHODS USED IN M A K IN G INDEX NUM BERS.
71
Here the arithmetic means can not, but the geometric means can,
be shifted from the 1913 base to the 1914 base or vice versa by
simply dividing the index number for one year b y that for the other.
That is, 100^75 = 13,31, not 150; but 100-f;70.71 = 141.42.83 By
shifting the base in this simple fashion geometric means can be made
to give direct comparisons between the price levels at any two dates
covered by the investigation, whereas with arithmetic means com
parisons not made in terms of prices at the original base period give
results that may present formal inconsistencies and results whose
meaning is difficult to grasp and put into words.
A third advantage of geometric means is that they are likely to
be nearer the modes of the distributions which they represent than
are arithmetic means. The importance of this point will be more
generally appreciated as statisticians come to study the whole array
of the price fluctuations with which they deal, instead of concen
trating their attention merely upon averages.
The chief objection to geometric means in an index number
intended for general use is that this form of average is unfamiliar and
therefore more likely to be misinterpreted than arithmetic means.
Further, geometric means do not have any direct bearing upon
changes in the purchasing power of money as do arithmetic means
and weighted aggregates of actual prices.84 Finally, geometric means
are somewhat more laborious to compute than arithmetic means or
medians. Instead of adding the relative priees just as they stand and
dividing the sums by their number, the computer must convert the
relative prices of every commodity into their logarithms, add these
logarithms, divide the sum b y the number of logarithms, and look up
the natural numbers corresponding to the quotients.85 Statisticians
are the more loath to incur the extra labor of this process since the
special merits of the geometric mean are shared in part by certain
53 See the discuss ion of shifting bases, pp. 83-90.
84 This point is more fully explained on pp. 76 and 77.
83If relative prices are not needed for any other purpose, it is quicker to compute the geometrie mean from
the logarithms of the successive actual prices and then to find the ratios between the results. But even
that is a somewhat longer process than calculating relative prices, casting them up, and dividing by their
number.
That geometric means can be computed either with or without the use of relative prices is readily shown.
Let %>o,. V f l
stand for the actual prices of n commodities in the two years o and x.
Then the relative prices of these articles, in the year as on the-hasis of actual pciees in the year o are
,
Vx, p X,
Po ’ P'o
n
X
„ n
ft
The geometric mean of these relatives is
But this, expression is equal to
VO*) 0 0 • • • ( O
•Vw
••• «>
And the latter expression,, of course, is the ratio between the geometric means of the actual prices in the
two years.
72
THE M A K IN G AND USING OF INDEX NUMBERS.
other forms of index numbers. Like geometric means, aggregates
of actual prices, or relatives made directly from them, can be
shifted to any base desired without raising difficult problems of
interpretation. Like geometric means, again, medians are not
more affected b y cases of exceptionally great advances in price than
by cases of exceptionally great declines. Hence in practice most
makers of index numbers who distrust arithmetic means abandon the
practice of averaging relative prices or use medians instead of geo
metric means.
Medians, indeed, have several distinguished champions among
theoretical statisticians.86 It is generally claimed that of all averages
medians are the easiest to compute, for a quick ordering of the data,
followed by a counting of the items, takes the place of casting sums
and dividing by the proper number. But in this day of adding
machines the palm for ease of computation has shifted to the arith
metic mean and the aggregate of actual prices. More important
is the fact demonstrated by Prof. Edgeworth that the median is
safer than the arithmetic mean when, as m the case of index numbers,
the items to be averaged are samples drawn from a larger field.
For, according to the theory of probabilities, the probable error
of the median can not in any case be much greater than that of the
arithmetic mean and in other cases it may be very much less.87
But medians have their drawbacks. (1) They are not perfectly
reversible; that is, they can not be shifted from one base to another by
simple division without ambiguity. (2) Medians of different groups
can not be combined, averaged, or otherwise manipulated with ease as
can arithmetic means. For example, in making up its old form of
index number the Bureau of Labor Statistics could add the sums used
for making arithmetic means of the relative prices of farm products,
foods, cloths and clothing, etc., and from the sum of these sums strike
the grand average for all commodities. It could not handle medians
in this convenient fashion; instead of combining the sums from the
groups it would have to reexamine and rearrange the relative prices
of those commodities which fell near the respective medians. Simi
larly, a reader who finds arithmetic means of two groups in different
sources can compute the arithmetic mean of these means, provided
the number of items in each group be stated, with no greater error
than that arising from the dropping of fractions in the published
data; but he can not approximate except in the vaguest way the
median of two medians.88 (3) When the number of items to be
averaged is small, medians are erratic in their behavior. For in
such groups there is often a considerable interval between the mid
most relative price and the relative prices standing next above it
and next below. No change in any of the items, large or small, can
alter the position of the median unless it shifts an item from the
86Compare, for example, F. Y . Edgeworth, “ Index numbers,” Dictionary of Political Economy, Vol. II,
p. 386; Irving Fisher, The Purchasing Power of Money, revised edition, p. 425: A. L. Bowley, Elements of
Statistics, second edition, p. 224. Walsh, however, does not recognize the median as a mean. See Quar
terly Publication of the American Statistical Association, March, 1921, p. 542, and the numerous
references to medians in his Problem of Estimation.
w See his paper “ On the use of analytical geometry to represent certain kinds of statistics,” Journal of
the Royal Statistical Society, June, 1914, Vol. L X X V II, p. 733.
83 It is a convenient feature of arithmetic means computed from relatives based on average prices over a
period of years that the mean of these means for the base period must be 100—again barring discrepancies
caused by dropping fractions. For example, the arithmetic means of the Bureau cf Labor Statistics oldstyle index numbers for the 10-year period 1890-1899 would always add up to 1,000.0, had all the fractions
been kept and had all commodities been quoted in every year of the decade. If medians made from these
figures add up to 1,000.0 in 1890-1899, it is accidental.
METHODS USED IN M A K IN G INDEX NUM BERS.
73
upper half of the list to the lower half, or vice versa. But any
change of this character, large or small, will make the median jump
over the whole interval between its former position and that of the
next highest or next lowest relative price, unless the change happens
to place a new item within these limits. In large groups such erratic
jumps are less likely to occur, because the intervals between the
median and its nearest neighbors are usually slight.89 (4) If the num
ber of commodities included in an index number is even, the position
of the median may be indeterminate, though within a determinate
range.
Most of the advantages and defects of arithmetic means have been
mentioned incidentally, but it is well to list them all together: (1)
Arithmetic means (and aggregates of actual prices) stand first in
ease of computation, when an adding machine is available, especially
when the items are to be averaged first in small and later in large
groups. (2) Their familiarity to all readers is supposQd to be an
advantage in work intended for wide reading though perhaps this
familiarity means a dangerous lack of curiosity rather than clear
understanding of the figures. (3) They can themselves be averaged
and manipulated algebraically in various other ways.90 On the other
side of the score it must be said (4) that arithmetic means are liable to
distortion from the occurrence of a few extremely high relative prices,
(5) that arithmetic means of relative prices can not consistently be
shifted from one base to another without recomputation in full,91
and (6) that they may conceivably give contradictory results con
cerning the direction in which prices are moving, according as relative
prices are computed on one base or on another.92
Concerning the numerical value of the three averages under dis
cussion, it can be proved that the geometric mean is always less than
the arithmetic. On the other hand, the median may be either above
or below the ari thmetic mean, and likewise either above or below the
«9 “ This objection is met,” says Prof. Edgeworth, “ by denying that the interval between two adjacent
observations at the middle of the group is likely to be ‘ considerable’ ; large relatively to the magnitude
with which it is proper to compare that interval—that is, the minimum mensurable, as we may say—that
interval which is equal to (or of the same order as) the smallest degree which the compared method of
measurement is capable of distinguishing with accuracy. For this minimum we may take at the least
the ‘ probable error’ incident to the arithmetic mean. „ That the interval between adjacent observations is
likely to be small compared with this minimum is sufficiently evidenced by the following proposition:
When the number of observations (n ) is large the interval at the middle of the group, which is as likely as
not vacant, within which it is an even chance that no observation falls, is most probably very small com
pared with the probable range of the arithmetic mean (in the ratio of about 1: j n ) . When the number of
observation is not large the proposition is less accurate. But it remains roughly true, as the number can
not be supposed very small consistent with the applicability of the theory of probabilities.” Economic
Journal, June, 1918, Vol. X X V I II, p. 193.
Granting the justness of these general remarks, the writer has found several cases in his own work where
the medians of groups numbering 25 or more items moved in a way not representative of the whole array.
For examples see ‘ ‘A critique of index numbers of the prices of stocks,” Journal of Political Economy, July,
1916, Vol. X X I V , pp. 674, 675. It may, indeed, be set down as an advantage of medians that working with
them may bring the full array of fluctuations under the eye and lead to the detection of peculiarities which
would have escaped notice had arithmetic means been employed. When medians are used in averaging
small groups the practice of scrutinizing the whole set of data is strongly recommended as a means of guard'
ing against the occasional cases of erratic movement.
90 See, for example, G. U. Yale, Introduction to the Theory of Statistics, pp. 114-116.
91 See section 8 below.
92 Take, for example, the following data:
1913
Wheat, per bushel.........................................
Com, per bushel............................................
50 1
.48 j
1914
$1. 00
.24
T4‘
TH E M A K IN G AND USING OF INDEX NUM BERS.*
geometric mean. For example, if the relative prices of the 145 com
modities represented in the second index number oi Table 6 be aver
aged in these- three ways, the results are as follows for 1913:
Geometric mean, 125.7; median, 126.9; arithmetic mean, 131.3.
A more striking example of differences among the averages was
incidentally remarked above. The relative prices of Table 15 yield
the f ollowing figures:
Geometric meanr 194; median, 191; arithmetic mean, 217.
A fuller study of the relations between medians and arithmetic
means is provided for by the following table.93 In the chain index 94
the two forms of average never quite coincide; the median is smaller
in 20 cases and larger in 3; it is also steadier than the arithmetic mean
in the sense that it indicates an average annual change of 2.22 per
cent from prices in the preceding year, as against 3.64 per cent for
the arithmetic mean. In the fixed-base series for 1890-1913, in
cluding 145 , commodities, the median is likewise steadier than the
arithmetic mean, showing a smaller percentage of change, except dur
ing the middle nineties, when the price level was at its lowest. The
second series for these years illustrates the behavior of medians and
arithmetic means when used to average small groups. Here the
median is greater than the arithmetic mean in 13 years, the same
in 1 year, and less in 10 years. Moreover, it shows a greater aver
age change from one year to the next than the arithmetic mean.
Finally, the median drops 9 points in 1913 while the arithmetic mean
rises 2 points. Scrutiny of the full array of relative prices in this year
as compared with 1912 shows that this violent drop is not an apt
Then compute index numbers on the basis 1913—100:
1
1913
Wheat, relative prices...................................
Corn, relative prices.......................................
Index numbers................... ...............
1914
100
100
200
50
20a
250
125
100
Also,,compute index numbers on the basis 1914=100:
1943
1914..
Wheat, relative prices...................................
Com, relative prices.....................................
50
200
100
100
Index numbers....................................
250
125
200
100
Thus it appears that prices were 25 per cent higher in 1913 than in 1914 and also that they were 25 per cent
higher in 1914 than in 1913.
Much stress is often laid upon illustrations of this sort, but they are not seriously damaging to the good
repute of arithmetic means when properly interpreted. What they really say is: The arithmetic mean
variation of prices from 1913 to 1914 may conceivably be upward in percentages of prices in 1913, and at the
same time be downward in percentages of prices in 1914. N oreal inconsistency is involved in that statement
to one who can keep the meanings of the two results in mind. It should be added that cases in which such
apparent inconsistency occurs, while common in theoretical discussions, seldom if ever occur in the practi
cal computation of wholesale-price index numbers. In retail-price indexes they are not unknown. An
example has been pointed out in the British Board of Trade’s reports upon cost of living of the working
classes. See the reviews by J. M. Keynes in the Economic Journal, September and December, 1908.
93 For numerical examples of geometric and arithmetic means computed from the same data, see F. Y .
Edgeworth, “A defense of index numbers/' Economic Journal, Vol. VI (1896), p. 137, and A. W . Flux,
“ Modes of constructing index numbers,” Quarterly Journal of Economics, Vol. X X I (1907), p. 627.
94 On the character of chain indexes, see the following section (pp. 81 to 91).
75
METHODS USED IH M A K IN G m D B X NUM BERS.
summary of the combined movements.93 The figures for prices dur
ing the period of irredeemable paper money (1862-1878, inclusive)
show how far arithmetic means may depart from the medians when
a few commodities attain very high relative prices. The maximum
difference occurs in July, 1864, when the arithmetic mean exceeds the
median by 42 points, or more than 20 per cent. This excessive dif
ference is due to the high prices of cotton, tar, and other southern
products. It is precisely in cases such as this that the median is
distinctly safer to trust than the arithmetic mean.
T able
1 6 .—COMPARISONS OF MEDIANS AND ARITHM ETIC MEANS AS AVE R A G E S OF
R E L A TIV E PRICES.
[Data from Bulletin No. 149 of the Bureau of Labor Statistics.]
Chain index number
("prices in preceding
year= 100).«
Year.
Medians.
1890................................................
1891................................................
1892................................................
1893................................................
1891................................................
1895................................................
1898................................................
1897................................................
1898................................................
1899................................................
1900................................................
1901................................................
1902................................................
1903.............................................. .
1904................................................
1905................................................
1906................................................ ^
1907................................................
1908................................................
1909................................................
1910................................................
1911................................................
1912................................................
1913................................................
±0
-3 .1
±0
-7 .1
- 2 .4
- 1 .2
±0
+ 1.8
+5. 5
+7. 5
-1 .5
+ 2 .2
+ 1.3
±0
+ .7
+5* 1
+ 3 .9
- 3 .8 ±0
+ 1.5
- .9
+ 1.0
+ .5
Arithmetic
means.
+
+
+
+
+
+
+
+
+
+
+
+
+
0.2
4.4
.2
8.7
1.5
2. 8
.2
4.8
10.4
9.4
1.1
4.6
1.2
.1
2.9
5; 8
6.0
5.6
3.2
4.1
1.9
3.4
1.2
Averages, 1890-1899'....................
1900-1909...................
1910-1913....................
Average change from one year
to the next...................... \ ___ ;
a Compare Tables 2 and 17.
2.22
3.64
Relative prices of 145
commodities (aver
age prices in 18901899= 100). &
Medians.
Arithmetic
means.
Relative prices of 25
commodities
(aver
age prices in 18901899=100).e
Medians.
Arithmetic
r means.
112
111
107
104
96
94
90
91
94
100
109
107
110
111
• 112
114
119
129
119
121
124
125
.127
127
114
113
106
105
96
93
89
89
93
103
111
110
114
114
114
116
122
130
121
124
131
130
134
m
116
109
106
102
90
94
89
92
99
108
117
112
115
112
124
126
131
133
125
13ft
126
131
136
127
115
112
103
103
92
95
88
90
96
107
113
111
116
118
122
123
130
133
124
133
133
129
140
142
100
115
126
100
118
132
101
123
130
100
122
136
3.61
&Compare Table 6, second series.
4.13
5.70
5.09
c Compare Table 6, fifth series.
» Of the 25 commodities. 13 rose in pfico and 12 fell; the median percentage of change from prices in the
year before is 4-1.0.
THE M A K IN G AND USING OF INDEX NUMBERS.
76
Table
1 6 .—COMPARISONS OF M ED IAN S AN D AR IT H M E T IC M EANS AS A V E R A G E S OF
R E L A T IV E PRICES—Concluded.
[From W . C. Mitchell, Gold Prices and Wages under the Greenback Standard, pp. 59, 60.]
92 commodities at wholesale (prices in 186G=100).
Year.
Arith
Me metic
dians. means.
1860, January
April___
J u ly ....
October.
1861, January
April. . .
July----October.
1862, January
A pril...
J u ly ....
October.
1863, January
April. . .
J u ly ....
October.
1864, January
A pril.. .
J u ly ....
October.
1865, January
April. . .
J u ly ....
October.
1866, January
April. . .
J u ly ....
October.
100
100
100
100
100
96
96
97
100
100
100
111
125
137
134
135
156
169
194
200
216
190
158
175
182
173
181
173
102
102
100
102
100
98
95
103
115
112
120
126
142
160
155
155
179
197
206
183
205
199
186
191
188
Year.
1867, January
A p ril.. .
J u ly ....
October.
1868, January
A p ril.. .
J u ly ....
October.
1869, January
A p ril.. .
J u ly ....
October.
1870, January
A p ril...
J u ly ....
October.
1871, January
A pril.. .
J u ly ....
October.
1872, January
April. . .
J u ly ....
October.
1873, January
A pril.. .
July----October.
Arith
Me
dians. metic
means.
169
166
150
162
158
162
154
159
159
159
158
153
147
140
132
135
133
131
130
129
133
140
130
133
135
137
130
131
179
175
170
172
171
176
165
166
165
165
158
157
152
146
145
143
142
140
137
139
141
145
139
143
142
144
140
140
Year.
1874, January
April. . .
J u ly ....
October.
1875, January
April. . .
J u ly ....
October.
1876, January
April. . .
July___
October.
1877, January
April. . .
July----October.
1878, January
April. ].
J u ly ....
October.
1879, January
April. . .
J u ly ....
October.
1880, January
A p ril.. .
July___
October.
Me Arith
dians. metic
means.
130
129
130
130
127
125
121
120
117
115
110
108
114
108
100
102
99
96
90
94
140
141
138
138
138
132
129
127
122
122
118
117
121
118
114
110
107
105
99
102
88
100
84
85
95
108
107
102
101
103
114
116
110
111
Average change from one quarter to the next: Medians, 5.66 points; arithmetic means, 5.65 points.
Wise choice of the average to use in making an index number,
then, involves careful consideration of the materials to be dealt with
and of the purpose in view. (1) If that.purpose be to measure the
average raho of change in prices, the geometric mean is the best;
indeed, in strictness, it is the only proper average to employ— on one
interpretation of that somewhat indefinite problem. For, alone
among our averages, the geometric mean always allows equal in
fluence to equal ratios of change in price, quite irrespective of the
previous levels of the prices in question, the amounts of money rep
resented by the changes themselves, or any other factor. As has
been said already, in a geometric mean the doubling of one price is
precisely offset by the halving of another price— though if the two
prices were originally the same the rise amounts in money to twice
the fall. And further changes of 10 per cent from the two new prices
will again be precisely equal in their influence upon a geometric
mean, although 10 per cent of the price that has doubled represents
a sum of money four times as great as 10 per cent of the price that
has been halved. (2) But these same examples show that geometric
means are not proper averages for measuring alterations in the
amount of money that a given bill of goods costs. And as a rule our
interest does center in the money cost of goods rather than in the
average ratio of changes in price. For example, when we are inves
tigating the increased cost of living, the doubling of one item in
the family budget may well be twice as important as its halving;
and when we are studying the “ relation of prices to the currency, a
METHODS USED IN M A KIN G INDEX NUM BERS.
Ch a r t
77
11.—A COMPARISON OF MEDIANS AND ARITHM ETIC MEANS OF THE R E L A TIV E
PRICES OF 145 COMMODITIES.
(Based on Table 16.)
TH E M A K IN G AND USING OF INDEX NUMBERS.
large upward variation should count for more than a small down
ward variation, for it requires more currency; ” 96 provided always
that the changes in prices are not offset or more than offset by con
trary changes in quantities bought. For such purposes the arith
metic mean is the logical average to use. (3) Frequently, however,
the very fact that an article has advanced greatly in price cuts down
its market, so that the increase in money cost represented by the
arithmetic mean exists on paper rather than in fact.97 When such
cases of extreme advance are numerous among the relative prices to be
averaged, the median may give more significant results than the
arithmetic mean. (4) When the number of commodities included
in the index number is small, however, medians may occasionally
prove erratic, representing less the general trend of prices than the
peculiarities of the data from which they are made. (5) If the index
number is designed for the public at large, the familiarity of arith
metic means is an argument in their favor; but it counts for nothing
in the case of figures intended for specialists. (6) Often the useful
ness of a new index number may be enhanced without detriment to
its srpecial purpose by throwing it into a form directly comparable
with that of index numbers already in existence. Then, of course,
not only the form of average but also the base period employed in
making the existing series has special claims for imitation. (7) Fi
nally, the desirability of making index numbers that can be shifted
from one base to another without raising difficult problems of inter
pretation, deserves more consideration than is commonly accorded
it. On this count the score is in favor of the geometric mean. If
geometric means were invariably used, all index numbers could
readily be compared with one another, whatever the bases on which
they were originally computed. And that would be a great gain to
all students of priees.
No single form of average made from relative prices, then, is with
out its merits and its defects. Can we not escape the necessity of
relying upon any one of them by giving up the use of relative
prices and falling back upon aggregates of actual prices %
Index numbers made on this latter plan practically compel the
compiler to adopt a systematic scheme of weighting. This should
constitute a great safeguard against crude work, though in view of
Bradstreet’s method o f weighting one can not claim that it always
is effective. For the haphazard weighting involved in merely adding
up the raw quotations for different commodities in terms of their
ordinary commercial units is far more dangerous than the haphazard
weighting involved in using the same materials after reduction to relalative prices.98 It is true that sums in dollars and cents are likely to
run in amounts awkward for comparison; but these sums can quickly be
turned into a series of relatives on the scale of 100. The same device
96 Irving Fisher, The Purchasing Power of Moisey,, revised; edition, p. ^ 6 , note 2. Mr. Flux and Mr. Yule
hold that to measure changes “ m the money cost af the things we buy” is “ the retail-prices problem,”
and is not the appropriate aim of a wholesare-price index; but they do not consider the arguments which
Prof. Fisher advances. Journal of the Royal Statistical Society, March, 1921, pp. 175-9, and 200, 201.
97 Such cases might be met by reducing the weight allowed the article in question; but we have seen thal
revising weights blurs the meaning of the index number, by making it impossible to say how far the finat
results measure the change in prices and how far they measure the change in weights. (See p. 65.)
98 See the example from Hunt's Merchant’s Magazine given on pp. 31 and 32. However, a very rough
system of weights based upon guesswork may give quite as good results as the haphazard weighting of
relative prices. Prof. Irving Fisher suggests to the writer a “ lazy man's index number,” made by adding
actual prices for ordinary commercial units, with their decimal points shifted forward or backward, or
left unchanged, according to the estimated importance of each article.
METHODS USED IN MAKING INDEX NUMBEBS.
n
meets the objection that the introduction of new commodities, neces
sary at intervals in any large index number that is kept up to date, dis
turbs a sum of actual prices more than it disturbs an average of rela
tive prices. This statement is valid because the quotations for new
commodities, however adjusted, are just so much added ta the old
sum; while the relative prices of new commodities may be either
above or below the old average, and often exercise but a trifling net
effect upon its value. But by noting the ratios between the sums of
actual money which include and which exclude the new commodities,
and by using these ratios to adjust successive aggregates, the compiler
meets this difficulty quite as well as if he were averaging relatives
from the start.
The technical difficulties attending the construction of index num
bers made of actual prices, then, can be surmounted. Offsetting these
difficulties are numerous and substantial advantages. Aggregates of
money prices weighted according to the importance of the several arti
cles are even easier to understand than arithmetic means of Relative
prices. They are less laborious to compute than any other form of
weighted series, for no relative prices are used; the original quotations
are multiplied directly by the physical quantities used as weights,, and
the products added together. They are not tied to a single base
period; but from them relative prices can quickly be made upon
the chain system or any fixed base that is desired, and these relative
prices themselves can be shifted about at will as readily as geometric
means." Hence they are capable of giving direct comparisons
between prices on any two dates in which an investigator happens:
to be interested. Hence, also, they can be compared with any index
numbers covering the same years, on whatever base the latter are
computed.. They can not be made to give apparently inconsistent
results like arithmetic means. When published as sums of money,
they can be added, subtracted, multiplied, divided, or averaged
in any way that is convenient. When comprehensive in scope and
weighted on a sound system, they are not likely to be unduly dis
torted by a very great advance in the price of a few articles, and yet,
unlike medians, they allow every change in the price of every article
,9 The legitimacy of shifting these relatives by the “ short” method is best shown by the use of symbols*
p'o ^ 4 ^ p /y}rePreseri*: k*16 money prices of the two commodities p and p' in three years o, x, and y.
Then tiie sums of these actual prices will be—
p o+ p 'o in the year o.
Px+P'x in the year x.
Py+P'y in the year y.
Relative prices in the year x computed from these sums will be—
p 0- f p r0011
^as*s °* prices in the year o, and
g + £ .?on the basis of prices in the year u.
Py+Py
Relative prices in the year y will be— .
Py+P yQn
of prices in the year o.
Now the relative price in the year x, computed on the basis of prices in the year oy can be turned into
the relative price for the year x ©n the basis of prices in the- year y, by dividing the relative for the year
x on the basis of prices in the year o by the relative for the year y on tha basis of prices- in the year o-. Foe
p x + p 'x . Py+P'y P x+ P rx
P o+ P 'o Po+P'o Py^P^y
The reason why ordinary arithmetic means of relative prices can not be consistently shifted to another
base by this simple method is explained on p. 83.
80
TH E M A K IN G AND USING OF INDEX NUMBERS.
to influence the result. In fact, they combine most of the merits
and few of the defects characteristic of the various methods of averag
ing relative prices.
But the main issue has still to be faced. Do we wish to know
how certain sample prices have changed on the average, or do we
wish to know how the total cost of a sample bill of goods has changed ?
This is practically the same question we considered on pages 76 to
78 in discussing how best to average relative prices. And the answer
given there is valid here. If our interest really lies in measuring
average ratios of change, then geometric means are best. But (lj
the unfamiliarity of this average outside technical circles is itself an
objection to measuring average changes in an index number designed
for wide use, and (2) a measure of change in the money cost of goods
probably serves more uses than a measure of average ratios of. change
in prices. Now, the weighted aggregate of prices is the best measure
of change in the money cost of goods; it is better in several ways
than the simple arithmetic mean of relative prices, and in addition
it has all the merits of the latter form of average. For the relatives
which can be computed from these aggregates with little trouble are
identical with arithmetic means of relative prices, when the latter are
weighted by the money value of the physical quantities used as
weights for the corresponding actual prices.
This identity of the variations of a weighted aggregate of actual
prices and the arithmetic-mean variations of similarly weighted rela
tive prices can readily be demonstrated. Suppose that we have
collected the price quotations and the quantities to be used as weights
in an index number, and have decided what period to make the base
for comparisons. Then if we want an aggregate of actual prices, we
merely multiply the quotations of each commodity at each date by the
physical quantities used as weights, and add these products. To
measure the variations of these aggregates in terms of prices at the
base period, we have only to divide the aggregate for each period by
the aggregate for the base period. But if we plan to make a weighted
arithmetic mean of price variations, we begin by turning the quota
tions into relative prices; That is, we divide the actual price of each
commodity at each date b v its price in the base period. Then we
weight these relatives, not by physical quantities as in the first case,
but by the money values of the physical quantities at the prices of
the base year. But in this step the prices of the base year, which
were just used as divisors to get relative prices, are used again as
factors by which the relative prices are multiplied. Hence our results
are the same as if we had neither multiplied nor divided by the prices
of the base year; in other words, the same as if we had multiplied the
quotations of each commodity in each year by the physical quantities
used as weights. But that is just what we did when we set out to
make an aggregate of actual prices. So far, then, the two processes
are identical in their outcome. And the remaining steps are also the
same. The products must be added, and the sums divided by the
physical quantities used as weights times the actual prices of the base
year. Therefore, to make relative prices from aggregates of actual
METHODS USED IN M A K IN G INDEX NUM BERS.
81
prices is a shorter way of getting the same results as are obtained by
making similarly weighted arithmetic means of relative prices.1
But while an arithmetic mean of relative prices is always equiva
lent to some aggregate of actual prices turned into relatives, this fact
does not mean that the arithmetic mean of relatives is as desirable
a form of general-purpose index number as its rival. For the par
ticular aggregate of actual prices to which a given arithmetic mean
of relatives corresponds is one difficult to grasp. It is that aggregate
in which the price of each commodity included, quoted in terms of
its ordinary commercial unit, has been multiplied by the number of
commercial units which is necessary to make its price in the base
period some predetermined multiple of 100. Now this is a much
more complicated idea to carry in mind and to make clear to readers
than the idea of the price of the commodity multiplied by the num
ber of units that are ordinarily produced, exchanged, or consumed.
In other words, the arithmetic average of relatives has the same
relation to its corresponding aggregate of actual prices that a com
plicated mathematical expression has to the same expression .reduced
to a simpler form. The difference is one of form, but simplicity of
form greatly increases the efficiency of thinking.
8. BASE PERIODS, CHAIN INDEX NUMBERS, AND FIXED-BASE SERIES,
When relative prices are used it is necessary to select the quota
tions of some given period as a base. The actual prices in this base
period are called 100; all antecedent and subsequent prices are
divided by the base prices, and the quotients, multiplied by 100,
make the relatives which are usually summed and divided by the
number of commodities to get the final index number. When aggrei The explanation given in the text may be put in the form of algebraic formulas for readers willing to
study symbols.
Let p 0, Px)
V'o> p'x Irepresent the prices of the commodities from which an index number is to be made in the
» «| base year o and m some other year designated by the subscript r.
^ o’ x|
Let q, qf and <p respectively represent the physical quantities of these commodities to be used as weights.
Then an unweighted arithmetic mean of relative prices is represented by the following formula, in wnich
n stands for the number of commodities included:
. . .3?
’ ™
V jc.V'x
Vo P’ o
_0
n
A weighted aggregate of prices reduced to relatives is represented by the following formula:
Px q + p 'x q '+ .
. p” <P
Vo q + p 'o < ? ' + . . . p no qn
A weighted arithmetic mean of relative prices with money weights corresponding to the physical weights
of the expression immediately above is represented by the'following formula:
fo(qPo)+po(.S'p'o)+ . . . h (tn p n
^)
_______________ ______________ o___________
Vo q + v'o q’ + • . . • p n qn
o
But in the numerator of this fraction, p 0, p '0, and p n cancel out. Then formula (3) becomes identical with
formula (2). That is, the weighted aggregate of prices gives the same results when turned into relative
as the weighted arithmetic mean of relative prices, and gives them with less work.
33226°— 21— Bull. 284------- 6
82
THE M A K IN G AND USING OF INDEX NUMBERS.
gates of actual prices are first made and then turned into relatives the
problem of selecting a proper base period has to be faced at the end
of the computation.
In some cases the prices of a single day have been used as the base?
but as a rule average prices for a year, five years, a decade, or an even
longer period have been preferred. For this preference there is a
simple justification when arithmetic means are used as averages of
the relative prices.2 If the price of any commodity happens to be
unusually high or unusually low in the base period, its relative prices
at other periods will be correspondingly low or high, and the very
high relative prices will exercise much more influence upon arith
metic means tlian the very low ones. If an appreciable proportion
of the commodities in the list be very high or very low, the final
index number may be distorted. Though numerically correct, the
results have less significance than if they showed changes in terms
of prices that men consider “ normal.” 3 Of course exceptionally
high or exceptionally low quotations are less likely to last for a year
than for a day, and less likely to last for a decade than for a year.
The period chosen as base for the relative prices should be that
period with which accurate comparisons are most significant for the
purpose in hand. Probably most users of general-purpose index
numbers prefer to make their comparisons with recent dates. Hence
the case for “ chain” indexes is very strong— that is, for indexes
like the medians of Table 2, which sliow the average rise or fall of
prices on the basis of prices in the preceding year.4 Hence, also,
any index number with a fixed base becomes in one respect less signifi
cant the longer it is maintained. For example, when the Bureau of
Labor Statistics series was established in 1902, the public was inter
ested to know how much prices in that year had changed in terms of
average prices in the decade 1890-1899. In 1918, however, when
people cared less about knowing changes in terms of what prices had
been 19 to 28 years earlier, the Bureau shifted its base to 1913. Sim
ilarly, Sauerbeck’s index number, which uses prices in 1867-1877 as a
base, suffers in significance for recent comparisons because it forces
one to make all comparisons in terms of prices in a period that ended
before most of the people now living were old enough to know the
meaning of prices.
Index numbers made on a base many years in the past, moreover,
encounter all the difficulties that inhere in the problem of measuring
price variations through long periods of time. As was shown in
Section II I of this bulletin (pp. 11 to 23), price variations become
dispersed over a wider range and less concentrated about their mean
as the time covered by the variations increases. That is, the longer
a fixed-base series is maintained, the more scattered as a rule be
come the relative prices. The difficulty is particularly serious
when arithmetic means are used. The commodities that have a con2
If geometric means are used the ratios between the index numbers for different dates are not influenced
at all by the selection of the base, affidif medians are used they are likely to be affected but slightly, provided
the number of commodities included be large.
s The selection of a proper base period, however, does not guarantee immunity from the exercise of undue
influence by certain articles. More important than the base is the choice of proper weights. Or, to speak
with more precision, the choice of base is itself part of the problem of weighting in its inclusive sense.
4 This form of index number was invented by Prof. Alfred Marshall. See Contemporary Review, March,
1887.
METHODS USED IN M A K IN G INDEX NUM BERS.
83
sistent long-period trend gradually climb far above or fall far below the
average relative price. Then the high relative prices of the com
modities that have risen exercise a much stronger pull upon the
position of the arithmetic mean than do the low relative prices of the
commodities that have fallen. For most purposes this constitutes
a defect, since commodities that have increased greatly in price are
likely to have become scarce, and commodities that have become
cheaper are likely to be more abundant. The changes in the influ
ence exercised on the mean by the relative prices are likely to be in
inverse ratio to the changes in the importance of the commodities.
In other words, the use of the distant base itself introduces a sur
reptitious set of weights into the figures to be averaged, and a set
which may well counteract in large measure the formal set of weights
which the investigator uses to show the importance of his articles.
It is not uncommon, of course, to shift fixed-based index numbers
from a remote to a recent base. For example, Sauerbeck’s index as
continued by the Statist was 85 in 1913 on the 1867-1877 base. If
one wishes to find how much English prices rose in 1914-1918 as
compared with their prewar level, he may put 85 = 100, and recast
the indexes for the years of war on that scale. But this is a purely
formal manipulation of the results. It does not diminish the scat
tering of the relative prices from which the averages are computed,
and it does not give the same result that recomputing the relative
prices of the 45 commodities on the 1913 base and averaging them
afresh would give. The first point is obvious; the second requires
explanation.
Averages of relative prices on a given base may be regarded as
averages of actual prices made with a peculiar scheme of haphazard
weights. That is, the quotation of every commodity is in effect mul
tiplied by the factor necessary to make its price in the base period
equal 100.5 To change the base is of course to change this set of im
plicit haphazard weights for another set, which may be better or
worse— the computer is unlikely to know which— but which will be
different unless the ratio of change in prices between the old and new
base periods has been precisely identical for all the commodities
included. Of course, different sets of weights applied to the same
set of price quotations will probably alter the average variations
somewhat. Hence, if one really wants to know how a given set
of prices have varied with reference to their standing at any given
time, the only way to find out accurately is to weight the varia
tions of each commodity by the factors which the chosen base de
termines; that is, in practice, to compute new relative prices article
by article. But if the purpose in hand is such that one set of hap
hazard weights will serve as well as another, then there is no objection
to shifting the base by the short method of manipulating merely
the averages, provided the results are properly explained.
5 Compare F. R . Macaulay, “ Index numbers for retail prices/’ American Economic Review. Decem
ber, 1915, Vol. V, pp. 928, 929.
84
TH E M A K IN G . AND USING OF INDEX NUM BERS.
It is easy to arrange examples in which wide discrepancies appear
between the results of the two methods of shifting the base.6 But
the difficult and the important thing is to find out how serious the
discrepancies are in actual practice. For to use index numbers
effectively, it is often necessary to shift the base, and sometimes the
short method must be followed, either because recomputation in full
requires a prohibitive amount of labor, or because the original data
necessary for recomputation have not been published. The next
table gives three pertinent examples. In the first case when Sauerbeck’s index is shifted from 1867-1877 = 100 to 1890-1899 = 100 the
discrepancies are fairly regular and rather small both absolutely and
relatively. In the last case, when the same series is shifted to 1860 =
100, the discrepancies are highly irregular from year to year, and are
rather large both absolutely and relatively— several times exceeding
5 per cent of the recomputed figures. In the remaining case the
discrepancies are small absolutely, though often large relatively to
the recomputed figures, and also highly variable from year to year.7
The conclusion which these experiments suggest is that the two
methods almost always give different results; that the discrepancies
are by no means constant from year to year in a given case, and that
their magnitude both absolutely and relatively differs much from one
case to another. Hence it is well to avoid the short method of
6 For example, suppose that an index number includes only wheat and corn, and that their prices are as
follows:
j
Wheat, per bushel__
Corn, per bushel........ j
1913
1914
$1.00
.40
$0.50
.40
If 1913 be made the base, the relative prices and index numbers will be:
1913
1914
Wheat,relative prices.
Corn, relative prices..
100
100
"
50
100
Sums...................
Index numbers..........
200
100
150
75
If now the base be shifted from 1913 to 1914 by the short method, the index number for 1913 will be
(lOO-s-75) 100= 133$. But if the figures be recomputed on the basis of prices in 1914, the result is an index
number of 150 in 1913:
1913
1914
Wheat,relative prices.
Corn, relative prices..
200
100
100
100
Sums...................
Index numbers..........
300
150
200
100
7 The discrepancies shown in the table do not result wholly from the mathematical inconsistency of the
short method, but partly from the fact that when an index number is shifted to a new base by recomputa
tion in full it is commonly impossible or undesirable to utilize all the original data. Some commodity, for
example, may not be quoted for the dates used as the new base, and therefore has either to be dropped or
introduced at a later dat e by means of some doubtful assumption as to what its price would have betn had
it been quoted for the full period. Of course this observation makes the objection t o using the short method
stronger rather than weaker. It means that this method often leads the statistician into uses of the original
data which he would have avoided had he undertaken the recomputation of the index number.
METHODS USED IN M A K IN G INDEX NUM BERS.
85
shifting bases whenever possible; and when that method must be
used, its results should not be treated as showing what the index
number would have been had it been made originally on the new base.
EX AM P LES OF DISCREPANCIES B E T W E E N THE RESULTS OF TW O METHODS
OF SHIFTING THE BASES ON W H IC H IN D E X NUM BERS AR E COMPUTED.
T a b l e 1 7 .—
(Arithmetic means.)
Sauerbeck’s index number,
1890-1913.
Bureau of Labor Statistics
index number
(old series).
Recom
Bu
Year. Orig Shifted puted
reau’s Chain
to
inal
basis Dis- series on index
1890- on1890form,
crepmade
1867- 1899= 1899= an- basis
by
short
1877= 100,by 100,by cies. 1890short
1899=
100.
method.
method. long
100.
method.
1890..
1891..
1892..
1893..
1894..
1895..
1896..
1897..
1898..
1899..
1900..
1901..
1902..
1903.J
1904..1
1905..'
1906.. 1
1907.J
1908..
1909..
1910..
1911..
1912..
1913..
72
72
68
68
63
62
61
62
64
68
75
70
69
69
70
72
77
80
73
74
78
80
85
85
109
109
103
103
95
94
92
94
97
103
114
106
105
105
106
109
117
121
111
112
118
121
129
129
i
i
i
i
i
i
1
109
109
103
103
95
94
92
93
97
104
115
107
106
106
108
111
119
123
112
114
120
123
130
130
1
1
1
1
1
1
2
2
2
2
1
2
2
2
1
1
112.9
111.7
106.1
105.6
98.1
93.6
90.4
89.7
93.4
101.7
110.5
108.5
112.9
113.6
113.0
115.9
122.5
129.5
122.8
126.5
131.0
129.2
133.6
135.2
- 1 .1
- 5 .0
- .5
-9 .0
-2 .6
- 3 .4
- .8
+ 4 .1
+ 8; 9
+ 8 .7
-1 .8
+ 4 .1
+ .6
- .5
+ 2 .6
+ 5.7
+ 5.7
- 5 .2
+ 3 .0
+ 4 .0
-1 . 8
+ 3 .4
+ 1.2
J
Chain
index
Dis
made crepan
cies.
iby
long
method.
- 0.2
- 4.4
.2
- 8.7
— 1. 5
- 2.8
+ .2
+ 4.8
+ 10.4
+ 9.4
- 1.1
+ 4.6
+ 1.2
.1
+ 2.9
+ 5.8
+ 6.0
— 5.6
+ 3.2
+ 4. 1
- 1.9
+ 3.4
+ 1.2
0.9
.6
.3
.3
1.1
.6
.4
.7
1.5
.7
.7
.5
.6
.4
.3
.1
.3
.4
.2
.1
. 1
]
Sauerbeck’ s index number,
1860-1891.
Year.
Orig Shifted Re
to
com Disinal
1860
form, = 100, puted crepon
1867basis anby
1877=
cies.
short
1860=
100.
method. 100.
1860.
1861.
1862.
1863.
1864.
1865.
1866.
1867.
1868.
1869.
1870.
1871.
1872.
1873.
1874.
1875.
1876.
1877.
1878.
1879.
1880.
1881.
1882.
1883.
1884.
1885.
1886.
1887.
1888.
1889.
1890.
1891.
1
99
98
101
103
105
101
102
100
99
98
96
100
109
111
102
96
95
94
87
83
88
85
84
82
76
72
69
68
70
72
72
72
100. 0
99.0
102.0
104.0
106,1
102. 0
103.0
101.0
100. 0
101. 0
99.0
97.0
101.0
112.1
103.0
97.0
96.0
95.0
87.9
83.8
88.9
85.9
84.9
82.8
76. 8
72.7
69.7
68.7
70. 7
72.7
72.7
72.7
100. 0
99.6
105. 5
109.3
112.3
105.8
106.5
103.9
103.1
101.9
100.3
102.6
112.5
116.6
107.0
100.3
97.5
97.4
91.2
86.7
91.8
88. 5
88.0
86.0
79.3
75.4
72.4
70. 7
73.9
76.7
76. 0
75.4
0.6
3.5
5.3
6.2
3.8
3.5
2.9
3.1
2.9
3.3
1.6
2.4
4.5
4.0
3.3
1.5
2.4
3.3
2.9
2.9
2.6
3.1
3.2
2.5
2.7
2.7
2.0
3.2
4.0
3.3
2.7
Chain index numbers on the base, prices in the preceding year = 100*
have the advantage pointed out in Section III, that the variations
which they represent are highly concentrated and therefore apt for
. averaging. That is, year-to-year variations are relatively easy to
measure with approximate accuracy. It is true that makers of
index numbers find chain relatives more troublesome to compute than
fixed-base series, since most of the prices used as divisors change
every year; but that fact weighs lightly with such laborious folk
in comparison with an improvement in their results. Why, then,
should they not make successive averages of vear-to-year variations
covering as iong a period as desired and weld the successive links
together by multiplication to form a continuous chain ?
For example, in Table 17 it is shown that the old Bureau of Labor
Statistics index in 1890 on the 1890-1899 base was 112.9 and that
prices fell 0.2 per cent in 1891. On multiplying, we get 112.9 X 0. 998
THE M A K IN G AND USING OF INDEX NUM BERS.
86
= 112.7. In 1892 the average change of prices was a fall of 4.4
per cent. 112.7x0.956 = 107.7. Once more, in 1893 prices fell
0.2 per cent on the average. Adding this new link to the chain, we
have 107.7x0.998 = 107.5. The next table shows this process
carried through to 1913. The result is a new index number covering
24 years, in which each successive step is taken b y averaging rela
tives which are probably better fitted for averaging, since they are
more highly concentrated, than the corresponding relatives on the
1890-1899 base. Is it not better than the old index on the fixed
base?
One may answer, first, that while each successive step in the chain
index may be taken with confidence, any errors which do inhere in
the steps are likely to accummulate. There is no magic in the
year-by-year computation which makes the final comparison be
tween prices in 1913 and 1890 more reliable on the one basis than on
the other. Second, the interpretation of the final result is certainly
simpler in the case of the fixed-base than in the case of the chain
index. The figures say in the first case that between 1890 and 1913
there was an average net increase of prices equal to 22.3 per cent of
average prices in 1890-1899. The chain index says that there was an
increase between these two years of 37.1 per cent; but when one asks,
“ Per cent of w hat?” the answer is complicated. Third, the chain
index which was begun arbitrarily on a par with the fixed-base series
drifts away from it upward, and by the end of the period has opened
a gap of nearly 15 points, or more than 11 per cent— a notable dis
crepancy. Stated in another way, the chain series makes the per
centage increase in prices from 1890 to 1913 more than half again
as great as the fixed-base series makes it.
1 8 — A F IX E D -B A SE IN D E X N UM BER , A CHAIN IN D E X NU M BER M ADE FROM
TH E SAME D A T A , AN D TH E CHAIN IN D E X M ADE INTO A CONTINUOUS SERIES.
T able
[Data from Bulletin No. 149 of Bureau of Labor Statistics.]
(Arithmetic means.)
Year.
..........
►
1890.
1891.
1892.
1893.
1894.
t
1895.
1896.
1897.
..........
..........
..........
I
I
1900.
1901.
Bureau's
index num
ber on basis
prices in
1890-1899=
100.
112.9
111.7
106.1
105.6
96.1
93.6
90.4
89.7
93.4
101.7
110.5
108.5
Chain in
Chain in
dex num
dex num
ber, on basis ber made
into a con
prices in
preceding
tinuous
year= 100.
series.
99.8
95.6
99.8
91.3
98.5
97.2
100.2
104.8
110.4
109.4
98.9
112.9
112.7
107.7
107.5
98.2
96.7
94.0
94.2
109.0
119.3
118.0
Year.
1902.................
1903.................
1904.................
1905.................
1906.................
1907.................
1908.................
1909.................
1910.................
1911.................
1912.................
1913.................
Bureau’s
index num
ber on basis
prices in
1890-1899=
100.
112.9
113.6
113.0
115.9
122.5
129.5
122.8
126.5
131.6
129.2
133.6
135.2
Chain in
dex num
ber, on basis
prices in
preceding
year= 100.
104.6
101.2
99.9
102 9
105.8
106.0
94.4
103.2
104.1
98.1
103.4
101.2
Chain in
dex num
ber made
into a con
tinuous
series.
123.4
124.9
124.8
128.4
135.9
144.1
136.0
140.3
146.1
143.3
148.2
150.0
Why should the annual shifting of the base on which.relatives are
computed make such a difference in the results ? On looking at the
figures in Table 17 from which the continuous chain in Table 18 is
forged, we see that when prices are falling the percentage of change
on the preceding-year base is generally smaller than the corresponding
change on the fixed base. On the contrary, when prices are rising
METHODS USED IN M A K IN G INDEX NUM BERS.
87
the preceding year base gives the larger percentage of change. In
two years the percentages are the same (1912 and 1913), and in two
other years the rule is reversed (1908 and 1911); but the rule holds in
19 cases out of 23.8 The problem is to account for the fa<?t that chain
relatives usually rise more than fixed-base relatives when prices are
rising and fall less when prices are falling.
The following numerical examples give the clue to the solution.
We have in the first two columns of each example two relatives on
a fixed base, for two successive years. First the larger of the two
relatives is made to increase 25 per cent in the second year, and then
to fall 25 per cent in the second year, the smaller relative remaining
constant. Afterwards the smaller of the two relatives is made to
rise and then to fall by 25 per cent in the second year, the larger
relative being constant. In the third column the figures for the sec
ond year are turned into chain relatives. Index numbers are com
puted for both sets of relatives and the percentages of change on the
two bases are given.
1.
When a relative above the average of the relatives rises, its rise
makes a smaller percentage addition to the chain than to the fixedbase index.
Fixed base.
First year.
<
Second year.
Preceding-year base—
Second year.
240
160
300
160
125
100
2)400
2)460
2)225
200
230
112.5
Per cent of change........ +15
'
i
5
|
Percent of change.. +12.5
2.
When a relative above the average of the relatives falls, its fall
makes a smaller percentage subtraction from the chain than from the
fixed-base index.
Fixed base.
First year.
Preceding-year base—
Second year. '■
Second year.
240
160
180
160
75
100
2)400
2)340
2)175
200
170
Per cent of change
.
'
87.5
—15
Per cent of Change. . . —12.5
s The fact was pointed out and the explanation of it suggested by Professor F. R. Macaulay, in American
Economic Review, March, 1916, Vol, VI, pp. 207, 208.
88
THE M A K IN G AND USING OF INDEX NUMBERS.
3.
When a relative below the average of the relatives rises, its rise
makes a larger percentage addition to the chain than to the fixedbase index.
Fixed base.
Preceding-year base—
First year.
Second year.
240
160
240
200
100
125
2)400
2)440
2)225
200
220
Second year.
112.5
Per cent of change.. +12.5
Per cent of change........+10
!
4.
When a relative below the average of the relatives falls, its fall
makes a larger percentage subtraction from the chain than from the
fixed-base index.
Fixed base.
Preceding-year base—
First year.
Second year.
Second year.
240
160
240
120
100
75
2)400
2)360
2)175
200
180
87.5
Per cent of change........ —10
Per cent of change.. —12.5
All that these figures show is that in certain cases the fluctuations
will be greater in the chain relatives and in other cases greater in the
fixed-base relatives. The vital point is, however, that cases 2 and 3
occur in price quotations much more frequently than c.ases 1 and 4.
Relative prices above the average seem more likely to fall than to rise
further; relative prices below the average seem more likely to rise than
to fall further. That is, the prices of individual commodities tend to
conform to the average movement, and when they have already di
verged from this average they move back toward it more often than
they* move away. These cases that occur more frequently than the
others are those that make the chain relatives rise more (case 3) or
fall less than the fixed-base relatives (case 2) .9
The net difference to be expected on this ground in a large body of
quotations between the movements of the relatives on the two bases
9Of course this argument can be more generally, as well as more compactly, stated in algebraic terms.
Prof. W . M. Ogburn contributes the following formulation:
Letp'i,
. . . stand for relative prices of commodities during the first year, and p'2,
. . .
stand for relative prices of commodities during the second year. Let n be the number of commodities
and mi the arithmetic mean of the relative prices during the first year.
The fixed-base index is obtained'by getting the average of the relative prices; the fixed-base index for
the first year is:
P\'+P \" + __•__;__•
n
And that for the second year is:
Pi'+Ih"+ • • •
n
METHODS USED IN M A K IN G INDEX NUM BERS.
89
is small in any one year. A glance at the figures in Table 18 will show
that the observed differences are generally less than 1 per cent. But
though small the differences are tolerably constant in direction, and
therefore when cumulated by multiplication they become significant
in 10 or 20 years.
The conclusion is that close agreement is not to be expected
between efforts to measure the change of prices between years far apart
when the measures are made first on a fixed base and then by the cnain
method. The ch&in method is perfectly legitimate, of course, when
its results are carefully interpreted; but, as remarked above, the
interpretation is difficult to put into words. Where means permit it
is well to make from the original quotations two series of index num
bers, one a chain index, the other a fixed-base series, and then to call
attention to the differences between the two.
The per cent increase, or the rise, is the ratio of the second to the first, or
J g /= g ? !.+ P 2" + . '
'
'
(1)
Pi'-»-Pi"+ • • •
Let the per cent increase from first year to second in the prices of individual commodities be r ',r " • • •
then the relation between the prices during the first and second year can be expressed by the following
equations:
Pi'=*Pi' (1 +r')
p 2” = p i " (1 + r ” ), etc.
By substitution in (1):
Rf-
.
p i '( l + r ') + p l" ( l + r " ) +
2j>/_P i '+ P i " +
■ ■
• • •
Pif+ P i " +
•__• • + p i ,r ' + p l" r ” +
• ■ .
P i'+ P i"
Rf
'ZPi+?P'r _.j.+ 2 Pir
2pi
2pi
Putting p\—m\=Xi where x\ is the size of the relative, we have:
,2(mi+zi)r
+
2pi
R f= 1+^
2 Pl + 2 px
2pi= 7177*1
7lTfl\
2pj
K )
i + 7l+ 2p i
The chain index is obtained by averaging the ratios of the individual prices of commodities and is ex*
pressed in the following manner:
P?'
Rc=?V\
Rc=
P /'
pa'__________
Pi'(l
+ r ') _____________
, p i" Q + r ” )'_ l. . . .
_______
Px
__________
P i"
n
” +n2 r - l + n
?
( l + r ' ) + ( l + r " ) -f
n
» • •
(3)
by subtracting (3) from (2):
In other words the fixed-base index number will not equal the chain index number unless 2ziri= o
(which is true when r is constant). When 2 x1^ is negative the chain-index number will be larger and
when positive the fixed-base index will be larger. 2xiri is positive when x (the size of the relative) is
correlated (positively) with r (the percentage of increase), which is rarely if ever the case. The exact
difference can be measured by
90
THE M A K IN G AND USING OF INDEX NUM BERS.
Even this combination, however, is far from meeting all the needs
of users of index numbers. For certain users may require for special
purposes accurate measurements of price fluctuations in terms of the
price level in any given month or year, or any given stretch of time
in the whole period covered by the investigation. If such users are
few as compared with all the people who note or quote the popular
index numbers, they are precisely the few most interested in price
fluctuations and most likely to increase knowledge by their use of the
figures. But of course compilers can not foresee what base periods
would serve best all these special purposes, and they can not be
expected to work out index numbers on all the bases made possible
by their original data. It is therefore highly desirable to have index
numbers that can be shifted from one base to another readily and
without involving difficulties of interpretation.
It is this desideratum, in large part, that has led to the recent
reaction against index numbers made by striking arithmetic means
of relative prices and in favor of index numbers made by adding
actual prices. For the latter form of index, being a sum of dollars
and cents, with an explicit scheme of weights, can be thrown into
the form o f a series of relative prices on any base that is desired, with
slight labor and with no ambiguity. Geometric means, of course,
possess the same advantage.
Another problem in base periods has recently been developed by
Prof. Fisher. Should the period to which the weights refer be the
same as the period used as the base for computing relative prices,
or should the weights be taken from a different period? Suppose
that the index number is to be an arithmetic mean of relative prices
weighted by the values of the commodities exchanged in some year.
Then “ if the weights used are the values of the base year (that is,
the base year for the relative prices) they impart a downward bias
to all the index numbers of any given year calculated thereby , while,
on the other hand, if the weights used are the values of the given
year itself, they impart an upward bias”
To understand this effect one must note that the commodities
which have unusually high market prices in the base year will tend to
have both high values (prices multiplied by quantities) in that year
and low relative prices in other years. Vice versa, the commodities
which have unusually low market prices in the base year will tend to
have both low values in that year and high relatives in other years.
Then the multiplication of the low relatives by the high values and
of the high relatives by the low values will tend to reduce the index
numbers for all other years in comparison with the base year. Chang
ing the weights from values in the base year to values in any other year
will tend to reverse these combinations. For commodities that have
unusually low market prices in the base year and therefore high rela
tives in other years will tend to have higher values in the latter years,
and the commodities with high market prices in the base year and
low relatives in other years will tend to have lower values in the latter
years. The index number with “ given-year” weights will therefore
tend to combine high relatives heavily weighted and low relatives
lightly weighted, and so give figures that run high for all other years
in comparison with the base year.
How considerable this “ biasing” of the results by the choice of the
period to which the weights refer will prove in practice depends upon
METHODS USED IN M A K IN G INDEX NUM BERS.
91
whether the prices and quantities of commodities usually fluctuate in
the same or in opposite directions, for the influence of high and low prices
on the values as weights may be offset, or more than offset, by contrary
changes in the quantities. Little is positively known concerning
the run of these facts. Prof. Fisher believes, however, that the
quantity factor is almost as likely to influence the weights in one
direction as in the other. If so, the price factor has a fair field to in
fluence the values used as weights and the above argument holds
good.
On this basis Prof. Fisher advises that in making arithmetic means
of relative prices the weights be taken from the base year, in order
that the downward bias of these weights may run counter to the
upward bias of the arithmetic mean (caused b y the greater influence
exercised by high than by low relatives upon this form of average).
Harmonic means, on the contrary, have a downward bias (are more
influenced by low than by high relatives) and should therefore be
weighted by values taken from some other year than the base.
Geometric means, medians, and modes, which have no inherent bias,
he holds, should be weighted by values both in the base and in the
given year; for otherwise they will be affected by the bias of the
weights.10
9. THE “ IDEAL” FORMULA.
A more complicated formula for making index numbers than those
heretofore discussed has recently been invented independently by
three high authorities and recommended as the best for making
general-purpose series. It may be written thus:
t>
__
J^Pntfn ,
l n ~ M X p 0qn
2p0q0
where X indicates “ the sum of such terms as”
f n = the price of any commodity in a given year (or period).
^ = t h e quantity of tha4) commodity in the given year.
p0—the price of that commodity in the base year.
g0= the quantity of that commodity in the base year.11
To use this formula it is necessary to have data concerning the
prices and the quantities of every commodity in every year covered
by the index number. From these data four sets of aggregates of
actual prices multiplied by quantities are made for each year:
(1)
Prices in the given year times quantities in the given year, (2) the
same prices times quantities in the base year, (3) prices in the base
Irving Fisher: “ The best form of index number.” Quarterly Publication of the American Statistical
Association, March, 1921, pp. 535, 536.
Prof. W . M. Persons has tested Prof. Fisher’s contention that a geometric mean weighted by prices in the
base year will have a downward bias. He finds that “ Indices of quantity or of prices of agricultural prod
ucts of the United States for the period 1880-1920 when measured relative to a fixed base (1910 in this case)
show the same general movement whether the Fisher method or the geometric average is used . . .
no
cumulative divergence of the two indices is evident.” —Review of Economic Statistics, May, 1921, p. 111.
11 Mr. Walsh mentioned this formula in a footnote in his Measurements of General Exchange Value, 1901,
but did not then exploit its merits. In 1912 Prof. A. C. Pigou published the same formula in Wealtn and
Welfare (p. 46); but failed to note that the square root of the product should be extracted. This oversight
he remedied in his Economics of Welfare, 1920 (p. 78). In 1921 Prof. Irving Fisher having invented the for
mula in his turn, presented it before the American Statistical Association. Meanwhile Mr. W alsh in review
ing his earlier work had concluded that his footnote formula was perhaps the best of all. (See Quarterly
Publication of the American Statistical Association, March, 1921, pp. 536,539, and “ The Problem of Esti
mation,” p. 102.)
1 have adopted Prof. Persons's notation as clearer than that of the inventors.— Review of Economic
Statistics, May, 1921, p. 107, note.
92
THE M A KIN G AND USING OF INDEX NUMBERS.
year times quantities in the given year, and. (4) the latter prices
times quantities in the base year. Then the first and second aggre
gates (prices in the given year weighted in-two ways) are reduced to
relatives by dividing them respectively bv the third and fourth
aggregates (prices in the base year weighted in the same two ways).
Finally these relatives are multiplied together and the square root
of their product extracted.
What advantages does this formula possess to compensate for the
great amount of labor it entails ?
Prof. Pigou uses it in an index of changes in the volume of “ real
income.” He finds it necessary to use weights for two periods be
cause of “ The root fact . . . that in the first period our group
expends its purchasing power upon one collection of commodities,
and in the second period it expends it on a second and different col
lection. ” The change in real income can not be accurately measured
unless these alterations in the quantities of goods bought are repre
sented in the index of prices used in reducing money income to real
income.12
Prof. Fisher wants this formula for use in his equation of exchange.
It serves admirably there, because an index number of prices made
by it when multiplied by a similarly constructed index number of
quantities will show the changes in the total values of goods
exchanged.
Mr. Walsh's purpose is more general, “ to measure variations in
the exchange value or purchasing power of m oney,” and his argu
ment concerning its merits is more technical. The first of the two
ratios included in the formula is equivalent to an harmonic mean of
relative prices weighted by values in the given year, while the second
ratio is equivalent to an arithmetic mean of price relatives weighted
b y values in the base year. B y using imaginary examples covering
four years, in which the last year has the same prices and quantities as
the first year, Mr. Walsh tests arithmetic and harmonic means
weighted in his way. He finds that they yield different results
which “ lie on opposite sides of the truth, and apparently equally
above and below it proportionately.” This result suggests the pro
priety of taking the geometric mean between the two averages.
That step yields the “ ideal” formula. Mr. Walsh adds: “ Note that
it involves the arithmetic average, the harmonic average, the weight
ings of the first and second periods, and the geometric mean. . . .
It
seems to contain everything that could be desired.” 13
We may agree with Prof. Pigou that this formula is well adapted to
use in a measure of change in real income and with Prof. Fisher that
it is well adapted to use in the equation of exchange. Can we agree
with Mr. Walsh that it is the best formula for making generalpurpose index numbers ?14
If the end in view is to compare the change in prices between any
two years, then this formula is more desirable than an aggregate of
actual prices weighted by quantities in either year alone. That
proposition holds true o f every year-to-year comparison however
far extended. Hence the “ ideal” formula is admirably adapted for
making chain index numbers, whenever it is possible to secure the
Economics of Welfare, p. 72.
is The Problem of Estimation, p. 102.
n Mr. Walsh is explicit upon this point.
12
(See The Problem of Estimation, p. 118.)
METHODS USED IN M A K IN G INDEX NUMBERS.
93
necessary annual data for quantities as well as prices and to meet
the necessary expense of computation.
But can the separate links in such a chain index be welded together
to make an equally admirable index covering long periods? Two
objections lie against it on this score. (1) The ideal formula changes
weights in each successive link in the chain. The quantities for 1920
and 1921 used in computing the link for that year are not likely to be
the same as the quantities for 1921 and 1922 used in computing the
latter link. As pointed out in section 6 above a change in the weights
makes it uncertain what part of the net result is due to price fluctua
tions and what part to fluctuations in quantities. Whenever the
purpose in view requires that the price factor shall be isolated, it is
therefore undesirable to use the “ ideal” formula for any comparisons
except those between two specified years.15 (2) It has been shown
in section 8 that an arithmetic mean of relatives on the precedingyear base when forged into a continuous chain drifts away upward
from the corresponding fixed-base series made from the same data.
Now the ideal formula does not use relative prices, but is made from
aggregates of actuals which can not drift in this fashion, provided they
are made with constant weights. Does the annual change of weights
required by the “ ideal” formula introduce errors that cumulate and
so cause the chain index to part company from a fixed-base series?
Prof. Persons has answered that question by an actual trial. Taking
the prices and quantities of 12 leading crops of each year of the decade
of 1910-1919, he has made first for the quantities and second for the
prices two index numbers, one using the “ ideal” formula computed
directly to the fixed base 1910, another using the “ ideal” formula
chain fashion. Both of the chain indices are found to diverge from
their fixed-base mates by a distance that is rather wide considering
that the errors are cumulated for no more than nine years. The chain
index for quantities drifts upward and the chain index for prices
drifts downward. In both cases the discrepancies reach 4 per cent
in 1919.16 Hence the “ ideal” formula is ill-fitted for making index
numbers covering a long period of years, when it is applied in the way
which its logic strictly requires, namely, year-by-year comparisons.
And a fixed-base series made by this formula— that is, one in which
the index for each year is made by compounding the weights of that
year with some base year (instead of the year before)—yields accurate
comparisons only between the base year and any given year and not
comparisons that are accurate as between any two given years. If it
is desired to make possible comparisons between any years of a period
longer than two years aggregates of actual prices or geometric means,
both made with constant weights, seem better than the “ ideal”
formula, as well as far easier to compute.17
is This objection is reduced but not removed if the indices for each year are computed directly to a fixed
base, say 1913. Then the prices for the year 1920 would be weighted by quantities in 1913 and 1950, the prices
in 1921 by the quantities in 1913 and 1921, etc. The weights would still change, but not so much as m the
chain index.
16 Review of Economic Statistics, May, 1921, pp. 113, 114.
17 Concerning the difference in labor of computing Prof. Persons gives an interesting note. The relative
times required to compute the “ ideal ’ index numbers and the geometric means in his test of the two wereas
follows:
Relative times
required.
Geometric means, constant weights................................................................................................. 27
“ Ideal” index number, fixed base.................................................................................................... 51
“ Ideal” index number, chain series................................................................................................. 100
Of course the difference would be much larger if the time were counted in that is spent in collecting yearly
data concerning quantities called for by the “ ideal” formula. A sum of actual prices made with fixed
weights takes still less time for computation than a weighted geometric mean.
THE M A K IN G AND USING OF INDEX NUM BEKS.
94
V.—A COMPARISON OF THE LEADING AMERICAN INDEX
NUMBERS FOR THE YEARS 1890 TO 1918.
Many of the threads running through the preceding sections can
be woven into a comparison oi the best-known index numbers cur
rently published in the United States— a comparison having intrinsic
interest of its own, as well as making a fitting summary of Part I of
this bulletin.
1. ANALYSIS OF THE SIMILARITIES AND DIFFERENCES BY YEARS, 1890
TO 1918*
Three general-purpose index numbers are available for the critical
study proposed, the latest form of the Bureau of Labor Statistics
series, feradstreet’s index, and Dun’s index. It seems hardly worth
while to include in the comparison index numbers made solely of
the prices of foods, because they do not profess to measure changes
in the commodity markets at large. It has been shown that these
special indexes are not in close agreement with series containing not
only foods but also minerals, forest products, textiles, chemicals, etc.;
and that demonstration need not be repeated.18
The first step toward comparing index numbers is to throw them
into similar form and establish them upon a common base. The
new series of the Bureau of Labor Statistics is a weighted sum of
actual prices, turned into relatives on the base, prices in 1913 = 100.
This series can be shifted to any base desired without appreciable
loss in accuracy. Dun’s and Bradstreet’s series are sums of actual
prices, and have no base of their own. Accordingly they may be
recast into relatives on the base, the average of the original figures
for 1890-1899 = 100. Dun’s figures for this decade average $84.32.
B y dividing the published figures by this sum and multiplying the
results by 100 we can make a new series strictly comparable with the
rest of our material. Shifting Bradstreet’s series is less satisfactory,
because it does not begin until 1892. The best that can be done is
to. equate Bradstreet’s average for 1892-1899 with the average made
from the Bureau’s figures for these years— that is, to put $6.7785 =
97.1— and then to apply the rule of three.19
These three series in comparable form are assembled in Table 19.20
is See subdivision 5, “ The numbers and kinds of commodities included,” especially pp. 52-56.
19 No violence is done by this procedure to Bradstreet’s series; but the comparision is not quite satis
factory, because our other series were not worked out on the basis, prices in 1892-1899=97.1, and would prob
ably have shown slightly different results if they had been.
20 The annual averages, made from the original figures published by Dun and Bradstreet’s, run as follows:
Year.
1890.
1891.
1892.
1898.
1894.
1895.
1896.
1897.
1898.
1899.
1900.
1901.
1902.
1903.
1904.
1905.
1906.
Dun’s.
$90.9
92.2
90.0
92.4
84.7
81.3
76.0
74.0
78.9
82.8
93.4
95.9
100.4
99.0
100.2
100.6
105.3
Bradstreet’s.
$7.78
7.53
6.68
6.43
5.91
6.12
6.57
7.21
7.88
7.57
7.88
7.94
7.92
8.10
8.42
Year.
1907..............................................
1908..............................................
1909..............................................
1910..............................................
m i
1912..............................................
1913..............................................
1914..............................................
1915..............................................
1916..............................................
1917..............................................
1918..............................................
Averages:
1890-1899..............................
1900-1909.............................
1910-1914..............................
1915-1918..............................
...........................
« Average of 1892-1899.
Dun’s.
Bradstreet’s.
$111.8
109.9
117.8
119.2
116.8
124.4
120.9
122.2
126.4
148.8
204.1
229.2
$8.90
8.01
8.52
8.99
8.71
9.19
9.21
8.90
9.85
11.83
15.66
18.73
84.3
103.4
120.7
177.1
a 6.78
8.11
9.00
14.02
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
95
The second and third divisions of the table bring out certain dif
ferences among the figures, and the summaries in the latter part
show the average or net movements in various periods.
T able
1 9.—A COMPARISON OF THE CHIEF AM ERICAN IN D E X NUMBERS FOR TH E
Y E A R S 1890 TO 1913.
The three index num
bers shifted to the
1890-1899 base.
Year.
Percentage differences
among the three in
dex numbers.
Bradstreet’s
greater
Bureau
( + ) or
of
less (—)
BradLabor
than
Dun’s.
street’s.
Statis
Bureau
tics.
of
Labor
Statis
tics.
Dun’s
greater
( + ) or
less (—)
than
Bureau
of
Labor
Statis
tics.
Percentage by which
each of the three index
numbers rose ( + ) or
fell (—) each year.
Bradstreet’s
Bureau
greater
of
Brad( + ) or street’s. Labor Dun’s.
less (—)
Statis
than
tics.
Dun’s.
Period of decline.
1890.......................................
1891.......................................
1802.......................................
1893.......................................
1894.......................................
1895.......................................
1896.......................................
Ill
108
96
92
85
Ill
111
103
106
95
95
90
108
109
107
110
100
96
90
+ 7 .8
+ 1.9
+ 1.1
-3 .2
—5.6
—2.7
—1. 8
+ 3 .9
+ 3 .8
+ 5.3
+ 1.1
±0
+
-
88
94
103
113
108
113
114
113
116
121
127
115
122
129
125
132
132
128
141
91
95
101
109
108
116
117
117
117
121
128
125
133
136
130
138
137
136
139
88
94
98
111
114
119
117
119
119
125
133
130
140
141
139
148
143
145
150
- 3 .3
- 1 .1
+ 2 .0
+ 3 .7
±0
-2 .6
-2 .6
-3 .4
— .9
±0
- .8
-8 .0
-8 .3
- 5 .1
-3 .8
- 4 .3
-3 .6
-5 .9
+ 1 .4
-3 .3
- 1 .1
-3 .0
+ 1.8
+ 5.6
+ 2 .6
±0
+ 1.7
+ 1.7
+ 3 .3
+ 3 .9
+ 4 .0
+ 5.3
+ 3.7
+ 6 .9
+ 7 .2
+ 4 .4
+ 6 .6
+ 7 .9
± o
± 0
+ 5.1
+ 1.8
- 5.3
- 5.0
- 2.6 '
- 5.0
- 2.5
- 3.2
— 4. 5
-1 1 .5
-1 2 .9
- 8.5
-1 0 .1
-1 0 .8
7.7
-1 1 .7
- 6.0
3.5.
6.8
9.6
9.7
4.4
4.6
.9
.9
2.7 :
4.3
5.0
9.4
+
6.1
+ 5.7
- 3.1
+
5.6
± o
- 3.0
+ 10.2
169
224
268
170
241
269
176
- .6
242 :„• - 7 .1
272 1 - . a :
+ 3 .5
+ .4
+ 1.1
-
+ 19.9 +22.3
+ 32. 5 : +41.8
+19.6 +11.6
3.7
1.8 '
4.0
4.2
5.6 ‘
- 2.7
-1 1 .1
- 4.2
- 7.6
± 0
':
— 7.2
+ 2.9
-1 0 .4
± o
- 5.3
+
+
+
-
.9
1.8
2.8
9.1
4,0
6.3
Period of gradual rise.
1897................................
1898.......................................
1899.......................................
1900.......................................
1901........................................
1902.......................................
1903.......................................
1904........................................ 1
1905........................................ j
1906....................................... !
1907....................................... i
1908....................................... ;
1909........................................
1910........................................
1911.......................................
1912.......................................
1913........................................
1914.......................................
1915.......................................
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
±
d =
+
+
+
+
+
+
I
1.1 | - 2.2
4.4 1 + 6 . 8
6.3
+ 4.3
7.9
+ 13.3
+ 2.7
.9
+ 4.4
7.4
.9
- 1.7
+ 1.7
o
o
± 0
+ 5.0
3.4
5.8 i + 6.4
- 2.3
2.3
6.4
+ 7.7
2.3
+ .7
4.4
- 1.4
6.2
+ 6.5
.7
- 3.4
.7
+
1.4
2.2
+ 3.4
Period of accelerated rise
due to war.
1&16.......................................
1917.................................. .
1918.......................................
4.0
7.4
1.5 ;
+ 17,3
+37.5
+ 12.4
96
THE M A K IN G AND USING OF INDEX NUM BERS.
T ab le 1 9 .— A
COM PARISON OF T H E CH IEF AM ER ICAN IN D E X NU M BER S FOR T H E
Y E A R S 1890 TO 1918—Concluded.
The three index num
bers shifted to the
1890-1899 base.
Item.
Averages b y 5 - y e a r
periods:
1890-1894.......................
1895-1899.......................
1900-1904........ .............
1905-1909.......................
1910-1914.......................
1915-1918
Averages b y 1 0 - y e a r
periods:
1890-1899.......................
1900-1909.......................
1910-1918...............
Maxima and minima:
1890-1914—
Maxima.................
Minima..................
Differences............
1914-1918—
Maxima.................
Minima.-...............1
Differences............
Net rise ( + ) or fall (—):
1890-1896.......................
1896-1907.......................
1907-1908.......................
1908-1914.......................
1914-1918.......................
Algebraic averages:
1890-1894.......................
1895-1899.......................
1900-1904.......................
1905-1909.......................
1910-1914.......................
1915-1918.......................
1890-1914.......................
1890-1918.....................
Percentage variations
among the three in
dex numbers.
Brad Dun’s
street’s com
com
Bureau
pared pared
with
of
Brad Labor
with
Dun’s.
street’ s.
Bureau Bureau
of
Sta
of
tistics.
Labor Labor
Sta
Sta
tistics.
tistics.
Percentage variations of
the yearly rise and fall
of each of the three
index numbers.
Brad
Bureau
street’ s
of
com
Brad
pared street’ s. Labor Dun’s.
Sta
with
tistics.
Dun’s.
105
92
112
120
129
201
105
94
113
125
135
205
107
93
116
129
143
210
3.6
3.0
2. 5
3.6
9.5
2.4
3. 5
1. 7
2.3
3.6
5.8
3.2
3.2
3.0
3.9
6.9
10.0
4. 7
6.9
6.3
4.1
5. 5
3.5
20.6
5.1
3.4
3.4
3.6
2.9
19.5
3.7
4.7
4.8
4.3
2.7
17.7
(100)
116
161
100
119
166
100
123
173
3.3
3.0
3.6
2.6
3.0
4.6
3.1
5.4
7.5
6.5
4.8
11.1
4.2
3.5
10.2
4.2
4.5
9.3
132
85
47
138
90
48
148
88
60
8.3
0
8.3
7.2
0
7.2
12.9
0
12.9
11.1
0
11.1
10.4
0
10.4
13.3
0
13.3
268
128
140
269
136
133
272
145
127
7.1
.4
6.7
7.9
.4
7.5
11.7
1.5
10.2
32.5
3.0
29.5
41.8
.7
41.1
37.5
1.4
36.1
21
38
3
11
133
- 18
+ 43
— 3
+ 15
+ 127
+ 3 .6
-2 .2
—1.0
-3 .6
-4 .5
-1 .7
+ 1.7
—1.3
+ 2 .3
+ 3 .6
+ 5.8
+ 3 .2
.7
— .9
- 3.2
- 6.9
-1 0 .0
- 4.7
- 6.9
+ 1.6
+ 2.0
+ 1.7
+ 1.0
+20.6
— 3.7
+ 1.3
+ 3.1
+ 2.7
+ .5
+ 19.5
— 1.8
— .3
+ 4.1
+ 3.4
+ .8
+ 17.7
- 2 .0
-2 .0
+ 2 .4
+ 2 .5
-
+ .8
+ 3.9
+ 1.0
+ 3.6
+ 1.4
+ 3.7
- 26
+ 42
— 12
+ 13
+ 140
+
+
+
4.6
4.6
A cursory examination of this table, or a glance at Chart 12, shows
that these index numbers made by three independent organizations
have a marked family resemblance. They all agree that prices fell
heavily in 1890-1896, rose still more sharply in 1896-1900, wavered
uncertainly in 1901-1904, rose rapidly again in 1905-1907, fell in
1908, more than recovered their lost ground in 1909-1910, dropped
back in 191J, rose to a new high record in 1912, receded somewhat
in 1912-1914, and finally shot up at an extraordinary rate during the
war. Further, the three index numbers agree that the general level
about which the yearly oscillations clustered was higher m 1910-1914
than in 1900-1909, and higher in 1900-1909 than in 1890-1899.
About the major facts of price history, in short, the testimony of the
leading American index numbers is unanimous.
On the other hand, Table 19 shows that the series differ in many
details. For example, not once in the 29 years covered by the pres
ent record are all three index numbers identical, and in only six years
Chart 12.—IN D E X NUM BERS OF TH E BU R E A U OF LABOR STAISTICS, DUN, AN D B R A D STR E ET, 1890 TO 1918.
(Based on Table 19.)
33220°—21.
(To face page 90.)
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
97
are any two indexes precisely the same. On the average of the whole
period the Bureau of Labor Statistics series varies from Bradstreet’s
by 3.3 per cent, from Dun’s by 3.4 per cent, while Bradstreet’s index
varies from Dun’s by 5.4 per cent. The maximum differences in any
one year rise to 8.3 per cent between the bureau’s index and Brad
street’s (1909), 7.9 per cent between the bureau’s and Dun’s (1915),
and 12.9 per cent between Dun’s and Bradstreet’s (1909). Concern
ing the direction in w^hich prices move from one year to the next, the
bureau’s series contradicts Bradstreet’s in one year (1893) and Dun’s
series in four years, while Dun’s and Bradstreet’s indexes contradict
each other in six years. If we count cases in which one index re
mains the same for two successive years while another series rises or
falls, we find four years of partial contradiction when we compare
the bureau’s index with Bradstreet’s, three years when we compare
the bureau’s index with Dun’s, and two years when we set Brad
street’s against Dun’s. In general, the bureau’s index steers a middle
course between the other two, averaging 2 per cent higher than Brad
street’s and 2.5 per cent lower than Dun’s, while the margin by which
Dun’s index exceeds Bradstreet’s averages 4.6 per cent.21
Most of the detailed differences among the annual figures of the
three index numbers may be regarded as resulting from differences
in respect to (1) secular trend and (2) degree of variability from one
year to the next.
1.
Chart 12 and the averages by decades in Table 19 show that on
the whole Dun’s index number has risen more than the bureau’s,
and the bureau’s more than Bradstreet’s. This long-period shifting
of the level about which the monthly and yearly oscillations occur
is technically called the secular trend. Graphically it may be repre
sented by a straight line. Two turning points occur in the 29 years
covered by the table. The great fall of prices which began in 1873
ended in 1896 or 1897, and a rise began. In 1915 the rate of this
rise was violently accelerated by the war, so that the slope of the
straight line representing the direction of the secular trend was sud
denly made steeper. Of the three periods marked off by these turn
ing points in the first half of Table 19, the middle one, 1896-1914,
alone is long enough to make the computation of the secular trend
significant.
The secular trends of the three index numbers during this period of
19 years, given in Table 20, are represented collectively on Chart 13
and. are shown separately with their respective curves on Charts 14,
15, and 16. They are summarized in the following table:
T able 2 0 .—
SECULAR T RENDS OF IN D E X NUMBERS OF BU R E A U OF LABOR STA
TISTICS, B R A D S T R E E T , AND DUN, 1895 TO 1914— SUMMARY.
Index numbers.
Annual
geometric
increment
of secular
trend in
1896-1914.
Bradstreet’s...................................................
Bureau of Labor Statistics.........................
Dun’s..............................................................
1. 0230
1. 0232
1,0269
Geometric
mean in
1896-1914.
113. 7
117.1
120.3
Ratio of
annual
increment
to geome
tric mean
(per cent).
0.90
.87
.85
Terminal
points of the
straight line
representing
the secular
trend.
1896
1914
92.7
95. 3
94.7
139.6
144. 0
152.6
Net per
cent of rise
in lines of
secular
trend,
1896-1914.
These averages are made, of course, from algebraic sums of the yearly percentage differences.
33226°— 21— Bull. 284------- 7
15.1
15.1
16.1
98
THE M A K IN G AND USING OF INDEX NUM BERS.
It is primarily these differences in secular trend that make the
bureau’s index number follow a course intermediate between Bradstreet’s and Dun’s indexes.
Chart 1 3.—SECULAR TRENDS OF IN D E X NUMBERS OF B U R E A U OF LABOR STATIS
TICS, DUN , AND B R AD ST R EET, 1896-1914.
(Based on Table 21.)
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
Ch a r t
99
14.—IN D E X NUM BERS OF BR A D ST R E E T, COMPARED W IT H THEIR SECULAR
TREND, 1896-1914.
(Based on Table 21.)
100
Ch a k t
the
m a k in g
and
u s in g
of
in d e x
num bers.
15.—IN D E X NUM BERS OF B U R EAU OF LABOR STATISTICS, COMPARED W IT H
THEIR SECULAR T R E N D , 1896-1914.
(Based on Table 21.)
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
Chart 16— IN D E X NUM BERS OF DUN, COMPARED W IT H THEIR
1896-1914.
(Based on Table 21.)
101
SECULAR TREND ,
102
the
M a k in g
and
u s in g
of
num bebs.
in d e x
T able 3 1 .—IN D E X NUM BERS OF B R A D S T E E E T , TH E B U R E A U OF LABO R STATISTICS,
AN D D U N , COMPARED W IT H T H E IR SECULAR T R E N D S, B Y Y E A R S , 1896 TO 1914.
Bradstreet’s.
Bureau of Labor Statistics.
Excess of—
Excess of—
Year.
1896...............
1897...............
1898...............
1899...............
1900...............
1901...........
1902...............
1903...............
1904...............
1905...............
1906...............
1907...............
1908...............
1909...............
1910...............
1911...............
1912...............
1913...............
1914...............
Secu Index Secular
lar
num trend
over
trend. ber.
index
num
ber.
92.7
94.9
97.1
99.3
101. 6
103.9
106.3
108. 7
111.2
113. 7
116.3
119.0
121.8
124.6
127.0
130.4
133. 4
136.4
139. 6
85
88
94
103
113
108
113
114
113
116
121
127
115
122
129
125
132
132
128
Per ct.
9.1
7.8
3.3
Excess of—
Index Secu Index Secular
lar
num trend
num
ber
over
trend. ber.
index
over
secular
num
ber.
trend.
Index Secu Index Secular
lar
num
num trend
ber
trend. ber.
over
over
index
secular
num
trend.
ber.
Index
num
ber
over
secular
trend.
Per ct.
Per ct.
Per ct.
3. 7
11.3
3.9
6.,3
4.8
1.6
2.0
4.0
6.7
5. a
2.1
1.2
4.3
1.0
3.3
9.0
Dun’s.
95.3
97. 5
99.7
102.0
104. 4
106.8
109.3
111.9
114.5
117.1
119. 8
122.6
125.5
128. 4
131.3
134. 4
137.5
140.7
144.0
90
91
95
101
109
108
116
117
117
117
121
128
125
133
136
130
138
137
136
Per ct.
5.8
7.1
5.0
1.0
4.4
1.1
6.1
4.6
2.2
.1
1.0
4.4
.4
3.6
3.5
3.4
.4
2.7
5.9
94.7
97.3
99.9
102.5
105.3
108.1
111.0
114.0
117.0
120.3
123. 5
126.8
130.2
133.7
137.3
141.0
144.8
148.6
152.6
90
88
94
98
111
114
119
117
119
119
125
133
130
140
141
139
148
143
145
Per ct.
5.2
10.6
6.3
4.6
1.1
.2
1.4
5. 4
5.5
7.2
2.6
1.7
1.2
4.9
4.7
2.7
2.2
3.9
5.2
2.
While steadier over a considerable period of time, Bradstreet’s
index changes more from one year to the next than does either the
bureau’s or Dun’s series. Dun’s index, further, is more variable than
the bureau’s.
Several different ways of measuring year-to-year variations all sup
port this conclusion: (1) If the upercentage by which each of the
three index numbers rose or fell each year” as shown in Table 19 be
averaged from 1892 to 1914, the results are Bradstreet’s 5.15 per
cent, Dun’s 4.37 per cent, and the Bureau of Labor Statistics’s 3.71
per cent. (2) The standard deviations of these annual percentages
of rise and fall are, Bradstreet’s 5.79, Dun’s 5.06, and the bureau’s
4.46. (3) If the figures showing the excess of the secular trend over
the index number or the excess of the index number over the secular
trend in Table 21 be averaged for 1896-1914, the results are, Brad
street’s 4.0 per cent, Dun’s 4.0 per cent, the bureau’s 3.3 per cent.
(4) If the yearly deviations from the secular trend are plotted as in
Chart 17, it appears that Bradstreet’s fluctuates through the widest
and the bureau’s series through the narrowest range, Dun’s index
being intermediate.
To show that these index numbers differ in detail, however,-means
little. The significant problem is whether these differences are due to
the inherent difficulty of measuring changes in the price level, to the
crudity of the method of measurement in vogue, or to technical dif
ferences in the construction of the particular index numbers in
question.
Unfortunately it is not possible to attack this problem effectively
on the lines of analysis suggested in the preceding sections. For the
compilers of Bradstreet’s and Dun’s index numbers do not give suffi-
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
Ch a r t
103
17.—Y E A R L Y DEVIATION S FROM SECULAR TREND OF IN D E X NUM BERS OF
B U R E A U OF LABOR STATISTICS, D U N, AN D B R A D ST R EE T, 1896-1914.
(Based on Table 21.)
104
THE M A K IN G AND USING OF INDEX NUMBERS.
cient data concerning the sources of information drawn upon for
quotations, the commodities included and the weights employed for
each commodity to make possible a close comparison with the bu
reau’s series. Bradstreet’s publishes quotations for 106 commodities,
but bases its index number on the prices of 96, and does not say
which 10 are omitted. Its prices per pound, which are added up to
give the index number, were published for a short time in 1897, but
are not disclosed in recent years. Dun’s Review does not publish its
list of commodities, to say nothing of their prices, ana explains
merely that it weights by per capita consumption, allowing 50 per
cent of the total for foods, 18 per cent for textiles, 16 per cent for
minerals, and 16 per cent for other commodities.22 With such scanty
information about these two series, statements concerning the rea
sons for the relatively slight differences between each of them and the
bureau’s index number would be subject to a relatively wide margin
of error.23
After all, the important fact is that the three index numbers agree
with one another very closely. The divergencies which do appear
are smaller than those which result from most attempts to measure
economic quantities. For example, two sets of experts employed
upon a valuation case are likely to arrive at results farther apart than
the maximum differences shown in Table 19. Again it is doubtful
whether the margin of error in the average balance sheets of business
enterprises, or in cost computations is as narrow as the average mar
gin between Bradstreet’s and Dun’s index numbers, to say nothing
of the narrower margins between the official series and either of these
commercial indexes.
To sum up the comparisons in the most definite form the coefficient
of correlation must be used. This coefficient is the standard statis
tical device for measuring the degree of agreement or difference be
tween two variables. Its extreme limits are —1.0 and +1.0, the
latter expressing perfect agreement.24 When such coefficients are
computed for the annual index.numbers in 1892-1914, inclusive, the
following results are obtained:
Coefficients
of correlation.
Bureau of Labor Statistics index num ber and Bradptreet’s.
Bureau ctf Labor Statistics in dex num ber and D u n ’s .........
Bradstreet’s in dex num ber and D u n ’s ..........................................
+ 0 . 964
+ .992
-f .959
High coefficients of correlation are to be expected, of course, when
the variables compared are different measurements of the same
quantity— in this case the general level of wholesale prices through a
period of 23 years. To get such high coefficients as the preceding
indicates that the measurements made by different hands are in close
agreement and therefore presumably reliable.
A severer test may be applied'by computing the coefficients of cor
relation between the percentage changes in the three index num
22 Compare J. P. Norton’s article in the Quarterly Journal of Economics, Aug., 1910, Vol. X X I V ,p . 754.
23 Most of the analytic comparisons among various American index numbers in Bulletin No. 173
dealt with series much more perfectly known than Dun’s or Bradstreet’s. The reader who turns
back to that discussion will probably share the writer’s belief that were all the necessary data available,
the differences among the three series dealt with would be found to result primarily from differences in
the lists of commodities and in the systems of weighting. But that belief will remain a mere probability
so long as the construction of Bradstreet’s and Dun’s indexes is not fully disclosed.
24 Most statistical text books explain the method of computing the coefficient of correlation in detail.
See for example, G. Udney Yule, Introduction to the Theory of Statistics, 2d edition, 1912, chs. I X
and X .
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
bers from one year to the next.
follows:
105
The results of this operation are as
Coefficients
of correlation.
Bureau of Labor Statistics index number and Bradstreet’s .....................................
Bureau of Labor'Statistics index number and D u n ’s .................................................
Bradstreet’s index number and D u n ’s ...............................................................................
+ 0 - 882
-j- •873
+ • 788
Here the coefficients, though less than in the preceding case, are
still high. Bradstreet’s agrees a bit better with the bureau’s series
than does Dun’s, whereas in the former comparison Dun’s had dis
tinctly the higher correlation. In both comparisons, the bureau’s
series makes the best showing. Other things being equal, among
different measures of a given quantity, that measure has the best
claim to acceptance which is nearest the mean of all the measures.
In the present case, however, other things are not equal. The bu
reaus’s series includes more commodities than either of its rivals, its
system of weighting is better, and its method of construction from
start to finish is disclosed with a fullness which justifies confidence.
On these grounds its superiority is clear. The fact that it agrees bet
ter with both the commercial indexes than they agree with each other
merely confirms the choice which would be made on a priori grounds.
2. COMPARISON OF FOUR LEADING AMERICAN INDEX NUMBERS, BY
MONTHS, JULY, 1914, TO DECEMBER, 1918.
The peculiar interest attaching to the revolution in prices during
the World War makes desirable a more detailed comparison of the
leading American index numbers in 1914-1918. For this period, there
are available besides the three series discussed in the preceding section,
the index number compiled by the Price Section of the War Industries
Board.
Table 22 and chart 18 present the four series on a common base—
namely, average prices in the twelve months preceding the outbreak
of war (July, 1913-June, 1914) = 100, giving by months first the index
numbers themselves, and then the percentage by wiiich each of the
four index numbers rose or fell as compared with the month preceding.
Study of the table and of the chart based upon it shows at once a
closeness of agreement for which even the results of the preceding
comparison scarcely prepare one. And this impression of close agree
ment is abundantly justified when the coefficients of correlation are
worked out. These coefficients, shown on page 108, approach even
more closely to the limit of perfect agreement (4-1.0) than the remark
ably high coefficients we have found for the yearly index numbers in
times of peace.
106
T
able
THE M A K IN G AND USING OF INDEX NUM BERS.
2 2 .—A COMPARISON OF FOUR LEAD IN G AM ERICAN IN D E X
MONTHS, JULY, 1914, TO DECEM BER, 1918.
The four index numbers shifted
to the base, July, 1913-June,
1914-100.
Year and month.
NUM BERS, B Y
Percentage by which each of the
four index numbers rose ( + ) or
fell ( —) each month.
i
i
:
War Bureau
War Bureau
of
of
BradIndus .Labor
Brad- ;
Indus
Labor
Dun’s.
Dun’s.
tries
street's.
tries
Statis street’s.
Board. Statis
Board.
tics.
tics.
1914.
97
101
101
99
98
98
99 :
103
103
99
98 J
97]
97
103
106
101
100 ■
102 =
99
102
103
102
102
102
±0
+ 4 .1
+0
-2 .0
-1 .0
±0
+ 0 .8
+ 3 .2
+ .5
-4 .2
- .9
- .8
+ 0 .5
+ 6 .3
+ 2 .9
-4 .7
—1.2
+ 1 .6
- 0 .1
+ 3 .0
+ 1 .1
-L I
+ .3
- .1
100
100
100
100
100
100
102
102
102
104
107 :
111
98
100
99
100
101
99
101
100
99
101
103
108
105
108
109
110
109
110
110
110
111
114
118
120 I
103
103
103
104
104
103
103
103
103
106
108:
111
+ 2 .0
±0
±0
±0
±0
±0
+ 2 .0
±0
±0
+ 2 .0
+ 2 .9
+ 3 .7
+1*1
+ 1 .7 *
-L G
+ .7
+ 1 .0
- 1 .6
+ 1 .8
- .7
- 1 .3
+ 2 .4
+ 1. 5;
+ 2 .9 '
+ 3 .5
+ 2 .5
+ ,0
+ .9
- .2
+ .4
+ .4
- .3
+ .8
+ 2 .9
+ 3 .3
+ 2 .6
+ 0 .6
±0
- .2
+L0
+ <4
- .7
- .4
- .1
+ .6
+ 2.3
+ 2 .5
+ 2 .7
115
118
121
123
123
122
123
125
127
132
141
144
110
111
114
116
118
118
119;
123
127
134
143
146
123
126 1
129
132
131
130
129
130
133
139
148
153
115
117
118
120
120
119
119
122
125
130
137
139
+ 4 .2
+ 3 .6
+ 2 .6
+L2.
+ 2 .5 •' + 2 .3
+ 1 .7 : + 2 .0
±0
+ 1 .6
- .8
+ .4 ,
+ .4
+ •8
+ 1 .6
+ 3. 5
+ 1.6
+ 3 ,3
+ 3 .9
+ 5 .0
+ 7 .2
+ 6 .8
+ 2 .0
+ 2 .1
+ 2 .3
+ 2 .1 ;
+ 2 .8
+ 1.6
- .3
- .9
- 1 .1
+ 1 .1
+ 2 .6
+ 4 .3
+ 6 .5
+ 3 .5
+ 3 .4
+ 1 .6
+ 1 .2
+ 1 .4
- .1
- .4
- .5
+ 2 .4
+ 2 .8
+ 4 .2
+ 5 .0
+ 1.4
June.......................................................
July.......... f ...........................................
August...................................................
September.....................................
October..................................................
November.............................................
December..............................................
148
151
156
170
178
183
189
187
186
182
183
182
150
155
160
172
181
185
135
185
182
180
183
182
155
157
161
166
171
176
182
185
188
190
194
199
142
149
155
164
173
175
177
178
179
181
181
182
+ 2 .8 '
+ 2 .0
+ 3 .3
+ 9 .0
+ 4 .7
+ 2 .8
+ 3 .3
-1 .1
- .5
-2 .2
+ .5
- .5
+ 2 .8
+ 3 .4
+ 3 .2
+ 7 .2
+ 5 .4
+ 2 .0
+ .4
- .4
-1 .4
- 1 .1
+ 1 .4
- .5
+ 1 .0
+ 1 .5
+ 2 .3
+ 3 .4
+ 3 .0
+ 3 .1
+ 3 .0
+ 1 .8 =
+ 1 .0
+ 1 .3
+ 2 .0
+ 2 .6
+ 2 .4
+ 4 .8
+ 3 .8
+ 5 .9
+ 5 .7
+ .8
+ 1.5
+ .7
+ .2
+ 1 .3
+ .1
+ .3
1918.
January.................................................
February...............................................
March.....................................................
April.......................................................
May........................................................
June.......................................................
July........................................................
August...................................................
September............................................
October..................................................
November.............................................
December..............................................
385
187
188
191
190
189
193
196
201
201
201
203
186
187
188
191
191
193
199
203
207
204
207
207
202
202
204
209
212
214
215
214
213
212
212
210
190
187
188
188
186
188
191
191
192
191
190
189
+ 1 .6
+ 1.1
+ .5
+ 1.6
— .5
- .5
+2.1
+ 1. 5
+ 2 .5
±0
±0
+ 1.0
+ 2 .2
+ .7
+ .4
+ 1 .9
- .1
+ 1 .0
+ 2 .9
+ 2 .2
+ 2.1
- 1 .5
+ 1.3
±0
+ 1 .4
+ ,3
+ 1 .1
+ 2 .3
+ 1.5
+ .7
+ .6
- .6
- .3
- .4
+ .1
- 1 .0
+ 4 .3
-1 .4
+ .7
- .3
-1 .2
+ 1.3
+ 1 .6
+ .1
+ .3
- .5
- .6
- .1
August...................................................
September............................................
October..................................................
November............................................
December........... .................................
1915.
January.................................................
February...............................................
March................................................... .
April.......................................................
June.......................................................
July........................................................
August...................................................
September............................................
October..................................................
November.............................................
December..............................................
1916.
January.................................................
February...............................................
March...............................................
April.......................................................
May........................................................
June.......................................................
August....... ..........................................
September............................................
October..................................................
November.............................................
December............. ................................
1917.
January.................................................
February...............................................
March-...................................................
April.....................................................
i
COMPARISON OF U3ADING AMERICAN IN DEX NUM BERS.
107
Since both Bradstreet’s and Dun’s index numbers are computed
from prices as of the first of the month while the Bureau of Labor
Statistics and War Industries Board use average prices for the month
or prices at various dates within the month, it is not quite accurate
to compute coefficients of correlation from the figures as they stand
after shifting to a common base. To overcome this difficulty as well
as may be, new monthly figures for Dun’s and Bradstreet’s have
been made by averaging the index for July and August to get a new
figure for July, then averaging the figures for August and September
to get a new figure for August, and so on.
108
TH E M A K IN G AND U SING OF INDEX NTJMBEKS.
Coefficients of correlation among four American index numbers in the 54 months July ,
i914 , to December, 1918.
A . Coefficients of correlation computed from the m onthly ind ex numbers.
Coefficients of
correlation.
(1)
(2)
(3)
(4)
(5)
(6)
Bureau of Labor Statistics’ and W ar Industries Board’s series.....................
Bureau of Labor Statistics’ and Bradstreet’s series.............................................
Bureau of Labor Statistics’ and D u n ’s series...........................................................
W ar Industries Board’s and Bradstreet’s series.....................................................
W ar Industries Board’s and D u n ’s Series.................................................................
Bradstreet’s and D u n ’s series..........................................................................................
+ 0 . 997
-j- •988
-j- .9 9 4
+ •986
-j- . 995
+ . 991
B . Coefficients of correlation computed from the percentage change in prices from
one month to the next.
(1)
(2)
(3)
(4)
(5)
(6)
Bureau of Labor Statistics’ and W ar Industries Board’s series.................... ..+ 0 - 866
Bureau of Labor Statistics’ and Bradstreet’ s series............................................. ..-j- • 633
Bureau of Labor Statistics’ and D u n ’s series............................................................+ •801
W ar Industries Board’s and Bradstreet’s series........................................................-j- • 640
W ar Industries B oard’s and D u n ’s series................................................................. ..+ • 761
Bradstreet’s and D u n ’s series.......................................................................................... ..- f * 616
Taking both sets of coefficients into account, we find that the
Bureau of Labor Statistics’ index.number has the closest agreement
with the other three series. Then, in order, come the War Industries
Board’s series, Dun’s, and Bradstreet’s— which is the most divergent
of the four. But there is a better test of reliability. In view of its
very comprehensive list of commodities (1,366 in number) and its
use of class in addition to commodity weights, the War Industries
Board’s series is probably the nearest approximation to a true measure
ment of the changes in the wholesale price level during the war.
Accepting it as the standard, we may ask which of the three index
numbers currently published is in closest agreement with it. Once
more the answer is in favor of the bureau’s series, when one considers
the correlation either of the monthly index numbers themselves or of
the monthly percentages of change. Dun’s comes second and Brad
street’s again ranks lowest.
3. CRITICAL EVALUATION OF THE BUREAU OF LABOR STATISTICS’,
BRADSTREET’S, AND DUN’S INDEX NUMBERS.
A few additional remarks are called for on the relative merits of
the three general-purpose index numbers now regularly published in
the United States.
In the publication of actual prices, the Bureau of Labor Statistics’
and Bradstreet’s stand foremost. The contribution they have thus
made to the knowledge of prices possesses great and permanent value
over and above the value attaching to then* index numbers. For, it
is well to repeat, all efforts to improve index numbers, all investiga
tions into the causes and consequences of price fluctuations, and all
possibility of making our pecuniary institutions better instruments
of public welfare depend for their realization in large measure upon
the possession of systematic and long-sustained records of actual
prices. And much of this invaluable material would be lost if it
were not recorded month by month and year b y year.
Critical users of statistics justly feel greater confidence in figures
which they can test than in figures which they must accept upon
COMPARISON OF LEADING AMERICAN INDEX NUMBERS.
109
faith. Hence the compilers of index numbers who do not publish
their original quotations inevitably compromise somewhat the repu
tations of their series. They compromise these reputations still
further when they fail to explain in full just what commodities they
include, and just what methods of compilation they adopt. Brad
street’s index number suffers a bit in comparison because readers are
not told which 96 commodities out of the 106 for which prices are
published are included in the index number, and because the method
of reducing prices by the yard, the dozen, the bushel, the gallon, etc.,
to prices per pound is not fully explained. Dun’s index number is
more mysterious still, because neither the list of commodities nor the
weights applied to each commodity are disclosed.
The number of commodities now included in the three series is
given as follows by the compilers: Bureau of Labor Statistics’ 328,
Dun’s “ about 300,” Bradstreet’s 96. Provided the commodities are
equally well chosen, of course the longer the list of commodities
included the better claim has an index number to acceptance as a
measure of changes in the general level of commodity prices.
The preceding study of the relations among the leading American
index numbers was made in the winter of 1919-20, just before the
reat fall in prices be^an. Early in the course of this fall marked
iscrepancies appeared between the Bureau’s series and both the
commercial indexes. These discrepancies presently became wider
than any that had appeared in the preceding 30 years. By Decem
ber, 1920, Bradstreet’s index was 22.4 per cent lower than the Bureau’s
index and Dun’s was 10.9 per cent lower.25
t
25 The following table continues, by months, from January, 1919, to May, 1921, the index numbers of the
Bureau of Labor Statistics, Bradstreet, and Dun in the form given in Table 22:
Comparison of three American index numbers, by months, January, 1919, to May, 1921.
Year and month.
Bureau
of Labor
Statistics.
Brad
street.
1919:
January..
February..
March____
April.........
M ay..
June..
July.
August.......
September..
October___
November..
December..
202. 35
195.02
193. 03
193.11
197. 64
206. 92
217. 62
220. 84
218.15
220. 56
224. 22
226. 80
185.13
179. 74
179. 72
181.83
185.12
189. 85
195. 48
197. 38
195. 01
195.10
198. 71
202.34
202.89
197.24
200.90
203. 43
206.85
206.91
218.74
226.42
220.55
222.98
229.89
238.28
January..
February..
March___
April....... .
May.
June............
July.............
August.......
September..
October___
November..
December..
230. 68
233. 09
232.22
231. 87
227.19
219. 46
213. 60
205. 89
195. 16
182.30
163. 93
147.08
206. 08
208. 39
210.10
214. 34
216.09
214. 89
214. 95
209. 95
199. 69
191. 03
180. 45
168. 70
248. 59
249.19
253.97
268. 80
272.14
268.62
263.20
249.80
241.93
225.36
207.33
189. 44
January..
February.
March___
April.......
May.........
140. 05
135. 58
130. 02
124.17
119. 93
158. 09
151. 23
146. 53
140.26
136.80
177.92
167.41
161.83
153.72
151.09
110
TH E M A K IN G AND USING OF INDEX NUMBERS.
These wide discrepancies mean, not that the index numbers had
become suddenly worse, but that the diversity among price fluctua
tions had become greater, so that differences among index numbers
in respect to the number of commodities included and methods of
weighting produced wider differences in the results. In other words,
we have here the demonstration of a significant fact about price
fluctuations: The great drop of prices in 1920-2.1 was characterized
by much more irregularity in the promptness and degree of readjust
ment of different markets to the new situation than was the great rise
of prices in 1915-1919. Presumably these great irregularities will
prove to be a feature of the transition period only, and the three
index numbers will approach one another again as the readjustments
are gradually worked out in all markets.
With reference to weighting, Bradstreet’s index number takes low
rank, for the plan of reducing all quotations to prices per pound
grossly misrepresents the relative importance of many articles.
That figures made thus should give results in close agreement with the
Bureau of Labor Statistics’ series is really remarkable and proves
that if prices, the raw materials from which index numbers are made,
are accurate the particular method used in computing the index
number is of secondary importance. Dun’s system of weighting is
distinctly better than Bradstreet’s in theory. Whether the practice
is as good as the theory is doubtful, for anyone familiar with the
deficiencies of American statistics of consumption must wonder whence
the compilers derive their estimates of the quantities of “ about 300”
commodities “ annually consumed by each inhabitant.” Moreover,
what little is known concerning the actual weights is not unobjec
tionable. Fifty per cent of the total is too large a weight to allow
to foods in a wholesale-price series. Even in the great collection of
budgets of workingmen’s families made by the Commissioner of
Labor in 1901 the average expenditure for food was less than 45 per
cent of total family expenditure, and in 1918 it was found to be
only 38.2 per cent.26 The bureau’s practice of weighting wholesale
prices by the quantities of commodities that enter into trade is
preferable to weighting by consumption. Moreover, the bureau
publishes its weights, and shows each year the percentage which
each weighted price makes of the total for the group in which the
commodity is put, as well as of the total for all commodities.
26 Eighteenth Annual Report of the Commissioner of Labor, 1903, p. 66. The data represented 25,400
families and 124,108 persons, both natives and immigrants. Also the Monthly Labor Review of the Bureau
of Labor Statistics, August, 1919, p. 118. The data represented 12,096 white families in 92 industrial centers.
COMPARISON OF LEADING AMERICAN INDEX NUM BERS.
I ll
Index number.
Bradstreet’s...................................................
Bureau of Labor Statistics’ .......................
Dun’ s..............................................................
1893
1895
1897
1901
1903
1904
Lit
In the form of presenting results, Bradstreet’s set an admirable
example, which was wisely followed by Dun’s. Their sums' of
actual prices can readily be turned into relatives on any base desired,
and hence can be made to yield direct comparisons between any two
dates. The bureau’s series shares this advantage, since it too is made
by adding actual prices multiplied by weights; but it is presented in
a form more convenient for comparison than the other two series.
The relatives on the scale of 100, into which the bureau throws its
figures in the last step of compilation, are easier to use than the
awkward sums of dollars and cents which Dun’s Review and Brad
street’s publish.
It is interesting, finally, to test the reliability of the several index
numbers as “ business barometers.” Monthly figures would be better
for this purpose than our yearly averages, but since they are not
available for all three series in the 1890’s, we must do the best we can
with the rougher gauge. In 17 of the 26 years since 1892 (when
Bradstreet’s index m its present form begins), the three series agree
concerning the direction m which prices were moving; they differ in
nine years. In the following schedule these nine years are repre
sented by columns in which each index number is credited with 4-1
when its change accords with the character of the alteration in
business conditions, debited with —1 in case of disagreement, and
marked ± 0 when it recognizes no appreciable change in the price
level.27 The net scores made by casting up the plus and minus entries
indicate roughly the relative faithfulness with which these series have
reflected changes in business conditions in the past quarter of a
century.
-1
±0
-1
+1
+1
-1
+1
+ 1
-1
-1
-1
+1
+1
±0
-1
.
1905
1913
1914
Net
score.
+1
±0
±0
±0
+1
+1
+1
+ 1
-1
+4
+2
-4
Of the three indexes, Bradstreet’s makes the best showing. Pre
sumably the poor quality of Dun’s index as a business barometer is
due chiefly to the heavy weight (50 per cent) which it ascribes to
foods. For foods are largely farm crops whose prices in a given year
depend at least as much upon the weather as upon the condition of
business. The bureau’s series in this respect stands intermediate
between the two commercial series, giving a lighter weight to foods
than Dun’s and a heavier weight than Bradstreet’s. Probably that
is why it is a better business barometer than the one and not so good
as the other.
Of course this conclusion that Bradstreet’s index number is a
better business barometer than the bureau’s series does not invalidate
the preceding conclusion that the bureau’s series is the best measure
of changes in the general level of prices. For when farm crops are
given their due weight in an index number, it is not to be expected
that the index will always rise with business prosperity and decline
27 For a summary of the changes in business conditions during these years, see Business Cycles, by
Wesley C. Mitchell, p. 88.
112
TH E M A K IN G AND USING OF INDEX NUMBERS.
with business depression. In making a wholesale price index number
for use as a business barometer, indeed, one should exclude.altogether
commodities whose price fluctuations are determined largely by the
weather. We have no such series at present, and it is high time that
this lack should be supplied. But if some one does make a wholesale
rice index that is a nearly infallible business barometer, it will not
e as reliable a measure of changes in the general level of prices as
the present Bureau of Labor Statistics series.
E
VI.— CONCLUSIONS.
1. Variations in the level of wholesale prices from one year to the
next are capable of being measured by a close approximation to
accuracy, for these variations are highly concentrated about a central
tendency. There are two American chain index numbers which for
a quarter of a century never differ by more than 5 per cent, and differ
on the average by only 2 per cent, although they were compiled from
start to finish quite independently of each other, based upon dis
similar sets of price variations, constructed by unlike methods, and
covered a period of violent fluctuations.28
2. Variations in prices that have been cumulating through several
or many years show much less concentration about a central tend
ency than variations from one year to the next. Hence, index num
bers become less accurate the greater the time over which they are
extended. Nevertheless, the discrepancies observed between the two
series just referred to (Dun's and the Bureau of Labor Statistics’
new series of index numbers) do not reach 8 per cent in a period of
26 years, and average 3.4 per cent. The coefficient of correlation
between these two series in 1892-1914 is +0.992, a close approach to
+ 1.0, the expression of perfect agreement.
3. The choice of methods to be employed in making' an index
number should be guided by the purpose for which the results are to
be used. These purposes are so numerous and so diverse that it is
impossible to make a single series well adapted to them all. Prob
ably the time is near when certain uses will be so standardized that
several divergent types of index numbers will be regularly compiled
to serve the needs of various groups of users. Even now we have
special index numbers of the prices of foods, of farm products, of
metals, etc. To this list there might well be added a series especially
designed to throw changes in business conditions into high relief, and
assist in the bettering of business forecasts. Most of the currently
zi These figures are computed from Table 19, by turning the percentages by which each index number
rose or fell each year into relatives on the precedmg-year bass and computing the percentage differences
between the resulting indexes. The results for three series are as follows:
1
i Average difference.
Index numbers.
1893-1914. 1893-1918.
Bureau of Labor Statistics and Bradstreet’s ......................................................................
Bureau of Labor Statistics and Dun’s ............................................................................
Bradstreet’ s and Dun’s . .............................................................................................................
P er cent.
2. 2o
1.95
2. 92
P er cent.
2.82
2.00
3.15
CONCLUSIONS.
113
published index numbers, however, are what may be called generalpurpose series, which undertake to measure changes in the wholesale
price level at large.
.4. The best form for these general-purpose series is a weighted
aggregate of actual prices or a weighted geometric mean. The latter
is preferable for measuring average ratios of change in prices; the
former is preferable for measuring average change in the amount of
money required to buy goods.
5. The more commodities that can be included in such an index
number the better, provided that the system of weighting is sound.
Certainly, each of the following classes of commodities should be
represented, and represented as fully as is feasible: Raw mineral,
forest, animal, and farm products, and manufactured products in
various stages of elaboration, bought for family consumption and for
business use.
6. Probably the best weights to apply are the average physical
quantities of the commodities bought and. sold over a period of years
without reference to the number of times their ownership is changed.
In making an aggregate of actual prices these weights should be ap
plied directly to the quotations of each commodity in making up the
totals for the several groups that have been mentioned, and then, if
the necessary data can be secured the .totals for the several groups
should be weighted again in making up the grand totals for “ all
commodities.”
7. In presenting such an index number, it is well to publish the
aggregate actual prices, both for the several groups and for the
grand totals. But it is highly desirable to publish also relatives
made from these actual prices on a percentage scale, since comparisons
can be made more easily from such figures than from the aggregates
of actual prices, which are likely to run in awkward quantities.
Indeed, several sets of these relatives, computed on the basis of actual
prices at different times, can readily be provided for readers inter
ested in knowing how prices have changed with reference to recent
or to past years. Among the relatives of greatest significance is the
set which shows the annual percentage of rise or fall as compared
with prices in the preceding year. In such chain index numbers it
is usually possible to include some commodities for which quotations
are lacking in certain of the years covered by the whole investigation.
8. Chain index numbers are best made by the “ ideal” formula,
when the chief aim is to attain the greatest possible accuracy in
measuring fluctuations from one year to the next. But when the
annual percentages of rise or fall in prices made in this way are
forged into a continuous series, their errors cumulate and vitiate
comparisons between the earlier and the later years. Such series
are also faulty for some purposes in that one can not tell what part
of the net results is due to changes in prices and what part to changes
in the quantities used as weights. When the chief aim is to forge a
chain which will yield reliable comparisons between prices in any
two years it is best to use constant weights and make aggregates of
actual prices or geometric means of price fluctuations, the choice
turning once again upon the specific purpose in mind.
33226°— 21— Bull. 284-------8
114
TH E M A K IN G AND USING OF INDEX NUMBERS.
9.
While index numbers are a most convenient concentrated
extract of price variations, they are far from being a competent
representation of all the facts which they summarize. Most' “ con
sumers of statistics” lack the time to go back of the finished produets
to the data from which they are made. But the increase of knowl
edge concerning the causes and consequences of price variations
depends much more upon intensive study of the ultimate data than
upon the manipulation of averages or aggregates. Upon the exten
sion of knowledge in this field depend in turn large issues of public
welfare. Hence it is highly important to collect and to publish in
full the actual prices of as many commodities as possible, even
though some of the quotations may not now be available for use in
making an index number.
PART II. — INDEX NUMBERS OF WHOLESALE
PRICES IN THE UNITED STATES AND FOREIGN
COUNTRIES.
INTRODUCTION.
In the following pages the attempt has been made to present a
brief though systematic description of all of the more important
index numbers of wholesale prices that are known, either current or
past. In the case of current series, the information has been brought
up to the end of 1919, or the latest date for which complete data coufd
be obtained. For discontinued series, the facts as published in Bul
letin No* 173 of the Bureau of Labor Statistics are reprinted herewith.
Owing to the situation brought about by war, several hitherto im
portant series of index: numbers have been allowed to lapse. This
applies particularly to index numbers for Germany and Russia. In
other countries the work of compiling and publishing the information
has been interfered with to a greater or less extent by the exigencies
of war. Several series of index numbers not in existence when the
former bulletin was being prepared (in 1914) are described in the
present bulletin.
UNITED STATES.
INDEX NUMBERS OF THE UNITED STAGES BUREAU OF EABOR
SfAWHCS;
PUBLICATION.
An index number is published in connection with the reporta on
wholesale prices issued by the Bureau of Labor Statistics of the
United States Department of Labor at Washington. These reports
are in bulletin form and appear annually*1 Since June, 1918, the
index number has been published currently in the Monthly Labor
Review, the figures for a given month being shown in the issue for
the second month thereafter.
HISTORY.
The publication of this index number was begun in 1902. Prior to
that time the Department of Labor, now the Bureau of Labor Sta
tistics, had conducted an inquiry into the course of wholesale prices
from 1890 to 1899, the results of which were published in March,
19O0.2 The purpose of thi& inquiry was to continue, so far as prac
ticable, the investigation made for the Senate Committee of Finance
for the years 1840 to 1891 under the direction of Roland P. Faulkner,
statistician to the committee.8 In the report of the Department of
1 Owing to the situation brought about by the war, the bulletins for the years 1917 to 1919 were published
as one volume.
2 Bulletin No. 27 of the Department of Labor.
3 Report from the Committee of Finance of the United States Senate on wholesale prices, wages, and
transportation, Mar. 3, 1893, 52d Cong., 2d sess., Rept. No. 1394.
115
116
INDEX NUMBERS OF WHOLESALE PRICES.
Labor alluded to, the index numbers appearing in the Senate
Finance Committee’s report were brought down to 1899, important
changes with respect to the base period and the method of weighting
being adopted. In 1902, however, when the material for the new
report on wholesale prices was being assembled, it was found that
many articles included in the report of the Senate Committee on
Finance were either no longer manufactured or had ceased to be
important factors in the market. On the other hand, a number of
articles not shown in that report had become of such importance as
to render necessary their inclusion in the new report. These facts
necessitated'the computation of a new series of index numbers based
on the revised list of commodities. It was found, however, that
prices of such commodities could be obtained for a period dating back
to 1890, so that the new series of index numbers, as published in the
1902 report,4 covered the 12 years from 1890 to 1901, inclusive.
This series was continued in subsequent wholesale-price reports to
1914, such reports being contained in the March issue of the bimonthly
bulletin of the bureau for the years from 1902 to 1911, inclusive, and
published in separate form for 1912 and subsequent years.
Beginning with the report for 1914,5 the number of commodities
included in the exhibit of wholesale prices was greatly increased and,
in addition, a radical change in the method of calculating the index:
numbers from the money prices quoted was introduced. The new
method is described in, detail in the appendix to the 1914 bulletin,6
its predominant feature being the use of quantities denoting the total
value in exchange of all commodities in each group in a given year
or month, as compared with a similar quantity in the period selected
as the base. This method of constructing group and general index
numbers from the aggregate value of commodities exchanged year by
year, thus producing a weighted index:, superseded the former method
of averaging the relative prices of individual commodities to obtain
unweighted group and general index numbers. This method has
been continued in subsequent reports.
SOURCE OF QUOTATIONS.
The commodities included in the reports have been selected, not
only with regard to their representative character, but also with
regard to their availability in the future in the continuation of the
price record. Standard trade journals, reports of boards of trade,
chambers of commerce, produce exchanges, and leading manufac
turers or their selling agents are the usual sources from which the
price quotations are obtained. It has been the aim to secure the
quotations for the various commodities from their primary markets.
Thus, for most grains, live stock, meats, etc., Chicago prices are
quoted; for iron and steel products, Pittsburgh prices; for cotton
and rice, New Orleans prices, etc. The prices for textiles are those
prevailing in the general distributing markets, such as Boston, New
York, and Philadelphia; and wThere no market is mentioned it should
be understood that the prices are for the general market.
« Bulletin No. 39, of the Department of Labor, March, 1902.
5 Bulletin No. 181, of the Bureau of Labor Statistics, pp. 252-256.
e Idem.
U NITED STATES---- BUREAU OF LABOR STATISTICS.
117
BASE PERIOD.
In the reports issued during the period from 1902 to 1913, inclu
sive, the standard used for measuring price changes was the average
price for the 10 years 1890 to 1899. This period, which embraced
the first 10 years of the bureau’s record of prices, was selected because
it was believed that an average price for a number of years would
better reflect average conditions and form a broader and more satis
factory base than would the price for any single year. In the cases
of a few articles for which prices for the entire 10-year period could
not be obtained, the average for such years prior to 1899 as were
covered by the data was chosen as the base.
In the 1914 report (Bulletin No. 181), issued in October, 1915, the
base period was shifted from the decade 1890-1899 to the last com
pleted year, 1914. This change was made for the purpose, first, of
utilizing the latest and most trustworthy price quotations as the
base from which price fluctuations were to be measured, and, second,
to permit of the addition of new articles to those formerly included
in the index number. For practically all articles which it was de
sired to add to the index in 1914 no prices were obtainable for the
years 1890-1899.
In the reports for 1915 (Bulletin No. 200) and 1916 (Bulletin No.
226) the plan was continued of using the last completed year as the
base. The abnormal conditions brought about by the war in Europe,
however, made it advisable to provide a prewar standard for measur
ing price changes, and in the preparation of the bulletin covering the
years 1917 to 1919 (Bulletin No. 269) the base period was shifted to
the year 1913. This plan conformed to the one adopted by the
bureau in its reports on retail prices and on union wages and hours
of labor.
PRICES: H OW SHOW N AND COMPUTED.
For commodities of great importance, such as cattle, wheat, butter,
eggs, milk, cotton, wool, pig iron, and lumber, more than one price
series has been included in the index. In no case, however, is an
article of a particular description represented by more than one series
of quotations for the same market. Weekly prices have been secured
for all articles subject to frequent fluctuations in price. In a large
number of instances, particularly since the beginning of 1918, it has
been possible to obtain average monthly prices based on daily fluc
tuations. For a few commodities whose prices are quite stable, as
salt and plate glass, only first of the month prices are used.
The average price for the year is obtained by dividing the sum of
the quotations for a given commodity by the number of quotations.
For example, the sum of the 52 Tuesday prices of No. 1 northern
spring wheat at Minneapolis for 1919 was $133.43. This total,
divided by 52, gives $2.5660 as the average price per bushel for the
year. When a range of prices is shown for a particular date, the
mean of the high and low prices is found and this figure is used in
computing the yearly average as above described.
Net cash prices are used for all articles whose list prices are subject
to large and varying discounts. In the case of a few articles, the
prices of which are subject to a small discount for cash, no deduction
has been made.
118
INDEX NUM BERS OF WHOLESALE PRIC£S.y
NUMBER AND CLASS OF COMMODITIES.
Considerable variations have perforce occurred in the number of
commodities included in the various reports on wholesale prices
compiled by the Bureau. In earlier years the number varied from
250 to 260, increasing to 340 in 1914. In the detailed record of
prices from 1890 to 1916 published in Bulletin No. 226 (issued in
December, 1917) 342 series of quotations (including cases of sub
stitution) were presented for the entire period or some part of thje
period. Of this number, 296 were weighted and used in computing
the index number. In the report for 1917 to 1919 (Bulletin No. 269)
the number of price series included in the weighted index was in
creased to 328, with 48 additional series for which satisfactory
weighting factors could not be obtained. The number of commodities
or price series in the weighted index classified into groups is as follows:
Farm products, 32; foodr etc., 91; cloths and clothing, 77; fuel and
lighting. 21; metals and metal products, 25; lumber and building
materials, 30; chemicals and drugs, 18; housefurnishing goods, 13;
miscellaneous, 21; all commodities, 328.
DESCRIPTION AND GROUPING OF COMMODITIES.
The following list, compiled from the report for 1917 to 1919 (pp.
175 to 180), shows the grouping and description of the articles, those
not included in the weighted index number being so noted.
Group I .— Farm products.
Cotton* middMagv upland:
N ew Orleans.
N ew Y ork.
F laxseed, No. 1, cash.
G rain :
Barley, fair to good, m alting.
Com, cash—
Contract grades.
No. 3* m ixed.
Oats, contract grades, cash.
R ye , N o. 2, cash.
W heat, cash—
Chicago, N o. 1 northern, sprtag.
Chicago,. No. 2 red winter. ^
Kansas C ity, No. 2 hard w inter.
M inneapolis, No. 1 northern spring.
Portland, Oreg., foluestem.
H ay:
A lfalfa, N o . 1.
T im o th y, No. 1.
H id es:
Calfskins, No. 1.
GoatsM os, Brazilian.
Greett,: salted, packers’—
H eavy,, native steers.
H e a v y , Texas steers.
Hope:
New Y ork State, prim e to choice.
Pacific coast.
L ive stock (for fo o d ):
Cattle, steers—
Choice ta prime h eavy beeves'.
Good to choice, corn fed.
UNITED STATES— BUBEAU OF LABOB STATISTICS.
Hogs:
H eav y.
Light.
S h eep :
Ewes, fed, poor to best.
Lambs, good to prime.
W ethers, fed, common to best.
Peanuts, No. 1 grade.
Poultry, live fow ls:
Chicago.
New York, choice.
Tobacco, burley, dark red, good leaf.
Group I I . —Food , etc.
Beans, m edium , choice.
B read:
Crackers.
Oysters.7
Soda.7
Loaf (before baking, per 16 o z .)—
Chicago.7
Cincinnati.7
New Orleans.7
New Y o r k .7
San Francisco.7
Butter:
Boston—
Creamery, extra.
Creamery, firsts.
Creamery, seconds.
Chicago—
Creamery, extra.
Creamery, extra firsts.
Creamery, firsts.
Cincinnati—
Creamery, extra.
Creamery, centralized firsts.
Creamery, centralized seconds.
New Orleans.
Creamery, fancy.
Creamery, choice.
New York—
Creamery, extra.
Creamery, firsts.
Creamery, seconds.
Philadelphia—
Creamery, extra.
Creamery, extra firsts.
Creamery, firsts.
St. Louis—
Creamery, extra.
San Francisco—
Creamery, extra.
Creamery, primary firsts.
Canned goods:
C om , New Y ork , standard.
Peas, State and western, No. 5 sieve.
Tomatoes, standard New Jersey, No. 3.
Cheese:
Chicago, whole m ilk, American twins.
New Y ork, whole m ilk, colored, average, fancy.
San Francisco, California flats, fancy.
Coffee, R io, No. 7.
7Not weighted.
119
120
INDEX NUMBERS OF WHOLESALE PRICES.
Eggs:
Boston, firsts, western.
Chicago, firsts, fresh.
Cincinnati, firsts, fresh.
New Orleans, candled, western.
New Y ork, firsts, fresh gathered.
Philadelphia, extra firsts, western.
San Francisco, fresh, selected, pu llets’ .
F ish :
Cod, large, shore (pickled, cured).
Herring, pickled, domestic, sp lit, large, N o. 1.
Mackerel, salt, large, No. 3s.
Salmon, canned, A laska, red.
F lour:
B uckw heat.
R y e , pure, m edium straight.
W heat—
Kansas City, winter patents. 8
Kansas City, winter straights. 8
M inneapolis, standard p aten ts.8
Minneapolis, second patents. 8
Portland, patents. 8
St. L ouis, first p a te n ts.8
St. Louis, second patents. 8
Toledo, patents. 8
T oledo, straights. 8
F ruit:
A pples—
Evaporated, choice.
Fresh, Baldw in.
Bananas, Jamaica, 8s.
Currants, uncleaned, in barrels.
Lemons, California.
Oranges, California.
Prunes, California, 60s to 70s, in 25-pound boxes.
Raisins, California, coast seeded (bu lk).
Glucose, 4 2 ° m ixing.
Lard, prim e, contract.
M eal, corn:
W h ite, Terre H aute.
W h ite, table, Philadelphia.
M eat:
Bacon—
Rough sides.
Short clear sides.
Beef, fresh—
Chicago:
Carcass, good native steers*
Steer, loin ends (hips). 7
Steer, ribs, No. 2. 7
Steer, rounds, No. 2. 7
N ew Y ork:
N ative sides.
Loins, No. 2, city. 7
R ibs, No. 2, city. 7
Rounds, No. 2, city. 7
Beef, salt, extra mess.
H am s, smoked, loose.
Lam b, dressed, round.
Mutton, dressed.
Pork, fresh—
Chicago, loins. 7
N ew Y ork, loins, western. 7
7Not weighted.
8 Only flour conforming to United States Food Administration standard was quoted in 1918.
UNITED STATES— BUREAU OF LABOR STATISTICS.
Meat— Continued.
Pork, salt, mess, old to new.
P oultry, dressed fowls—
Chicago, iced.
N ew York, western, dry packed.
Veal, city dressed, good to prime.
M ilk, fresh:
Chicago (vicin ity).
New Y ork (vicin ity).
San Francisco (vicin ity).
Molasses, N ew Orleans, open kettle.
Oleomargarine, standard quality, uncolored.
Oleo oil, extra.
O live oil, Spanish.
R ice:
B lue Rose, head.
Honduras, head.
Salt, Am erican, m edium.
Soda, bicarbonate of, A m erica n .7
Spices: Pepper, black, Singapore.
Starch, corn, for culinary purposes.7
Sugar:
96° centrifugal.
Granulated, in barrels.
T allow , packers* prime.
Tea, Formosa, fine.
Vegetables, fresh:
Onions.
Potatoes, white, range from good to choice.
Vinegar, cider, 40-grain, in barrels.
Group I I I. —Cloths and clothing.
Boots and shoes:
Children’s, gun m etal, button.
Little boys’ , gun m etal, blucher.
M en’s—
Gun metal, Goodyear welt, blucher.
Gun m etal, Goodyear welt, button.
Split seamless, Creedmore.
V ici calf, blucher.
V ici kid, Goodyear welt.
Misses, vici kid, patent tip, button.
W om en ’s—
Gun metal, Goodyear welt, button.
Gun metal, M cK a y , sewed, button.
Patent leather, pum p, M cK a y , sewed.
Y o u th ’s, gun metal, blucher.
Carpets:
A xm inister, Low ell, 3 -4 .
Brussels, 5-fram e, Biglow.
W ilton, 5-fram e, Biglow.
Cotton goods:
Blankets, 2 pounds to the pair.
Calico, American standard prin ts, 7 yards to the pound, in gray.7
Denim s, Massachusetts, 2.30 yards to the pound.
Drillings—
Brown, Pepperell, 2.85 yards to the pound, 29-in ch .
Massachusetts, D standard, 30-in ch .
Flannels—
Colored, 2 f yards to the pound.
Unbleached, 3| yards to the pound.
Gingham — ■
Amoskeag, 6.37 yards to the pound.
Lancaster, 6.50 yards to the pound.
7 Not weighted.
121
122
INDEX NUM BEBS OF WHOLESALE PRICES.
Cotton goods— Continued.
Hoisery—
M en ’s half hose, combed yarn, fast black.
W om en ’s—
F u ll fashioned, combed peeler yarn, double sole.
Seamless, combed yarn, double sole, 220-needles,
Percale, Scout, 3 6-in ch , 5.35 yards to th e p ound.7
Print cloths, 2 7-in ch , 7.60 yards to the pound.
Sheeting—
Bleached—
P epperell, 1 0 -4 .
W am sutta, 10-4.
Brown—
Indian H ead, 4 -4 , 2.85 yards to the pound.
Pepperell, R . 4 -4 , 3.75 yards to th e pound.
W are Shoals, L . L ., 4 -4 , 4 yards to the pound.
Shirting, bleached m uslin, 4 -4 .
F ruit of the Loom.
Lonsdale.
Rough Rider.
W am sutta,
Thread, 6-cord, 200-yard spools, J. & P. Coats.
T icking, Amoskeag, A . C. A ., 2.05 yards to the pound.
Underwear—
M en ’ s shirts and drawers.
W om en ’ s union suits.
Y arn, carded, w hite mulespun, northern, cones— ■
10- 1 .
22-1.
Y a m , twisted, carded, ordinary, for w eaving— •
20- 2 .
4 0-2.
Leather:
Calf, chrome, dull or bright finish, B grade.
Glazed kid, black, top grades, from Brazilian skins.
Harness, California oak, N o. 1.
Side, black, chrome, tanned, B grade.
Sole— H em lock, m iddle, N o. 1.
Oak, scoured backs, heavy.
L in en shoe thread, 10s, Barbour.
S ilk, raw:
Japan—
K ansai, No. 1.
Special, extra, extra.
W oolen goods:
Blankets, 5 pounds to the pair.
Flannel, w hite, 4 -4 , Ballard V ale, N o. 3.
Hosiery—
M en ’s seamless half hose— •
Cashmere.
W o o l.7
Overcoating, soft-faced, black, plain tw ill, 24-ounce.
Suitings— ■
Clay worsted, diagonal— •
12-ounce.
16-ounce.
M iddlesex, w ool-dyed, blue.
Serge, 11-ounce.
Trousering, worsted, w ith silk decoration, 11 to 11J ounces to the yard.
Underwear, merino— ■
M en ’ s shirts and drawers, 50 per cont wool.
Union suits, 33J per cent wool.
7Not weighted.
UNITED STATES---- BUREAU OF LABOR STATISTICS.
W oolen goods— Continued.
W om en ’s dress goods:
A ll wool—
Broadcloth, 53-54 inches.
French serge, 35-inch.
Storm serge, double-warp, 50-inch,
Cotton warp—
Cashmere, H am ilton M ills,
Poplar cloth, 36-inch.
Sicilian cloth, 50-inch.
W ool, Ohio, scoured, fleece— •
Fine clothing.
F ine delaine.
Half-blood.
M edium (£ and f grades).
Yarn—
2 -3 2s, crossbred stock.
2-40s, half blood.
2-50s, fine domestic.
Group I V .—•Fuel and lighting.
Alcohol, denatured, 180 proof.
Coal:
Anthracite— ■
Broken.
Chestnut.
Egg.
Stove.
Bitum inous— Chicago—
M ine run.
Prepared sizes.
Screenings.
Cincinnati, run-of-m ine.
Pittsburgh, prepared sizes.
S t. Louis, run-of-m ine.
S em ibitum inous—
Cincinnati, N ew R iver smokeless, nmrof-mine.
Norfolk, Pocahontas.
Coke, Connelsville, furnace.
Gasoline, motor.
M atches, average of Safe H om e, B ird ’s E y e , and Searchlight brands.
Petroleum :
Crude—
California.
K ansas-Oklahom a.
P ennsylvania.
R efined— •
Standard w hite 110° fire test, for export.
W ater-white 150° fire test (jobbing lots).
Group V.— Metals and metal products.
Augers, regular, 1-in ch .7
Bar iron:
B est refined, P hiladelphia.
Common, from m ill, Pittsburgh.
B u tts, loose-pin, wrought-steel, 3^ b y 3| in ch e s/
Chisels, regular, socket firmer, l-in c li.7
Copper:
Ingot, electrolytic.
Sheet, hot-rolled (base sizes).7
W ire, bare, No. 8.
Door knobs, steel, bronze-plated.7
F iles, 8-inch, m ill bastard.7
Ham m ers, M aydole, No. 1|.7
Not weighted.
123
124
INDEX NUMBERS OF WHOLESALE PRICES.
Iron ore, M esabi, Bessemer.
L ead :
Pig, desilverized.
Pipe.
Locks, common, mortise, knob-lock, 3J in ch .7
Nails, 8-penn y, fence and common.
C u t?
W ire.
P ig iron:
Basic.
Bessemer.
Foundry—
N o.
northern
N o. 2, southern.
P ipe, cast-iron, 6-inch.
Planes, jack planes.7
Q uicksilver.7
Saws, Disston:
Cross-cut, N o. 2, 6-in ch .7
H and, N o. 8, skewback, 26-inch.7
Shovels, Allies, No. 2 .7
Silver, bar, fine.
Steel:
B illets, Bessemer.
Plates, tank, J-inch w ide.
Rails, standard—
Bessemer.
Open-hearth.
Sheets, bo x annealed, N o. 27.7
Structural.
T in :
Pig.
Plate, coke.
Trowels, Johnson’ s, brick, lO^-inchee.7
V ises, solid box, 50-pound.7
W ire:
Barbed, galvanized.
Plain, annealed, Nos. 0 to 9.
W ood screws, 1-inch, N o. 10, flathead.7
Z in c:
Sheet.
Spelter (pig zinc) western.
Group VI.—Lumber and building materials .
Brick, common:
Chicago, run-of-kiln, salmon.
Cincinnati, red, building.
New Y ork , red, domestic, building.
Cem ent:
Portland, domestic.
Doors, western white p in e.7
Glass:
Plate, polished, glazing—
Area 3 to 5 square feet.
Area 5 to 10 square feet.
W indow , American single, 25-inch—
A grade.
B grade.
Lath, eastern spruce, IJ-inch slab.
Lim e, eastern, common.
Lum ber:
Douglas fir—
No. 1.
No. 2 and better
H em lock.
Maple.
f
Not weighted.
UNITED STATES— BTJBEAU 01- LABOR STATISTICS.
L um ber— Continued.
Oak, w M te—
Plain.
Quartered.
Pine— •
W hite, boards, rough, No. 2, barn.
W hite, boards, rough, uppers.
Y ellow , flooring.
Y ellow , siding.
Poplar, yellow.
Spruce, eastern.
Paint materials:
Lead, carbonate of (white lead), American, in oil.
lin s e e d oil, raw.
Turpentine, spirits of.
Zinc, oxide of (zinc w hite), American, extra drv.
Putty.
Rosin, common to good, strained.
Shingles, 16 inches long:
Cypress.
Red cedar.
rf'ar, p in e.7
Group V II.—Drugs and chemicals♦
A cid :
A cetic, 28°.
Muriatic, 2 0°.
Nitric, 4 2°.
Sulphuric, 66°.
A lcohol:
Grain, 190 proof, U . S. P.
W ood, refined, 95 per cent.
A lu m , lum p.
Am m onia, anhydrous.
Borax, crystals and granulated, sacks.
Copper sulphate, 99 per cent crystals (blue vitriol),
Glycerine, refined, chem ically pure.
Opium, natural, in cases.
Quinine, manufacturers’ quotation.
Soda:
Carbonate of (sal soda).
Caustic, 76 and 78 per cent, solid.
Nitrate of (Chile saltpeter), 95 per cent.
Soda ash, light, 58 per cent.
Sulphur (brimstone), stick, crude.
Group V III. —Ilouse-furnishing goods.
Earthenware:
Plates, 7-inch, white granite.
Teacups and saucers, white, granite.
Furniture:
Bedroom sets, 3 pieces.
Chairs— •
Bedroom, rocker, oak.
K itchen, hardwood.
Tables, kitchen, 2 b y 3J feet, with drawer.
Glassware, common:
Nappies, 4-in ch .
Pitchers, ^-gallon.
Tumblers, table', |-pint.
Table cutlery:
Carvers, stag handles.
K nives and forks, cocobolo handles, metal bolsters,
Wooden ware, oak-grained:
Pails, 3-hoop.
Tubs.
125
126
raD E X NUM BERS OF WMO&ESALE PRICES.
Group I X
.—■Miscellaneous.
Beer, western, liglit or dark.7
Bran.
Cottonseed meal, prime.
Cottonseed oil, prime, summer yellow.
Jute, raw.
Lubricating oil, paraffine.
M alt, standard keg beer.
Paper:
New , wood, roll, annual contracts.
W rapping, manila, No. 1 jute.
Phosphate rock, Florida land pebble, 68 per ceat.
Hope, pure manila, f-in ch (base size) ana larger.
R ubber, Para, island, fine.
Soap, laundry:
100 bars to a box of 6 8 f pounds.
100 bars to a box of 75 pounds.
Starch, laundry.
Tobacco:
Plug, Climax.
Smoking, granulated, B lackw ell’s B u ll Durham.
W h isky:
Bourbon—
4 years in bond, 100 proof, in barrels, K en tuck y.
Bottled in bond, K en tu ck y.
Proof spirits.1
R ye—
4 years in bond, 100 proof, in barrels, Pennsylvania.
Bottled in bond, Pennsylvania.
Wood pulp, sulphate, domestic, unbleached.
SUBSTITUTIONS AND ADDITIONS.
Since the issuance of the first wholesale price report covering the
years 1890 to 1901, a number of changes h4ve been madfe necessary
m the list of articles included. Certain articles n& longer com^
mercially important, or for which satisfactory price quotations could
no longer be obtained, have had to be discontinued and other articles
substituted therefor. In making such substitutions articles were
supplied corresponding as closely as possible to feose which were
previously used.
Additions to the number of articles included in the compilation
have been made from time to time as circumstances demanded.
For 1890-the number of articles in the weighted index (as shown in
Bulletin No. 181, p. 160) was 192. In 1914 the number had increased
to 297. In 1915 the number was 30#. In 1916 it was reduced to
296, owing to the temporary dropping of 4 lines of children’s shoes.
In the report for 1917 to 1919, the 4 series of children’s shoes were
restored and 28 other series added, bringing the total te 328.
INTERPOLATION.
In a few instances during the years 1914 to 1917 interpolated
figures were used in calculating the index numbers for months in
which actual price quotations could not be obtained. For example,
the closing of the New York and New Orleans cotton exchanges
following the outbreak of war m 19*14 made it impossible to secure
satisfactory prices for this staple during a part of tftat year. There
fore, in computing the monthly index numbers for the farm proctucts
7Not weighted.
UNITED STATES-----BUREAU OF LABOR STATISTICS.
127
group and for all commodities combined, the plan was followed of
repeating for months in which quotations were lacking the figure for
the last month prior to the closing of the exchange. A similar
plan was followed in the cases of other articles for which prices were
unobtainable in one or more months of the year. In computing
yearly averages, however, the interpolated figures were not used,
the averages in such cases being made from the actual prices avail
able. In no case was a price interpolated for an entire year, or for
more than a few months of the year.
Beginning with 1918, the plan was adopted of computing and
publishing the index number from month to month, instead o f after
the close of the year as before. Under this plan of month-lo-month
comparison of weighted price aggregates it is the practice in cases
where prices from correspondents are not received in time for use
in computing the monthly index number, to interpolate the latest
reported prices and issue the index number as a “ preliminary”
figure, revising the computation when the correct figures are received.
WEIGHTING.
Prior to 1914 weighting in its technical sense was not attempted
in computing the index numbers shown in the wholesale price reports
of the Bureau. Instead of this, the plan was followed of using simply
a large number of representative staple articles, selected in such a
manner as to make them a large to extent weight themselves.^ The
actual money prices o f the different commodities were converted into
relatives o f the average price in the base period 1890 to 1899, taken
as 100, and these relatives were then combined into index numbers
o f groups o f commodities and into a general index number of all
commodities taken together. The simple arithmetic average of the
relative prices in each group and in all commodities combined con
stituted the index numbers under this method.
Beginning with the 1914 report, the plan was adopted of con
structing group and general index numbers from the aggregate
values o f commodities exchanged year by year. These aggregate
values are obtained by multiplying the price in any given year or
month by the physical quantity of the article placed on the market
in a given year, 1909. The latter year was selected because in 1914
it represented the latest year in which a Federal census of agriculture
and manufactures had been taken, and consequently it was the most
recent period for which complete information was available.
To obtain the quantities ol the various commodities marketed in
1909, recourse was had wherever possible to sources of an official
character— particularly to reports of the Census Bureau, the Depart
ment o f Agriculture, the Bureau of Foreign and Domestic Commerce,
the Geological Survey, the Forest Service, and the Commissioner of
Internal Revenue. In instances where satisfactory information from
Government reports was lacking, data believed to be authoritative
were obtained from nonofficial sources, such as the American Iron
and Steel Institute, the National Association of Wool Manufacturers,
produce exchanges, boards of trade, chambers of commerce, leading
trade journals, etc.
9 See Bulletin No. 181, Wholesale Prices 1890 to 1914, pp. 239-252.
128
INDEX NUMBERS OF WHOLESALE PRICES.
In the case of such articles as barley, corn, oats, hay, live poultry,
apples, eggs, and milk, of which a considerable part of the quantity
produced is consumed at home by the farmer, care was taken to
include only that portion actually marketed. A similar plan was
followed with regard to cotton and worsted yarns, coal, copper wire,
pig iron, steel billets, and other manufactures used to a large extent
m establishments where produced. Of cattle, hogs, and sheep, only
the quantities consumed in the slaughtering and meat-packing
industry were included, while the figures for the several kinds of
meat products were restricted to the output of this industry, with
the addition of the relatively small amount of imports.
For convenience in computation and to avoid the appearance of
overrefinement of accuracy, the physical quantities used as weights
for the prices of the different commodities have been expressed to
the nearest thousands only. For example, it was found that the
quantity of barley sold by growers in 1909, plus the imports, amounted
to 75,300,538 bushels. This was written as 75,301. Similarly, the
products representing value in exchange were expressed in thousands
of dollars. Thus, the average price of barley in 1914 was $0.6151
per bushel. This multiplied by 75,301 (000 omitted) produces
$46,318 (000 omitted) as the approximate value of barley marketed
in 1914.
To illustrate the method followed in determining the weights for
the different commodities, wheat is taken as an example. The
Census Bureau reported the production of wheat in the United States
in 1909 as 683,379,259 bushels. To this was added the imports
during the fiscal year ending June 30, 1909, of 41,031 bushels, and
from the sum was subtracted the reexports, amounting to 3,762
bushels, leaving 683,416,528 bushels as the quantity of wheat pre
sumably marketed in 1909. From data furnished by the Depart
ment of Agriculture it was found that 269,843,000 bushels, or approx
imately 39."5 per cent, of this was spring wheat and 413,573,000
bushels, or 60.5 per cent, was winter wheat. In preparing the 1914
report, prices for both 1913 and 1914 were collected for spring wheat
in Minneapolis and Chicago and for winter wheat in Kansas City,
Chicago, and Portland, Oreg. B y reference to reports of boards of
trade, supplemented by information from other sources, the relative
importance of each of these cities as a wheat market in 1909 was
established with reasonable certainty. The total quantity of spring
wheat marketed in the United States in 1909 was accordingly ap
portioned between Chicago and Minneapolis, and the total quantity
of winter wheat was apportioned among Chicago, Kansas City, and
Portland in proportion to their relative importance. As only one
series of price quotations for wheat had been included in the Bureau’s
compilation for years prior to 1913, the prices for such years were
weighted by the combined figures for spring and winter wheat stated
above— 683,416,000 bushels.
The physical quantities used as “ weights” for the different series
of price quotations in calculating the group and general index num
bers are based, as has been stated, on the most careful estimates from
official and nonofficial sources, and represent with approximate
accuracy the quantities of the different commodities actually mar
keted in the census year 1909. These quantities, representing the
UNITED STATES---- BUREAU OF LABOR STATISTICS.
129
amount of each commodity exchanged in 1909, have been multiplied
by the average prices for each month and year to obtain the value of
the same quantity of the commodity at the price prevailing in the
different months and years. The values of individual commodities
were summed by groups and for all commodities to get the cost of
the same bill of goods in the months and years for which an index
number was to be calculated.
If, during any period of time under consideration, there were no
changes in the list of commodities included in the index numbers,
the percentage changes in the cost of the different groups of com
modities would be accurately measured by dividing the aggregate
for the base period into the corresponding aggregate for each month
or year to be covered. In a number of years, however, the bureau's
reports of wholesale prices have contained additions to the list of
articles previously included, while numerous substitutions of one
article for another at a different price have been made from time
to time as circumstances demanded. Therefore, in carrying the
index numbers over the period since 1890, a method had to be adopted
that would allow for variations in the number and kind of commodi
ties from year to year. This method, which is identical in principle
with that employed by the bureau in its reports on retail prices and
on wages, consists in computing two separate aggregates for any year
or month in which an addition or a substitution occurs— the first
aggregate being made from the list of articles before making additions
or substitutions, and the second aggregate from the revised list of
articles. In this way comparison between any two consecutive
years or months is based on aggregates made up of identical com
modities only. Table 1, relative to the farm products group, which
was published on page 255 of Bulletin No. 181, will serve to illustrate
the method employed.
T a b l e 1 . — IN D E X
NUMBERS BASED ON AGG REGATE VALU ES IN EXCH A N G E OF FARM
PRODUCTS IN 1912, 1913, AN D 1914.
Value in exchange (000 omitted) in—
Commodity.
1912
1913-A
Cotton, upland, middling:
New Orleans......................................................................
$692,016
New York..........................................................................
$622,285
Flaxseed, No. 1........................................................................
38,581
27,123
Grain:
Barley—
52,266
Choice to fancy..........................................................
76,717
Fair to good, malting...............................................
Corn, cash—
288,032
Contract grades.......................................................
315,863
No. 3, mixed...............................................................
100,662
Oats, cash..........................................................................
117,323
18,781
Rye, No. 2, cash...............................................................
23,567
Wheat, cash.......................................................................
716,903
651,500
Chicago, No. 1 northern spring..............................
Chicago, No. 2 red winter........................................
Kansas City, No. 2 hard winter.............................
Minneapolis, No. 1 northern spring......................
Portland, Oreg., bluestem......................................
Hay, timothy, No. 1...............................................................
218,106
171,284
Hides:
Calfskin, No. 1 ..................................................................
Green, salted, packers’—
162,315
169,601
Heavy, native steers................................................ I
Heavy, Texas steers................................................
33226°— 21— Bull. 284------- 9
1913-B
1914
$458,026
230,636
27,123
$406,093
218,194
30,664
47,086
46,318
170,165
115,213
100,662
18, 781
188,716
128,151
112,260
22,657
24,639
122,372
217,548
212,137
38,417
171,284
28,015
124,705
233,032
243,612
41, 717
168,202
24,775
27,573
113,067
55,611
120,691
58,870
130
T
able
INDEX NUM BERS OF WHOLESALE PRICES.
1 .—I N D E X NUM BERS BASED ON AG G R E G A T E V A L U E S IN E X C H A N G E OF F A R M
PRODUCTS IN 1912, 1913, A N D 1914—Concluded.
Value in exchange (000 omitted) in—
Commodity.
1912
Hops:
New York State, prime to choice.................................
Pacific coast.......................................................................
Live stock (for food);
Cattle, steers—
Choice to prime.........................................................
Good to choice...........................................................
Hogs—
Heavy..........................................................................
Light............................................................................
Sheep—
Ewes, fed,. poor to best............................................
Lambs, good to prime..............................................
Wethers, fed, common to best...............................
Wethers, western......................................................
Peanuts, No. 1 grade..............................................................
Poultry, live fowls:
Chicago...............................................................................
New York, choice.............................................................
Tobacco, Burley,, dark red...................................................
Total.................................................... ..........................
Index number...................................................................
J913-A
1913-B
mu
$19,712
$12,803
$5,600
4,128
14,971
6,143
387,900
696,542
370,090
705,230
370,090
705,230
400,066
749,290
145,148
432,931
159,863
484,665
159,863
484,665
159,775
480,509
9,448
46,960
10,250
50,789
10,780
62,736
6,149
11,600
65, 319
6,705
68,429
125,753
77,138
145,274
4,224,483
98
4,187,367
15,470
17,518
35,505
38,569
145,274
33,431
38,016
161,25Q
4,191,601
97
4,334,063
100
Starting with the 1914 aggregate ($4,334,063) as the base (100),
$4,191, divided by $4,334,063 gives 96.7 as the index number for
1913. This is rounded off to 97, when published. The ratio of the
aggregate in 1912 to the aggregate in 1913, based on quotations of
identical articles, is $4,224,483 to $4,187,367 or 100.9 to 100. The
index number for 1912 is obtained by multiplying the index for 1912
on the 1913 base (100.9) by the index for 1913 on the 1914 base (96.7)
which gives 97.6, or 98, as the index number for 1912 on the 1914
base. B y continuing this operation the index numbers for preceding
years are obtained. Index numbers for the 8 remaining groups of
commodities and for all commodities combined are found in similar
manner.
TESTING.
Several comparisons have been made of the Bureau’s weighted index
number with the index numbers compiled by other agencies. Since
March, 1918, the Monthly Labor Review has carried quarterly a table
showing in parallel columns the index numbers of the Bureau of Labor
Statistics, several series for foreign countries, all being reduced to
percentages of the 1913 index number for purpose of direct com- parison. The Federal Reserve Bulletin, published monthly by the
Federal Reserve Board, has contained since October, 1918, a com
parison of the Bureau’s index with several series constructed by that
board and designated by groups as farm products, animal products,
forest products, mineral products, total raw materials, p rod u ced
goods, and consumers’ goods, respectively. The year 1913 consti
tutes the base period in this comparison. In the summary to a
History of Prices During the War, prepared by Wesley C. Mitchell
and issued as War Industries Board Bulletin No. 1, by tKe War In
dustries Board in 1919, a comparison is made of the four series of
index numbers compiled by the Price Section of the War Industries
Board, the Bureau of Labor Statistics, Bradstreet’s, and Dun’s for
UNITED STATES— BUREAU OE LABOR STATISTICS.
131
the six years from 1913 to 1918, inclusive. A second table compares
the Bureau of Labor Statistics index with that of the Price Section
for each quarter and month of the period 1913 to 1918. In both of
these comparisons the base period used is the fiscal year, July, 1913,
to June, 1914.10 Comparisons of the Bureau’s index with other series
is also contained in certain issues of the Monthly Labor Gazette of
Canada, the quarterly Bulletin de la Statistique Generale of France,
and other publications of foreign countries.
TABLES OF RESULTS.
Table 2, which has been reproduced from the Monthly Labor Re
view for February, 1921, page 45, shows the yearly movement in
wholesale prices by groups of commodities for the period from 1890
to 1920, inclusive.
T
able
2 .—IN D E X
NUM BERS
OF W H O LESALE PRICES, 1$90 TO 1920, B Y
COMMODITIES.
GROUPS OF
(Average prices in 1913=100.)
Fuel
and
light
ing.
Metals
and
metal
prod
ucts.
Build
ing
mate
rials.
Chem
icals
and
drugs.
94 !
91
91
88
78
69
68
66
66
61
114
102
93
85
72
72
70
67
68
66
90
92
91
90
83
72
72
71
68
67
92
92
88
91
86
81
82
76
77
69
74
67
71
76
75
78
75
75
79
82
67
69
62
61
71
77
80
71
71
108
64
63
62
65
71
88
91
89
93
96
62
58
56
61
62
82
80
80
79
82
7(T
66
67
69
74
69
73
81
75
80
79
80
85
82
87
88
82
84
88
89
80
78
92
105
91
106
98
97
96
88
76
73
77
80
80
97
98
97
96
97
69
69
73
74
73
91
90
92
94
94
80
79
85
85
86
1905..............................
1906..............................
1907..............................
1908..............................
1909..............................
77
78
85
85
97
86
84
89
94
99
91
97
104
94
98
87
90
93
91
88
98
113
120
94
92
85
94
97
92
97
96
94
96
100
101
95
97
101
97
109
85
88
m
91
97
1910.............................
1911.............................
1912.............................
1913..............................
1914..............................
103
93
101
100
103
100
99
108
100
103
99
96
98
100
98
84
82
89
100
96
93
89
99
100
87
101
101
100
100
97
102
103
101
100
101
80
85
91
100
99
116
104
101
100
99
99
95
101
109
100
1915..............................
1916.............................
1917.............................
1918..............................
1919.............................
1920..............................
105
122
189
220
234
218
104
126
176
189
210
236
100
128
181
239
261
302
93
119
175
163
173
238
97
148
208
181
161
186
94
101
124
151
192
308
114
159
198
221
179
210
99
115
144
196
236
337
99
120
155
193
217
236
101
124
176
19&
212
243
Farm
prod
ucts.
Food,
etc.
1890..............................
1891.............................
1892..............................
1893.............................
1894.............................
68
73
66
67
59
89
89
80
87
77
1895..............................
1898..............................
1897.............................. 1
1898..............................i!
1899..............................
60
54
58
61
62
1900..............................
1901.............................
1902.............................
1903..............................
1904..............................
Year.
Cloths
and
cloth
ing.
House
fur
Miscel
nish
lane
ing
ous.
goods.
71 I
74
80
78
77
All
com
modi
ties.
In Table 3 the index numbers are shown for each month from
January, 1913, to March, 1921, inclusive.
10 See also pp. 105 fco 108 of the present bulletin.
INDEX NUMBERS OF WHOLESALE PRICES.
N U M B E R S OF W H O L E S A L E P R IC E S B Y G R O U P S OF
A N D B Y M O N TH S, 1913 TO M A R C H , 1921.
(Average prices in 1913=100.)
Cloths
and
cloth
ing.
Fuel
and
light
ing.
Metals Build Chem
and
ing
icals
metal
and
ma
prod terials.
drugs.
ucts.
House
fur
nish
ing
goods.
arm
rodcts.
Food,
etc.
100
100
100
100
100
100
100
100
100
97
97
99
97
98
99
101
101
104
103
101
101
99
98
97
96
95
99
102
104
105
102
105
101
100
101
100
100
100
99
100
100
100
100
100
99
103
103
102
98
98
100
99
100
100
100
99
99
107
105
102
102
102
100
98
99
99
99
96
92
100
101
101
101
101
101
101
99
99
98
98
98
101
101
101
101
100
100
99
99
99
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
99
98
98
100
100
101
102
101
99
101
102
103
103
104
104
104
109
108
103
101
99
102
100
97
95
96
100
104
112
116
107
106
105
98
99
99
99
99
99
99
99
98
97
96
96
99
99
99
98
95
94
95
94
95
93
93
94
92
92
92
91
87
86
85
85
86
83
81
83
98
99
99
99
98
98
97
97
96
96
95
94
100
100
100
100
100 I
100
99
99
104
105
105
104
99
99
99
99
99
99
99
99
99
99
99
99
100
99
99
98
98
99
100
103
104
99
98
98
102
105
105
107
109
105
108
107
103
105
102
103
106
108
104
105
105
102
104
102
100
103
107
111
96
97
97
99
99
99
99
99
100
103
105
107
93
92
92
89
89
89
90
92
94
96
98
100
83
87
89
91
96
100
102
100
100
100
104
114
94
95
94
94
94
93
93
92
92
93
94
97
103
102
102
102
102
104
108
112
116
124
142
153
99
99
99
99
99
99
99
99
99
99
101
99
100
113
114
115
117
119
119
121
128
133
140
150
145
110
113
117
119
122
124
126
128
131
138
146
155
105
106
108
108
107
108
108
110
115
133
155
170
126
132
141
147
151
149
145
145
148
151
160
185
99
100
101
101
102
101
99
100
100
101
104
106
150
159
160
182
190
187
181
180
178
183
184
185
161
162
164
169
173
179
187
193
193
193
198
202
176
185
188
184
194
201
192
165
160
146
155
158
183
190
199
208
217
239
257
249
226
182
174
174
106
108
110
114
117
127
132
133
134
134
134
135
108
109
111
114
116
116
118
126
131
136
146
142
148
151
163
181
197
197
199
205
204
208
212
205
!
I
j
!
|
I
|
j
I
1
i
nd!S.
101
101
99
101
99
99
100
99
101
103
106
150 1
170
175
172
166
166
156
146
147
150
155
159
105
105
106
108
112
112
121
122
122
124
123
124
112
114
117
118
119
119
123
128
134
144
146
159
160
165
170
179
180
198
209
223
252
240
238
132
132
132
139
139
144
152
152
152
152
155
155
151
156
161
172
182
185
186
185
183
181
183
182
!
;
99
110
UNITED STATES— BUKEAXJ OF LABOR STATISTICS.
T able
133
3 .—IN D E X NU M BER S OF W H O L E S A L E PRICES B Y GROUPS OF COMMODITIES
A N D B Y M ONTHS, 1913, TO M ARCH, 1921—Concluded.
Metals
Build
and
ing
metal
ma
prod
ucts. terials.
Chem
icals
and
drugs.
House
fur
Miscel
nish
lane
ing
ous.
goods.
Farm
prod
ucts.
Food,
etc.
Cloths
and
cloth
ing.
Fuel
and
light
ing.
1918..............................
January...............
February............
March..................
April....................
May.....................
June.....................
July.....................
August................
September..........
October...............
November..........
December...........
220
207
208
212
217
214
217
224
230
237
224
221
222
189
187
186
177
178
177
179
184
191
199
201
206
210
239
211
216
223
232
237
245
249
252
255
257
256
250
163
157
157
158
157
160
159
166
166
167
167
171
171
181
174
176
176
177
178
178
184
185
184
187
188
184
151
136
138
144
146
148
150
154
157
159
158
164
164
221
232
232
232
229
223
219
216
222
220
218
215
195
196
161
161
165
172
173
198
199
221
226
226
226
227
193
178
181
184
191
194
196
190
191
194
196
203
204
196
185
186
187
190
190
193
198
202
207
204
206
206
1919.............................
January...............
February............
March..................
April..... .............
May.....................
June....................
July.....................
August................
September..........
October...............
November..........
December...........
234
222
218
228
235
240
231
246
243
226
230
240
244
210
207
196
203
211
214
204
216
227
211
211
219
234
261
234
223
216
217
228
258
282
304
306
313
325
335
173
170
169
168
167
167
170
171
175
181
181
179
181
161
172
168
162
152
152
154
158
165
160
161
164
169
192
161
163
165
162
164
175
186
208
227
231
236
253
179
191
185
183
178
179
174
171
172
173
174
176
179
236
218
218
218
217
217
233
245
259
262
264
299
303
217
212
208
217
216
213
212
221
225
217
220
220
220
212
203
197
201
203
207
207
218
226
220
223
230
238
1920.
January......................
February...................
March.........................
April...........................
May.............................
June............................
July.............................
August........................
September.................
October......................
November..................
December...................
246
237
239
246
244
243
236
222
210
182
165
144
253
244
246
270
287
279
268
235
223
204
195
172
350
356
356
353
347
335
317
299
278
257
234
220
184
187
192
213
235
246
252
268
284
282
258
236
177
189
192
195 !
193
190
191
193
192
184
170
157
268
300
325
341
341
337
333
328
318
313
274
266
189
197
205
212
215
218
217
216
222
216
207
188
324
329
329
331
339
362
362
363
371
371
369
346
227
227
230
238
246
247
243
240
239
229
220
205
248
249
253
265
272
269
262
250
242
225
207
189
1921.
January......................
February...................
March..........................
136
129
125
162
150
150
208
198
192
228
218
207
152
146 |
139 !
239
221
208
182
178
171
283
277
275
190
180
167
177
167
162
Year and month.
All
com
mod
ities.
INDEX NUMBERS OF THE FEDERAL RESERVE BOARD.
PUBLICATION.
The Federal Reserve Bulletin, issued monthly by the Federal
Reserve Board at Washington, contains a series of index numbers
based on the wholesale price information collected by the United
States Bureau of Labor Statistics. This publication is distributed
without charge to member banks of the system and to officers and
directors of Federal reserve banks. Others are charged a fee for
subscription.
HISTORY.
It was announced in the September, 1918, issue of the Federal
Reserve Bulletin that arrangements had been completed whereby
the index number of wholesale prices computed by the Bureau of
Labor Statistics, together with the data used in its construction,
134
INDEX NUM BERS OF WHOLESALE PRICES.
would be placed at the disposal of the Federal Reserve Board for use
each month in its bulletin. It was further explained that a regroup
ing of the commodities had been made in order to afford a more
satisfactory basis for the study of business conditions. The series of
index numbers, with the method of commodity grouping adopted,
appeared for the first time in the Federal Reserve Bulletin for October,
1918, and has been continued in each subsequent issue.
SOURCE OF QUOTATIONS.
All price quotations used in the preparation of this series of index
numbers are obtained directly from the Bureau of Labor Statistics
and are identical with those used in computing that bureau's series.
BASE PERIOD.
The last prewar year, 1913, forms the base period from which
changes in the price level are measured.
PRICES: H O W SH O W N AND COMPUTED.
To provide a more concrete illustration of actual price movements,
there are presented each month tables of absolute and relative prices
for certain commodities of a basic character. These are corn, cotton,
wheat, cattle, hides, hogs, wool, hemlock and yellow-pine lumber,
anthracite and bituminous coal, coke, copper, pig lead, crude petro
leum, pig iron, cotton yarns, sole leather, steel billets, steel plates, steel
railsr worsted yams, carcass beef, coffee, wheat flour, hams, iHumi
liating oils, and granulated sugar. In the October, 1918, number of
the Federal Reserve Bulletin average prices were shown for each
month from January, 1914, to August, 1918, inclusive. Yearly
averages for 1914 to 1917 also were given. In subsequent issues a
condensed statement for certain months in comparison with the latest
available information is published.
NUMBER AND CLASS OF COMMODITIES.
The number of commodities included in the calculation of this
series is identical with the number used in constructing the Bureau of
Labor Statistics series and, like the latter, is revised from time to time
as conditions make necessary. In 1920 this number was 327, in
cluding both raw materials and manufactured products.
DESCRIPTION AND GROUPING OF COM M ODITIES.
The commodities are divided into three main groups, designated as
raw materials, producers’ goods, and consumers' goods, respectively.
The group of raw materials is in turn divided into four subgroups, viz,
farm products, animal products, forest products, and mineral products.
Thus seven group index numbers, in addition to the general index of all
commodities combined, are currently published.
SUBSTITUTIONS AND ADDITIONS.
Following the plan used by the Bureau of Labor Statistics, additions
to the list of commodities and substitutions of one commodity for
another have been made from time to time as circumstances de
manded. The Federal Reserve Bulletin for October, 1918 (pp. 1006
U NITED STATES---- FEDERAL RESERVE BOARD.
135
to 1008), contains a complete list of the articles included at that time
in each commodity group. Subsequent changes in the list have been
announced as they were made. It is explained that the statistical
methods employed insure continuity in the index number, so that
figures from month to month are entirely comparable.
INTERPOLATION.
With respect to interpolation, the procedure followed has been
that of the source from which the information is taken.
WEIGHTING.
The index numbers are constructed from the aggregates of weighted
money prices in the same manner as the index numbers computed by
the Bureau o f Labor Statistics.
TESTING.
Comparison of the several group index numbers with the general
index number of the Bureau of Labor Statistics is carried in each
issue of the Federal Reserve Bulletin. No other means of testing is
considered necessary.
TABLE OF RESULTS.
Table 4, taken from the Federal Reserve Bulletin for March,
1921, shows the manner in which the information is currently pub
lished:
T
able
4 .—IN D E X NUM BERS OF W H O L ESAL E PRICES IN TH E U N IT ED STATES FOR
PRINCIPAL CLASSES OF COMMODITIES.
(Average prices in 1913=100.]
Raw materials.
Year and month.
July, 1914.....................
January, 1915..............
January, 1916..............
January, 1917..............
January, 1918..............
January, 1919..............
January, 1920..............
July, 1920.....................
August, 1920...............
September, 1920.........
October, 1920..............
November, 1920.........
December, 1920..........
January, 1921..............
Pro
ducers’
Farm
Forest
Animal
Mineral Total raw goods.
products. products. products. products. materials.
|
102
108
116
163
242
234
291
287
259
232
191
170
155
155
106
97
102
136
176
208
213
184
181
186;
172
159
132
119
97
94
95
99
130
147
273
359
351
344
339
289
278 :
245
91
90
112
181
172
179
190
256
265
277
272
246
224
215
99
98
107
148
184
196
239
258
251
248
230
205
186
174
93
95
320
170
181
196
245
251
238
224
209
193
175
169
Con
sumers’
goods.
103
102
111
147
193
216
259
272
250
240
224
214
196 '
184
All com
modities
(Bureau
of Labor
Statistics
index
number).
109
99
111
151
185
203
248
263
250
242
225
207
189
177
INTERNATIONAL PRICE INDEX.
It was stated in the January, 1920, issue of the Federal Reserve
Bulletin that plans were being formulated for the construction of an
international price index. In the May, 1920, issue it was announced
that the technical details had been worked out and an index number
constructed for the United States. The same method was to be
employed in the case of foreign countries. Arrangements for the
136
INDEX NUM BERS OF WHOLESALE PRICES.
cabling of prices each month from foreign countries had been made
with the State Department, and the preliminary work of choosing
commodities and determining quantity weights had been largely
completed for England, France, and Italy.
It was explained that the proposed international index would not
be a single index of world prices, but a group of index numbers for
different countries, all constructed in the same fashion and with the
same base year, the same type of quotations, and approximately the
same number of commodities. The plans contemplated the inclu
sion in the index for each country of a certain number of commodities
of special importance in its economic life, as well as the use of many
staple articles in all the indexes.
Classification o f commodities.— The system of classification of
commodities adopted for the international index conforms in part to
that used by the board for domestic wholesale prices, i. e., (1) raw
materials, (2) producers’ goods, and (3) consumers’ goods. An
additional classification was made in the case of the international
index, viz, (1) goods produced, (2) goods imported, (3) goods con
sumed, and (4) goods exported.
Period covered and base year.— The index number for the United
States covers the months of 1913 and all months since January, 1919.
No effort was made to follow prices through the war period from 1914
to 1918. In computing the index number the prewar year 1913 was
used as the base and prices in each month were measured in terms of
that year.
Source o f price quotations.— It is stated that quotations for the
United States index have been obtained for the most part from trade
journals, although a considerable number have been furnished by
private firms. In general the sources are the same as those used by
the Price Section of the War Industries Board in its study of prices
during the war. In many cases the quotations are the same as those
used b y the Bureau of Labor Statistics and are supplied by that
bureau. In all instances the effort is made to obtain the most rep
resentative quotations for each commodity line.
Number and Tcind o f commodities included.— In constructing the
index number for the United States about 60 different commodities
were used.
Repetition of commodities occurs in the classification
according to source and use, but no duplications have been carried
into the general index for all commodities. In choosing commodities
to represent home production, it is explained that the effort was
made to cover the most important agricultural, mineral, and forest
products. Quotations were selected, as far as possible, to represent
the most common types of the commodities in question. The same
method was used also in the case of imported articles, prices of com
modities received from leading foreign markets being allowed to
stand for the prices of all imports of a particular class.
Weighting.— The system of weighting adopted for the index is
said to be based upon the simple equation: The quantity of goods
produced in a year plus the quantity imported equals the quantity
consumed plus the quantity exported. Figures for the year 1913
were used for weighting purposes since, in most countries, that
represents the most recent normal business year.
In applying the weights to the prices, the commodity quoted was
allowed to represent other commodities in the same general class.
UNITED STATES---- FEDERAL RESERVE BOARD.
137
For example, the weight assigned to crude petroleum in the produc
tion index was the total production of petroleum in the United States,
and not merely the production in the two fields for which prices were
taken. A slightly different method was followed in the case of petro
leum products and hides and skins, the quotations for which rep
resented all but a small proportion of the total production of such
commodities and were therefore weighted by the production of each
kind instead of all kinds combined. It is stated that the production
and consumption weights are only approximately correct, as it was
necessary to resort to estimates in a large number of instances.
The method used for obtaining the weighted index number was
that of multiplying the price of each commodity each month by its
proper weight, adding the several results for each month together,
ana then converting the money aggregates into relatives or index
numbers on the 1913 base. The all-commodities index was obtained
by converting the sum of the aggregates of goods produced plus
goods imported to index numbers in like manner.
Table 5 has been reproduced from the Federal Reserve Bulletin
for March, 1921 (p. 330) :
T
5 . —IN D E X NUMBERS OF W H OLESALE
PRICES IN U N ITED STATES—CON
STRUCTED B Y TH E FE D E R A L R ESERVE BOARD FOR TH E PURPOSE OF IN T E R
NATION AL COMPARISON.
able
(Average price for 1913=100.)
Goods
pro
duced.
Date.
Im
ported.
Ex
ported.
Con
sumed.
Raw
mate
rials.
Pro
ducers’
goods.
Con
sumers’
goods.
All com
modities.
1913
Average for the year..
100
100
100
100
100
100
100
100
1919
Average for the year..
209 '
174
214
206
209
198
207
206
1920
January........................
February.....................
March...........................
April.............................
May...............................
June..............................
July...............................
August..........................
September...................
October........................
November...................
December....................
244
244
250
265
266
260
253
238
231
213
195
178
212
216
218
242
246
226
208
182
164
142
127
112
255
252
256
264
262
256
248
229
211
181
163
146
240
242
247
263
264
257
249
234
227
211
193
176
245
242
246
263
263
258
249
237
233
211
192
176
236
247
263
274
274
265
251
235
225
209
190
171
240
240
241
257
261
255
2.50
229
218
203
187
171
242
242
248
263
264
258
250
234
226
208
190
173
1921
January........................
168
114
142
166
164
166
163
164
INDEX NUMBERS OF WAR INDUSTRIES BOARD.
PUBLICATION.
An index number of wholesale prices in the United States from
1913 to 1918 was published by the Price Section of the War Indus
tries Board in its series of 57 bulletins issued under the title of
“ History of Prices during the War.” 11 These bulletins were pre
pared early in 1919 and given to the public as fast as they were
received from the press. With the dissolution of the War Industries
Board in 1919 its publications were suspended.
11 History of Prices During the War, Buis. 1 to 57, Washington, Government Printing Office, 1919.
138
INDEX NUM BERS OF WHOLESALE PRICES.
HISTORY.
In the preparation of the series of bulletins on prices during the
war period it was recognized that a definite idea of the general trend
of prices in an industry could not easily be gained from a study of the
price quotations for individual commodities. Satisfactory com
parisons of price fluctuations between two or more industries also
could not be made from such data. For persons desiring summaries
for an industry or comparisons for several industries index numbers
were accordingly provided. These index numbers end with Decem
ber, 1918, when the work of the War Industries Board was brought
to a close.
SOURCE OF QUOTATIONS.
The chief sources drawn upon for price quotations were standard
trade journals, newspapers making a specialty of price reporting,
manufacturers or merchants, produce exchanges, boards of trade,
trade associations, and Government bureaus. In general quota
tions from business houses of recognized standing were preferred to
transcripts from periodicals. All sources used were scrutinized
critically and no data were published that were not believed to
reflect faithfully the trend of price fluctuations.
BASE PERIOD.
Since the whole inquiry centered about the effect of the war, the
charts and index numbers were made to show the movements of
prices away from their prewar levels. This was accomplished by
using the 12 months from July, 1913, to June, 1914, as the base
period in the computations.
PRICES : H O W SH O W N AND COMPUTED.
In the 50 bulletins of the series dealing with classes of closely
related commodities average money prices for each commodity are
shown for the base period July, 1913, to June, 1914, and for each
year and quarter of the period covered. Mone}^ prices are also shown
for each month, these being averages of weekly quotations in some
instances and prices on a particular day of the month in other
instances. Each price is the average of the high and low quotations
on the day chosen. It is stated that variations in the method of
compiling prices were necessary because of the diverse sources
from which they were taken.12
In the several bulletins containing summaries for important
groups of commodities, as foods, clothing, and building materials,
as well as in the general summary, no absolute prices are shown, but
the information is presented in the form of index numbers for each
group, by years, quarters, and months.
NUMBER AND CLASS OF COMMODITIES.
It is stated in the introduction to the series of bulletins that in
order to make the investigation widely serviceable it was desired
to include as many industries and commodities as possible. The
“ See Bulletin No. 8, Prices of Feed and Forage, p. 16.
UNITED STATES— W AR INDUSTRIES BOARD.
139
plan adopted called for quotations on a much larger number of com
modities than had been included in other price investigations.
Table 6, taken from Bulletin No. 1 (page 5) shows the 50-classes of
commodities treated and the number in each class. Each individual
series of price quotations, not made by averaging other series shown
in the price table, is counted as one “ commodity.” The consecutive
numbers 8 to 57 assigned to the various classes shown in the table
are those of the bulletins in which the classes are treated.
T able 6 .—TH E 50 CLASSES OF COMMODITIES COVERED B Y THE “ H ISTO R Y OF PRICES
DURING TH E W A R " AND T H E NUM BER OF COMMODITIES INCLUDED IN EACH
CLASS.
Class.
Num
ber of
com
modi
ties.
Num
ber of
aver-
Class.
(37) Matches.............................
Total................................
Building materials group.
(38)
(39)
(40)
(41)
(42)
(4a)
(44)
Chemicals group.
eibihing group.
Total.
91
66
54
157
10
22
20
45
Rubber, paper, and fibers group.
(30) Rubber and rubber products
(31) Paper........................................ .
(32) Fiber and fiber products. . . . . .
Total.
Metals group.
(33) Iron, steel, and their products.
(34) Ferroalloys, nonferrous and
rare m eta ls..........................
Total.
Fuels group.
(35) Coal and coke............................
(36) Petroleum and petroleum
products.................................
Clay products................
Sand and gravel.............
Quarry products...........
Cement............................
Glass.................................
Lumber...........................
Paints and varnishes.,
Total.
Total.
Cotton and cotton products.,
Wool and wool products-----Silk and silk products......... .
Hides and skins and their
products............................... .
(27) Hatter’s fur and fur felt hats..
(28) Hair, bristles, and feathers...
(29) Buttons.....................................
Num
ber of
aver
ages.
Fuels group—Concluded.
Food' group.
(8) Feed and forage........................ .
(9) Wheat and wheat products,...
(10) Corn and corn products..........
(11) Oats, rice, buckwheat, and
their products.............. ........
(12) Barley, hops, rye, and their
products.................................
(IS) Sugar and related products...
(14) Vegetables and truck............. .
(15) Edible vegetable oils........ —
(16) Fruits, nuts, and wine...........
(17) Spices and condiments...........
(18) Tea,, coffee, and eocoa.............
(19) Tobacco, and tobacco pro
ducts.................................... .
(20) Live stock, meats, and fa ts...
(21) Poultry and dairy products.,
(22) Fish and oysters..................... .
(23)
(24)
(25)
(26)
Num
ber of
com
modi
ties.
(45) Mineral acids.............................
(46) Heavy chemicals......................
(47) Miscellaneous inorganic chem
icals..........................................
(48) Fertilizers...................................
(49) Soaps and glycerin...................
(50) Essential oils, flavoring and
perfumery materials............
(51) Wood distillation products
and navalstores....................
(52) Natural dyestuffs and tanning
materials..........................
(53) Coal-tar crudes ^intermediates,
and dyes.................................
(54) Drugs and pharmaceuticals..
(55) Proprietary preparations........
(56) Explosives.................................
(57) Miscellaneous organic chem
icals..........................................
Total.
19
Recapitulation.
Food group...................................
Clothing group..............................
Rubber, paper, and fibers group..
Metals group.................................
Fidels group...................................
Building materials group...........
Chemicals group..........................
332
420
104
117
65
177
281
Total............
Less duplications.
Number of commodities quoted.. .
1,474
1
45
19
1
I
140
INDEX NUM BERS OF WHOLESALE PRICES.
In addition to the 1,474 commodities included in the above table, a
considerable number of American commodities are to be found in the
bulletin on international price comparisons.13 These were not used
in calculating the index numbers. Of the 1,474 commodities in the
table, 108 were not included in the index numbers because they are
not in the market at certain seasons of the year, or for other reasons.
DESCRIPTION AND GROUPING OF COMMODITIES.
The individual commodities making up the 50 classes shown in the
table are described in more or less detail in the box heads over the
tables of average prices in the various bulletins. The grouping
adopted is seen in the foregoing table.
SUBSTITUTIONS AND ADDITIONS.
Owing to the comparatively short period of time covered by' the
index numbers, additions to tne list of commodities or substitutions
of one article for another were unnecessary. The ability to obtain
continuous price quotations for the entire six-year period was a pre
requisite in the selection of commodities.
INTERPOLATION.
In the preparation of the bulletins it was found necessary “ to treat
some of the series rather freely by interpolation to fill gaps, and occa
sionally splicing two series together, with allowance for a price differ
ential. Attention is called to all such cases by footnotes in the tables.
When a choice had to be made between omitting some important
commodity altogether or working up quotations that are not per
fectly regular, the latter course was chosen. ” 14
W EIGHTING.
The method of weighting employed in the construction of the index
numbers is explained as follows:15
The index numbers show the average fluctuations of all the commodities quoted to
represent an industry. To make averages w hich fairly reflect business conditions, it
is obviously necessary to give each com m odity included in an in d ex number an influ
ence upon the result proportioned to its industrial importance. T he meaning of the
figures can best be grasped b y noting the process b> w hich the> are made.
First, the m onthly prices of all the commodities to b e included in the industry were
entered in p aiallel columns on large sheets. Second, an estimate was made for each
com m odity showing the quan tity produced in the United States in 1917, plus the
quantity imported in 'that calendar year. These estimates are technically called
“ W e ig h ts.” T hird, the price of each com m odity each m onth was m ultiplied b y its
respective w eight. Fourth, the products of th e different com m odity prices tim es
com m odity w eights were cast up separately for each of the 72 months from January,
1913, to D ecem ber, 1918. F inally, the resulting aggregates in dollars and cents were
turned into relatives on the plan followed in converting actual com m odity prices-into
relative price for charting. T hat is, the average of the aggregates for the industry in
the 12 months July, 1913, to June, 1914, was called 100, and the aggregate for each
month from January, 1913, to December, 1918, was turned into a relative figure on
that scale.
One deviation from the standard practice should be noted. The materials bought
b y any industry were weighted, not b y production plus imports, b u t by the estimated
quantity consumed in that particular branch of trade.
The year 1917 was chosen as the base year for weighting because the inquiry aims to
show prices under wartime conditions. D ata for 1918 might have been still more ap
13 See War Industries Board Price Bulletin No. 2.
14 War Industries Board Price Bulletin No. 1, p. 7.
^ Idem, No. 1, pp. 9,10.
UNITED STATES---- WAR INDUSTRIES BOARD.
141
propriate, bu t they could not be had for many commodities at the tim e these bulle
tins were being compiled* I t was often difficult to find or make estimates of domestic
production even for 1917, and m any of the weights are subject to a wide margin of un
certainty . W henever i t was necessary to chose betw een guessing a weight or omitting
a com m odity, a guess was made and the com m odity included.
There is an elem ent of unreality in m ultiplyin g 1913 prices b y quantities produced
and imported in 1917. B u t it is clearly desirable to keep the weights constant for the
whole period covered. If one desires to know how prices have changed, then prices
m ust be the only variable adm itted into his computations. If he changes the weights
each year as well as the prices, then he can not tell what part of the results reached is
due to alterations in production and imports and what p artis due to price fluctuations.
A s matters stand, the results are unambiguous— they show the changes in prices and
nothing b u t that.
Made in th e fashion described, the ind ex numbers for the 50 classes of commodities
are strictly comparable not only among themselves b u t also w ith the relative prices
of single commodities shown b y the charts. T he particular form of index adopted—
the weighted aggregate of actual prices reduced to relatives at the end of the computa
tion— is lik e the relative prices of single commodities even in this respect, that i can
be shifted about from one base to another w ithout a tacit alteration of the w eights.16
Thus, the reader who so desires can make his comparisons in terms of prices in 1918 or
any other year or month covered, b v taking the relative figure for his desired base as
100 and converting the other figures into new relatives on the new base.17
The foregoing explanation relates to the manner in which the index
numbers for the 50 classes of commodities were obtained. To obtain
the index numbers for the 7 industry groups and for all commodities
combined a further step was deemed necessary. It is explained
that18—
The statistical materials available concerning the 50 classes of commodities included
in this survey vary so m uch that strictly uniform treatment of all classes is impossible.
In some industries one can obtain prices w hich represent nearly 100 per cent of the
business; in other cases the available quotations represent less than 25 per cent.
Sometimes a large part of the goods can be quoted first as raw materials, second as
intermediate products, and third as finished goods; for exam ple, raw cotton, cotton
yarns, and cotton textiles, or coal-tar crudes, intermediates, and dyes. In other cases,
for exam ple, fertilizers, no satisfactory figures can be had except for the materials,
and in still other cases the finished goods alone can be quoted. A further source of
difficulty is that the “ contract” prices at w^hich the bulk of the business is done in
certain industries are confidential and to use the market prices w hich are available
gives these trades exaggerated importance in such a period as 1913-1918 when the
market ran far ahead of the contract prices. T he class of mineral acids is a case in
point. F inally, in some industries m ost of the raw materials and even of the inter
m ediate products are not sold b u t used b y their producers in further manufacturing
processes, while in other industries each stage of manufacture brings a transfer of goods
to new owners. I f an in d ex num ber is to represent the average change in the prices
that are actually paid and received, an effort should be made to differentiate between
trades w hich call for the paym ent of several prices between the first producer and the
final customer and highly integrated trades where the whole process of manufacture
is performed by a singje business concern.
To repeat, these differences among industries in respect to business organization,
the abundance or scantiness of quotations, and the importance of confidential contract
prices m ake i t im possible to follow a uniform plan in collecting prices. In each
industry the investigator m ust conform to the conditions w hich prevail and accept
the quotations w hich are available. Y e t if the data thus collected are sim ply added
together, the industries for w hich quotations are abundant will be overweighted and
others underweighted in the grand average.
T h e only way to overcome this difficulty is to use a second set of weights. Just as
“ com m odity w eights” were em ployed to secure each article quoted its due influence
upon the ind ex numbers of the class to which it belongs, so “ class w eights” can be
em ployed to secure each class its due influence upon the ind ex numbers for “ all
commodities. ”
*6 For a discussion of this somewhat technical point see “ Themaking and using of index numbers,” in
Bulletin No. 173 of the Bureau of Labor Statistics, pp. 36-44, 91.
17 Of the 1,474 different commodities quoted in the class bulletins, 108 are not included in the index num
bers because they are not in the market at certain seasons o£the year, or for other reasons. Attention is
called to all such omissions in the appropriate places.
is War Industries Board Price Bulletin No. 1, pp. 22-23.
142
INDEX NUM BERS OF WHOLESALE PRICES.
The class weights were obtained by dividing the estimated value
of the goods represented in each class by the class aggregate, i. e., the
total found by adding the products of commodity prices times com
modity weights. The quotients thus found were used to multiply
the class aggregates in each month from January, 1913, to December,
1918, and the results were then added together to make totals for
the 7 groups and for all commodities. The index numbers for the
7 groups and for all commodities combined were made like the index
numbers for the 50 classes by treating the average totals for July,
1913, to June, 1914, as 100 and turning the totals for the 72 months
covered into relatives on that scale. It is stated that19—
A high degree of accuracy is not claim ed for the result. Some of the values m ay be
in error b y a w ide margin— there is no w ay of telling just how wide except b y taking
a census not only of manufactures, b u t also of farming, lumbering, m ining, and fishing,
with the classification here used as a base. A ll that is really required for the present
purpose, however, is a ranking of the industries in the order of their importance, and
the estimates m ade are believed to suffice for that. Rough as it is, the system of class
weights is m uch more trustworthy than the haphazard weighting of classes which
would result from m erely adding class aggregates.
TESTING.
A comparison of the Price Section (War Industries Board) series
with those of the Bureau of Labor Statistics, Bradstreet’s, and Dun's
for the years 1913 to 1918, is contained in Table 7, taken from page
28 of the summary bulletin.
T able
7 .—COMPARISON OF FOUR AM ERICAN IN D E X NUM BERS, B Y Y E A R S , 1913 TO 1918.
(Average prices in July, 1913, to June, 1914=100.)
Year.
The four index numbers on the
1913-14 base.
Rise ( + ) or fall (—) of prices as
compared with year preceding.
Bu
reau of BradLabor street’s Dun's
Statis index. index.
tics
index.
Bu
Price reau of BradSec
Labor
Dun's
tion
Statis street’s
index. index.
index.
tics
index.
Price
Sec
tion
index.
1
1913.............................................................
1914___ l.....................................................
1915.............................................................
1916.............................................................
1917.............................................................
1918.............................................................
101
99
102
126
175
194
100
99
100
123
175
197
103
100
110
132
175
209
99
101
104
122
168
189
- 2
+ 3
+24
+49
+ 19
!
!
j
1
j
- 1
+ 1
+23
+52
+22
- 3
+ 10
+22
+43
+34
+ 2
+ 3
+ 18
+46
+21
Differences am ong the Series.
Average differences.
Index number.
Maximum differences.
Bu
Bu
Price reau of
reau of Brad
Brad Dun’s Price
Labor
Sec
Labor street’s
Sec
Dun’ s
Statis street’s
tion
Statis from— from—
tion
from— from—
tics,
tics
from—
from—
from—
from—
Price Section............................................
Bureau of Labor Statistics...................
Bradstreet’s.............................................
Dun’s.........................................................
0.0
1.5
5.3
3.7
1.5
.0
5.8
3.8
5.3
5.8
.0
8.0
3.7
3.8
8.0
.0
0.0
3.0
15.0
7.0
3.0
.0
12.0
8.0
15.0
12.0
.0
20.0
7.0
8.0
20.0
.0
Averages of differences..........................
3.5
3.7
6.4
5.2
8.3
7.7
15.7
11.7
19 War Industries Board Price Bulletin No. 1, p. 25.
U NITED STATES---- W AR INDUSTRIES BOARD.
143
Table 8, also taken from the summary bulletin (p. 30), the Bureau
of Labor Statistics series and the Price Section series are compared
by years, quarters, and months of the period 1913 to 1918.
T a b l e 8 . — COMPARISON
OF THE IN D E X NUMBERS OF A LL COMMODITIES COMPILED
B Y THE BUREAU OF LABOR STATISTICS AND B Y THE PRICE SECTION OF TH E
W A R INDUSTRIES BOARD, B Y MONTHS, QUARTERS, AND Y E A R S, 1913 TO 1918.
(Average prices in July, 1913, to Jane, 1914=100.)
1913
Period.
1914
1915
1916
1917
1918
Bu
Bu
Bu
Bu
Bu
Bu
reau of Price reau of Price reau of Price reau of Price reau of Price reau of Price
Labor Sec Labor Sec Labor Sec Labor Sec Labor Sec Labor SeoStatis tion/ Statis tion. Statis tion. Statis tion. Statis tion. Statis tion.
tics.
tics.
tics.
tics.
tics.
tics.
Year...................
100
Quarter:
99
First...........
Second........
99
101
Third.........
Fourth
100
Month:
January___
99
February. .
100
March......... • 99
April...........
98
May............
98
June............
100
July............
101
101
August____
September.
102
October___
101
November.
101
December..
99
!
101
99
99
100
102
123
126
175
175
197
194
102
100
101
102
99
98
101
98
100
97
100
98
99
99
100
103
100
100
102
107
112
117
123
141
118
123
125
139
155
179
184
181
152
177
187
182
186
192
202
205
187
190
197
202
103
102
102
101
100
100
100
101
102
102
102
101
100
99
99
98
98
98
99
102
103
99
98
97
100
100
99
98
97
97
97
101
101
99
98
98
98
100
99
99
100
99
101
100
98
101
102
105
100
100
100
100
100
100
102
102
102
104
107
111
110
111
114
116
118
118
119
123
127
133
143
146
115
118
121
123
123
122
123
150
155
160
171
181
184
185
184
182
180
182
181
148
151
156
170
178
183
189
197
186
182
183
182 !
185
187
187
191
191
193
198
202
207
204
206
206
185
187
188
191
190
189
193
196
201
201
201
203
1 2 5
127
132
141
144
;
Comparisons are also made with the Sauerbeck-Statist series for
England, the General Statistical Office series for France, and the
Department of Labor series for Canada.
TABLES OF RESULTS.
Table'9, taken from the summary bulletin,20 shows the index num
bers of all commodities combined and of the seven main industry
groups, also index numbers of products used for the three purposes
stated, for the period 1913 to 1918.
aa War Industries Board Price Bulletin No. 1, pp. 52 and 53.
NUMBERS OF WHOLESALE PRICES.
LUMBERS OF SEVEN GROUPS AN D OF A LL
MONTHS, Q U A R TE R S, AN D Y E A R S , 1913 TO 1918.
(Average prices in July, 1913, to June, 1914=100.)
All
com
mod
ities.
Prod
Build
Rub
ing
Chem ucts
Food Cloth ber, Metals Fuels mate icals
used
ing
group. group. paper, group. group. rials group. for
fibers.
food.
group.
oa:ts
ed
>r
Ildg-
101
99
102
126
175
194
99
101
102
115
162
186
101
96
97
125
177
227
106
98
91
114
145
160
110
93
108
174
262
211
101
95
89
112
158
196
103
98
96
114
148
179
102
101
145
179
176
189
99
101
100
115
159
184
104
97
98
127
168
188
102
100
101
102
97
96
99
103
101
100
100
101
113
106
103
101
118
113
109
100
102
101
101
101
104
105
103
100
103
103
102
100
97
95
99
103
106
107
104
100
100
97
100
98
101
98
103
104
99
99
98
89
98
98
99
99
97
94
93
89
101
96
93
91
99
98
98
95
99
99
101
105
102
96
103
105
98
97
97
93
100
100
102
107
105
102
101
101
91
95
97
105
93
90
90
93
93
102
111
125
89
85
87
96
93
94
95
102
125
134
149
171
103
98
98
102
93
95
98
107
118
123
125
139
106
109
115
128
112
118
125
145
106
112
114
124
155
171
171
199
107
113
111
118
110
113
112
118
196
190
168
162
105
107
115
131
120
152
177
187
182
137
162
169
178
156
168
188
198
141
147
145
146
235
2(86
310
215
131
170
167
167
130
148
155
158
158
170
184 j
192 !
139
161
163
175
116
172
186
168
187
190
197
202
183
177
188
197
213
227
235
234
149
161
165
163
209
209
213
214
174
202
202
207
167
179
184
185
190
190
186
189
180
173
183
200
175
188
194
193
103
102
102
101
100
100
98
96
97
97
95
96
102
102
101
100
100
99
114
113
112
107
105
105
120
118
116
114
113
111
102
101
101
101
101
101
104
104
105
105
105
105
103
104
104
103
103
102
99
97
96
97
94
95
105
106
107
107
107
107
100
101
102
102
102
101
96
100
102
102
103
102
99
100
100
103
102
100
104
104
103
102
101
99
110
110
108
105
100
96
101
102
102
102
102
101
102
102
103
100
100
100
102
101
101
100
100
101
96
99
101
102
104
104
104
104
104
100
100
99
98
97
97
101
101
100
98
98
97
99
99
100
99
99
100
98
98
98
99
98
97
96
98
97
96
92
93
101
100
100
98
96
95
99
99
99
99
98
98
99
99
100
99
98
99
103
102
100
97
95
96
98
98
99
98
97
97
97
101
101
99
98
98
98
105
107
104
103
104
100
99
94
91
88
89
96
101
100
98
98
100
91
94
93
91
89
89
94
94
92
91
91
91
98
98
98
96
95
94
98
99
106
105
106
105
98
104
106
105
105
104
97
98
98
95
94
92
128
126
134
101
100
99
UNITED STATES---- WAR INDUSTRIES BOARD
145
T a b l e 9 . — IN D E X
NUM BER S OF SE V E N GROUPS AN D OF A L L COMMODITIES, B Y
MONTHS, Q U A R TE R S, AN D Y E A R S , 1913 TO 1918—Concluded.
Period.
1
Prod Prod Prod
Build
All
ucts
ucts
Cloth Rub
Chem ucts
ing
com Food
ber, Metals Fuels
used used
used
mod group. ing paper, group. group. mate icals
for
for
rials
group.
group.
for
ities.
fibers.
cloth build
group.
food.
ing.
ing.
Month—Concluded.
1915, January.................
February..............
March....................
April......................
May.......................
June......................
100
100
100
100
100
100
105
106
105
103
103
100
90
92
92
93
95
95
100
90
90
90
90
90
91
93
95
98
101
106
90
89
88
85
85
85
93
93 |
93
93
94
94
123
126
126
133
132
137
105
104
102
99
98
96
95
95
96
97
98
98
93
94
94
94
95
95
July.......................
August..................
September............
October.................
November............
December.............
102
102
102
104
107
111
103
101
99
99
102
103
96
96
98
103
106
107
91
90
89
90
92
95
110
110
114
116
124
136
85
86
90
92
95
100
96
95 j
95 1
101 ;
101
102
146
148
155
162
172
178
99
98
97
99
103
105
99
99
100
101
103
104
98
98
98
105
107
109
1916, January................
February..............
March....................
April.....................
May.......................
June......................
115
118
121
123
123
122
105
106
106
109
109
109
110
113
115
116
118
120
103
104
109
112
112
111
147
154
168
174
171
169
106
107
109
112
113
113
109
111
112
113
113
113
189
200
201
198
188
185
105
105
106
107
107
107
109
111
113
114
116
117
117
120
123
127
128
128
July.......................
August..................
September............
October.................
November............
December.............
123
125
127
132
141
144
111
115
118
125
130
129
122
125
129
135
146
154
112
114
117
120
123
129
167
170
172
176
202
218
113
no
109
111
120
122
112
112
112
116
118
119
175
166
162
162
163
162
111
115
119
127
133
133
119
123
125
130
136
139
126
126
127
131
134
137
1917, January................
February..............
March....................
April.....................
May.......................
June.......................
148
151
156
170
178
183
133
136
142
157
166
164
155
156
157
163
167
174
138
141
143
146
148
147
226
234
247
260
276
315
129
133
131
163
172
173
129
130
132
146
148
151
159
157
159
163
172
174
135
138
144
157
165
161
147
149
152
156
158
166
144
146
149
164
171
180
July.......................
August..................
September............
October.................
November............
December.............
189
187
186
182
183
182
167
168
173
177
182
178
187
189
189
191
199
202
144
143
149
147
146
145
333
313
283
228
209
208
168
169
165
164
167
170
155
155
156
157
159
159
180
183
190
193
191
193
160
162
165
174
176
176
173
178
182
186
187
190
188
188
182
168
168
169
1918, January.................
February..............
March....................
April.....................
May.......................
June.......................
185
187
188
191
190
189
152
184
182
180
177
175
209
212
218
218
226
228
148
148
150
155
162
165
208
209
209
208
209
210
173
174
175
200
204
202
165
165
169
176
179
181
186
192
192
192
190
189
182
182
178
176
173
169
200
204
210
220
223
226
173
174
178
186
188
190
July.......................
August..................
September............
October.................
November............
December.............
193
196
201
201
201
203
182
187
194
195
194
202
233
234
237
238
234
230
164
166
166
165
163
162
212
214
214
216
216
211
201
202
204
204
207
207
182
184
186
185
186
185
184
186
188
190
193
183
176
183
190
195
200
205
231
235
229
230
231
231
192
194
196
193
193
192
INDEX NUMBERS OF THE UNITED STATES FOOD ADMINISTRATION.
PUBLICATION.
A series of index numbers of wholesale prices extending over a
period of seven and a half years was published by the United States
Food Administration in a pamphlet entitled “ General Index Num
bers of Food Prices on a Nutritive Value B a se/’ issued in August,
1918.21 A comparable series of index numbers of producers7 prices,
based on figures from the Monthly Crop Reports of the United States
By Raymond Pearl, chief of the statistical division of the United States Food Administration.
33226 ° — 21— B u ll. 284------- 10
146
INDEX NUM BERS OF W H OLESALE PKICES.
Department of Agriculture, was. also included in the publication.
No subsequent presentation of either of these index number series
has been made.
HISTORY-
In April, 1918, the Food Administration published in a preliminary
way the results of certain computations in which food articles were
assigned varying degrees of importance in accordance with their
respective nutritive values. The pamphlet published in August con
tained a careful revision of the whole subject, with detailed data as
to methods, weighting, etc.
SOURCE OP QUOTATIONS.
The price quotations used in calculating these index numbers were
taken from the weekly report on wholesale prices of the Food Admin
istration issued in manuscript to certain of its members, but not
published. Most of the quotations used were obtained originally
from trade journal sources, some being supplied by manufacturers
or dealers.
BASE PERIOD.
The three years preceding the war, viz, 1911, 1912, and 1913, con
stitute the base period from which changes in the price level were
measured.
PRICES: H O W SHOWN AND COMPUTED.
No actual money prices are published in connection with the tables
of index numbers. It is stated, however, that a weekly range of
prices was obtained for all the commodities on the list. In most
cases Saturday quotations were used, because the majority of the
weekly trade papers close their columns on that day. For the sake
of uniformity quotations from daily papers were also taken on
Saturday. The weekly quotations on each commodity were averaged
for each month.
NUMBER AND CLASS OF COMMODITIES.
Twenty-six commodities were included in calculating the index
number of wholesale prices. These represent for the most part man
ufactured rather than raw food products.
DESCRIPTION AND GROUPING OF COM M ODITIES.
Table 10, which is reproduced as published, shows the principal
facts concerning the data on which the index numbers are based.
T able
Commodity.
1 0 .—QU OTATIONS, DESCRIPTIONS, A N D
Description.1
Price.
Taken.
Wheat flour.. 100 per cent Mill price f. o. b. Satur
day.
spring, in
Minneapolis.
98- pound
cottons.®
Rye flour------ In jute, car . Jobbers price f. o. ...d o ...
lots.
b. New York.
Oatmeal........ Carload tots, Jobber’s price to Satur
in barrels.
day
retailer f. o. b.
New York.
issue.
Corn meal___ W h i t e , in Mill price f. o. b. Satur
day.
bulk.
Terre Haute.
SOURCES OF PRICKS.
Source.
Northwestern Mil
ter.
■Unit.
Calculations.
Bbi. Weekly quotations
averaged for the
month.
........ do.....................
BbL
Do.
New York Com
mercial.
Cwt.
Do,
Furnished by the
mill.
Cwt.
Do.
1 All descriptions as written are for current quotations.
2 Quotations were taken on fCstandard patent” until issue of Jan. 2 , 1918; on “ standard war flour" until
issue of Mar. 27,1918, and standard 100 per cent subsequently*
UNITED STATES— FOOD ADMINISTRATION.
T ABLE
1 0 — Q U O T A TIO N S, D ESCR IPTION S, AND SOURCES OF PRICES—Concluded.
Commodity.
Sugar.
Potatoes, U.
S. average.
Description.
Price.
Taken.
Source.
Granulated,
refined, in
100-pound
bags.
White, bulk
or sacked.
Refiner’ s price,
n e t,!. o. b. New
York.
Satur
day
Weekly Statistical
Sugar T rad e
Journal.
Wholesaler’s price
f.o.b. New York
and Chicago.3
Satur
day
issue.
Journal of Commerce and Daily
Trade Bulletin.
..do..
Onions..
Yellow, bulk
or sacked.
Beans, navy
or pea.
Michigan..
Peanuts.........
No. 1 and
No. 2.4
Rice.
H o n d u ra s,
cleaned,do
mestic.
In 4816-ounce
tins.
Milk, evap
orated.
Milk,
con
densed.
Butter..........
Cheese..........
Margarine...
Lard.............
Ham.............
Bacon...........
Carcass beef.
Mutton.
Fowl___
147
Sweetened, in
4814 ounce
tins.
Fresh, firsts..
Fresh,
92
score.
Fresh, flats,
average run.
.do..
Jobber’s price f. o.
b. Michigan.
Satur
day.
Producer’s Price
C u rren t and
D a i ly T rad e
Bulletin.
F u r n is h e d b y
bean jobbers.
Wholesaler’s price
f. o. b. New
York.3
Satur
day
issue.
Producer’s PriceCurrent (UrnerBarry.)
Unit.
Calculations.
Cwt. Weekly quotations
averaged for the
month; current
price fixed.
Cwt. Average of New
York and Chica
go weekly quotations averaged
for the month.
Cwt.
Do.
Cwt. Weekly quotations
averaged for the
month.
F. o. b. Norfolk... ...d o ... Virginian-Pilot___ L b . Weekly average of
range of No. 1
and No. 2 aver
aged for th e
month.
,.do... N e w
O r le a n s L b .. Weekly quotations
New Orleans.
Board of Trade.
averaged for the
month.
Manufacturer’s ...d o ... Furnished by the Case.
Do.
price delivered
manufacturer.4
New York.
.do..
Case
.. ..d o ....................
Do.
.do .
.do.
Doz. Taken from UrnerBarry Statistical
Review of New
York Market.
.. .d o . . ....... do..................... L b ..
Do.
.do.
...d o ..
Standard high Manufacturer’ s
grade, in 60price, net f. o .b .
pound tubs.
Chicago.
Leaf, in 100- Packer’s price, f.
pound cans.
o. b. Chicago or
branch house.
Loose........
____do....................
B reakfast,
.do.
loose lots.
Good native
.do.
steers.
Legs............... ____d o ..................
Fresh, west Wholesaler’s price
ern, cornfed.
f. o. b. New
York.3
.do.
L b ..
Satur Furnished by the
day.
manufacturer.
...do ..
Provis-
Cwt.
Do .
Satur- Daily Trade Builetm (Howard,
day
Bartels).
issue.
...d o .. ........do....................
Cwt.
Do.
Lb.
Do.
Satur
day.
.. .d o ..
Satur
day
issue.
L b ..
Peas, canned. Standard Ear Canner’s price f. o. Satur
ly June No.
b. Baltimore.
day.
2, in case.
Tomatoes. . . Standard, No.
.. .do .
.do..
3, in ease.
Salmon........
Alaska pink, Canner’s pricef. o. . . -do .
tall No. 1,
b. coast.
case.
Sardines.
Oil keyless Canner’s price f.o. ...d o .
100 j4nch b. Eastport.
case.
National
ioner.
Weekly quotations
averaged for the
month.
L b ..
Do.
National Provisioner.
.do.
Producer’s PriceCurrent (UrnerBarry).
Lb.
Lb.
Canning trade.
.do..
Doz.
.do..
Doz
New Y ork Journal Doz.
of Commerce.
Do.
Do.
Taken from Urner-Barry Statis
tical Review of
New York mar
ket.
Weekly quotations
averaged for the
months
Do.
Do.
Do.
8 Wholesaler, i. e., the receiver of the country produce.
4 Previous to Jan. 25, 1913, peanuts No. 1 and No. 2 grades were called “ strictly prime” and “ prim e/’
respectively.
5 Current prices fixed on competitive market conditions.
6 Based on the views of the Federal Trade Commission of reasonable profit and the Food Administration’s
valuation of reasonable crop-hazard insurance. United States Food Administration, Canned Goods Divi
sion. Bulletin No. 38,
148
INDEX NUMBERS OF WHOLESALE PRICES.
SUBSTITUTIONS AND ADDITIONS.
It is stated that the prices are in most cases quoted from the same
source and for the same grade of commodity throughout the entire
period. In a few cases, where it was impossible to get consecutive
prices on a particular grade, several quotations were spliced. No
substitutions of one article for another was made, nor were there
any additions to the list of articles during the period covered.
INTERPOLATION.
In a few instances the particular grade of a commodity on which
prices were being secured was out of tne market for one or two months
of that period. In such cases, the quotations immediately preceding
and following the gap were averaged to supply the deficiency.
WEIGHTING.
A distinguishing characteristic of these index numbers is the
system of weighting adopted. The food value of the average pro
duction of each commodity during the base period 1911-1913 (except
in a few cases where crop years were used) was expressed in calories
and the value of wheat so expressed was taken as 100. The value
of the other commodities was each in turn related to the value of
wheat. The resulting relative figures were then divided by the
number of pounds in the unit used for price quotations to obtain the
weighting factor. For example, the relative food value of wheat
flour as compared with wheat, based on calories, was found to be in
the ratio of 61.69 to 100. Dividing 61.69 by 196 (the number of
>ou.nds of wheat flour in a barrel) yields 0.3147 as the weighting factor
or the barrel price of that commodity. Similar weighting factors
were worked out for the other commodities in the list. Each com
modity price was then multiplied by its appropriate weighting factor,
the several results added, and their sum divided by the sum of the
weights to produce the absolute index number. The following
formula illustrates the method employed:
f
r (alX^l)-j~(^2X^2)~l~ • • • ~Hfl»X&w)
&1+&2+ • • • -\~bn
where I is the absolute index number, a is the quoted price in what
ever unit given of a commodity denoted by the subscript 1, and i is
the weighting factor for the same commodity. “ In other words,
the absolute wholesale index number is the weighted average price
per pound of the several commodities entering into the index, when
the weighting of each quoted commodity price is in proportion to
the fooa value, expressed in calories, of the average production of
that commodity in the three years preceding the war. The techni
cal student in examining these indices critically should not forget
that the necessary adjustment for difference in the units on which
prices of different commodities are quoted (e. g., barrel, bushel,
hundredweight, pound, etc.) is incorporated in the weighting factor,
to the end of simplicity in computation.” 22 Most of the food values
used were taken from The Chemical Composition of American Food
22 General Index Numbers of Food Prices on a Nutritive Value Base, p. 4.
tration, Washington, D. C., August, 1918.
United States Food Adminis
UNITED STATES— FOOD ADMINISTRATION.
149
Materials, by W. O. Atwater and A. P. Bryant. Absolute index
numbers were reduced to relatives of the 1911-1913 base by dividing
each absolute figure by the average of the absolute figures for the
three years named.
TESTING.
No comparison of these index numbers with other wholesale price
series is made in the publication, nor is any other form of testing
employed. Index numbers of producer’s prices are, however, pub
lished in connection with those of wholesale prices.
TABLES OF RESULTS.
Tables 11 and 12, showing the producer’s and the wholesale price
index numbers for each month from January, 1911, to May, 1918,
are reproduced from page 8 of the pamphlet.
T
able
1 1 . — R E L A TIV E
PRODUCER’ S PRICE IN D E X .
(Average prices in 1911-1913=100.)
Month.
1911
1912
1913
1914
1915
1916
1917
January..................................................................................
February...............................................................................
March.....................................................................................
April.......................................................................................
May.........................................................................................
June........................................................................................
July.........................................................................................
August....................................................................................
September............................
................................................
October..................................................................................
November..............................................................................
December...............................................................................
92
89
88
88
88
92
97
99
101
100
100
100
103
104
107
116
117
113
109
106
103
99
91
90
90
92
93
96
96
99
101
108
109
106
107
106
106
106
106
106
107
106
107
115
114
108
106
108
115
115
114
117
116
111
112
-107
103
101
98
102
108
108
114
114
110
113
118
126
130
136
142
141
149
157
171
210
216
212
229
215
219
206
196
204
T a b l e 1 2 . — RELATIV E
1918
217
218
218
219
218
W H O LESALE PRICE IN D E X .
(Average prices in 1911-1913=100.)
Month.
1911
1912
1913
1914
1915
1916
1917
January..................................................................................
February........................................................................ i ___
March......................................................................................
April........................................................................................
May.........................................................................................
June........................................................................................
July .
......................................................................
August....................................................................................
September.............................................................................
October..................................................................................
November........................................................1....................
December...............................................................................
97
92
90
87
88
89
93
98
101
104
105
106
106
103
102
104
104
101
99
100
103
105
105
106
102
101
101
101
98
98
99
100
102
101
104
103
101
99
97
94
96
97
99
111
116
112
113
112
113
114
109
111
111
106
106
102
99
105
109
114
114
114
116
120
120
118
120
127
132
143
149
144
146
152
157
179
193
185
177
184
188
188
188
191
1918
190
191
178
180
179
I
INDEX NUMBERS OF THE UNITED STATES SENATE COMMITTEE ON
FINANCE.23
PUBLICATION.
The Committee on Finance of the United States Senate published
in 1893 an exhaustive report in which the course of wholesale prices in
the United States was shown by means of index numbers for the 52^
year period from 1840 to 1891. The report was of a special character,
involving an extensive research, and the price data contained therein
23 Report from the Committee on Finance of the U nited States Senate on W holesale Prices, Wages, and
Transportation. Mar. 3, 1893. 52d Congress, 2d session, R eport No. 1394.
150
INDEX NUM BERS OF WHOLESALE PRICES.
have not been continued except in a modified form for subsequent
years.
HISTORY.
A Senate resolution of March 3, 1891, authorized the Committee
on Finance “ to ascertain in every practicable way, and to report
from time to time to the Senate, the effect of the tariff laws upon the
imports and exports, the growth, development, production, and
prices of agricultural and manufactured articles at home and abroad.”
Pursuant to this resolution the committee undertook to ascertain
through accurate and adequate statistics of prices and wages the
changes which had taken place in the condition, as shown by the
relative purchasing power of their earnings, of the great mass of
people in the country for the preceding 50 years. The report of the
committee submitted on March 3, 1893, contained a mass of statistics
relating to wholesale prices compiled by the statistician of the com
mittee, Roland P. Falkner. (See Report on Wholesale Prices,
Wages, and Transportation, Part 1, Appendix A.)
A continuation of this series of prices has been published in Bulletin
No. 27 issued by the United States Department of Labor (now the
Bureau of Labor Statistics) bringing the data down to 1899. In this
latter series, however, two important changes of method were intro
duced. The first was in adopting as a basis the average price for
the nine quarters— January, 1890, to January, 1892, inclusive— in
place of the single-date basis, and the second in departing from the
simple average method of allowing to each article equal weight, and
instead combining the index numbers of similar articles to form one
index number, to be used as one article only in calculating the index
numbers for groups and for all commodities.
Another presentation of the data for the years 1860-1880 in some
what different form (by quarters) is contained in Gold Prices and
Wages under the Greenback Standard, by Wesley C. Mitchell.24
SOURCE OF QUOTATIONS.
The wholesale price quotations included in the report were col
lected mainly by the United States Department of Labor through its
corps of agents and experts. In some cases experts employed
directly by the committee furnished the data for the tabulation. As
a rule, the prices were obtained first-hand; that is, from records of
actual sales. In the selection of articles for quotations the committee
frequently consulted the representatives of leading industries.
“ The greatest care was exercised to secure absolutely accurate
statements, and the books of merchants and manufacturers were
ransacked in order to obtain figures worthy of every confidence.” 25
BASE PERIOD.
It is explained that the year 1840 was not used as a base because
a statement based on that year “ would have rendered comparatively
useless for purposes of comparison all the articles the quotations for
which begin later than 1840.” 26 For this reason the year 1860, which
would include most of the figures presented, was considered prefer24 University of California Publications in Economics, vol. 1, Mar. 2 7 ,1£08.
25 Report from the Committee on Finance of the United States Senate on Wholesale Prices. Wages,and
Transportation, Pt. I, p. 29
26 Idem, p. 28.
UNITED STATES---- SENATE COMMITTEE ON FIN AN CE.
151
able. ’ Moreover, it was believed that “ the year 1860 represents a
period in our industrial development midway between the older
methods of production that prevailed before the war and those which
have come into use since that period. It is also a period of compara
tively normal prices. The markets of the country had recovered
from the crisis of 1857 and the disturbances of trade caused by the
war had not yet taken place. ” 26
Also, a single year, 1860; rather than the average for a period of
years, was taken because “ it was not always practicable to secure for
the articles in question the average prices that would have covered the
period immediately prior to 1860, while in the following year some
prices already manifested the disturbances due to the unsettled state
of national affairs;’ ’ 1860 possessed all the aspects of a normal year.
“ Its price varies little from that of 1859 or 1858 on the one hand and
of 1861 on the other. It is therefore quite as proper a basis of com
parison as would be an average of these four years.” 26
PRICES: H OW SHOW N AND COMPUTED.
As a rule the prices used were actual prices obtained at certain
dates. In a few cases average prices for the year were used, when
such prices were considered representative. The index numbers were
calculated on the basis of the January prices in each year where the
prices were quoted by quarters. An exception was made to this rule
in the cases of those articles for which the January price was not the
representative price for the year, as for fresh vegetables, in which
cases the most appropriate month was selected.
NUMBER AND CLASS OF COMMODITIES.
In all there were 230 series of quotations presented, covering the
prices not only of food products and raw materials but also of a very
large number of manufactured articles. While all series of quota
tions did not cover the entire period, owing to the difficulty of
obtaining for the earlier years prices of articles in use during the
later years, prices for 85 articles quoted in 1891 were secured as far
back as 1840, and for 223 articles as far back as 1860. Those articles
which are articles of luxury only and whose price had increased so
immoderately that they could not be said to enter into consumption
in the same degree as formerly were omitted.
DESCRIPTION AND GROUPING OF COMMODITIES.
The 223 articles were grouped as follows:
Food (53).
Cloths and clothing (28).
Fuel and lighting (10).
Metals and implements (54).
Lumber and building materials (35).
Drugs and chemicals (18).
House-furnishing goods (15).
Miscellaneous (10).
26 R eport from the Comm ittee on Finance of the U nited States Senate on W holesale Prices, Wages,
and Transportation, Pt. I, p. 28.
152
INDEX NUMBERS OF WHOLESALE PRICES.
Following is an enumeration of the articles appearing on pages
30 to 52 of Part I of the report:
Food
Beans.
B read :
A ship bread.
B ship bread.
Boston crackers (two quotations).
N a v y ship bread.
Oyster crackers.
Ship biscuits.
Soda crackers.
Butter.
Cheese.
Coffee, R io, fair.
Eggs.
Fish:
Cod.
Mackerel, salt, shore, N o. 1.
Mackerel, salt, shore, No. 2.
Mackerel, salt, shore, No. 3.
Flour, wheat.
Flour, rye.
Fruit:
A pples, dried.
Currants, Zante.
Raisins.
Lard.
Lard, pure leaf.
Meal, corn, yellow , kiln-dried.
M eat:
Bacon, clear.
Beef, loins.
Beef, salt, mess.
Beef, ribs.
H am , sugar-cured.
L am b.
M utton.
Pork, salt, mess.
M ilk, fresh.
Molasses:
N ew Orleans, prime.
Porto R ico, best.
R ice, Carolina, prime.
Salt:
A sh ton ’s.
A sh ton ’s Li\erpool, fine.
Coarse, solar.
Fine, boiled.
T u rk ’s Island.
Spices:
Nutmegs.
Pepper, whole, Sumatra.
Starch, corn (two quotations).
Sugar:
Brown.
Cut.
Fair refining.
Refined, crushed, and granulated
Tallow, prime, city, in hogsheads.
Vegetables:
Fresh, potatoes, white (two quotations).
UNITED STATES— SENATE COMMITTEE ON FIN AN CE.
153
Cloths and clothing.
Blankets, 1 1-4, 5 pounds to the pair:
Cotton warp, cotton and wool filling.
Cotton warp, all-wool filling.
Broadcloths:
First quality, black, 54-inch, made from X X X wool.
Second quality, black, 54-inch, made from X X wool.
Calico, Cocheco prints.
Carpets:
Brussels, 5-frame, Bigelow.
Ingrain, 2-p ly , Lowell.
W ilton , 5-frame, Bigelow.
Cassimeres, all-w ool:
3 -4 , 7-ounce, Harris double and twist.
3 -4 , 12-ounce, Harris double and twist.
3 -4 , 12-ounce, Harris silk m ixed.
Fancy, 3 -4 , light weight.
Checks, black and white, all-wool, 3 -4 , 7-ounce, Harris.
Cotton, upland, middling.
D enim s, Amoskeag.
Drilling, 30-inch, Pepperell.
H ides, dry, Buenos Aires.
Horse blankets, 6 pounds, all-wool.
Leather, harness.
Print cloths:
28-inch, 64 b y 64, Metacomet.
28-inch, 7 yards to the pound, standard.
Shawls, standard, 72 b y 144 inches, weighing 42 ounces, made of X X
wool.
Sheetings, brown, 4 -4 , A tlan tic A .
Shirtings, bleaches, 4 -4 , New York mills.
Sole leather, first quality, m edium weight, Buenos Aires.
Tickings, Amoskeag, A . C. A .
W ool, Ohio, medium fleece, scoured.
W ool, Ohio, fine fleece, scoured.
Fuel and lighting.
Candles, best adamantine.
Coal, anthracite:
Chestnut.
Grate.
Pea.
S. lum p.
Stove.
Coal, bituminous.
M atches, 8-card.
Metals and implements.
A n vils, domestic.
Bar iron, best refined, rolled.
B utts, loose, joint, cast, 3 b y 3 inch .
Copper, ingot.
Copper, sheet.
Door knobs, mineral.
Iron rods, for making common wood screws.
Iron wire, market, N o. 10.
Lead, drop shot.
L ead:
P ig (two quotations).
P ipe.
L ock s:
Common mortise.
Common rim.
Ohio fleece,
154
INDEX NUMBERS OF WHOLESALE PRICES.
M eat cutters, H a le’ s, No. 12.
N ails, cut.
Pig iron, N o. 1, anthracite, found! y.
Pocket knives:
Redw ood, iron-lined handle, ‘2 J-inch, 1 blade.
Standard, blackhorn, brass G . S. handle, 3|-inch, pen, 2 blades.
Standard—
Cocoa, brass G . S. handle, 3J-inch, 2 blades.
Cocoa, brass G . S. handle, 3f-in ch , 2 blades.
Cocoa, iron-lined handle, 2 f-in ch , 1 blade.
Cocoa, iron-lined handle, 3|-inch, 1 blade (two quotations).
Cocoa, iron-lined handle, 3|-incli, 2 blades.
Cocoa, iron-lined handle, 3f-in ch , 2 blades.
Cocoa, iron-lined handle, 4-inch, 1 blade.
E bon y , brass G . S. handle, 3J-inch, 2 blades.
Ivo ry, brass G , S. handle, 3-inch, pen, 2 blades.
Ivo ry, brass G . S. handle, 3|-inch, pen, 2 blades.
Ivo ry, brass G. S. handle, 3^-inch, pen, 4 blades.
Ivo ry, brass G . S. handle, 3^-inch, pen, 3 blades.
Pearl, brass-lined handle, 3|-inch, pen, 3 blades.
Pearl, silver-lined handle, 3J-inch, pen, 3 blades.
Pearl, silver-lined handle, 3|-inch, pen, 4 blades.
R edw ood, iron-lined handle, 4-inch pruner, 1 blade.
Redwood, iron-lined handle, 5-inch primer, 1 blade.
Stag, brass G . S. handle, 4^-inch, 3 blades.
Stag, brass-lined handle, 3|-inch, pen, 3 blades.
Stag, brass-lined handle, 3|-inch, pen, 3 blades.
Stag, brass-lined handle, 3£-inch, 4 blades.
Stag, brass-lined handle, 4-inch, 4 blades.
Quicksilver.
R ope:
Manila.
Tarred, American.
Tarred, Russian.
Saws:
Circular, 52-inch, Disston’s.
Crosscut, 6-foot, Disston^s.
H and, common, Disston’s.
H and, standard, Disston’s.
Scythes.
Shovels, A m es No. 2, cast-steel, D handle, square-point, back-strap.
Spelter, imported.
W ood screws, 1-inch, No. 10, flat head, iron.
Lumber and building materials.
Brick, common domestic building.
Carbonate of lead, in oil.
Cement, Rosendale.
Chestnut, lumber, in the log, not sawed.
Doors, pine, unmolded, 2 feet 4 inches b y 6 feet 8 inches, 1J inches thick.
Hem lock, boards, first quality, 1-inch, not planed.
H em lock, lumber, in the log, not sawed.
L im e, Rockland.
M aple, boards, first quality, 1-inch, rough.
Oak boards, white, plain, first quality, 1-inch, rough.
Oxide of zinc, American, dry.
Pine, boards, white, clear, 1-inch, not planed.
P ine:
Boards, white, clear, extra, 1-inch, not planed (2 quotations).
Boards, white, common, 1-inch, not planed (2 quotations).
Boards, white, culls, 1-inch, not planed.
Flooring, white, extra, 1-inch, not planed.
Lumber, in the log, not sawed.
Plate glass, polished:
U nsilvered, area 1 to 3 square feet.
Unsilvered, area 3 to 5 square feet.
UNITED STATES---- SENATE COMMITTEE ON FIN AN CE.
Plate glass, polished— Concluded.
Unsilvered, area 5 to 10 square feet.
Unsilvered, area 10 to 40 square feet.
U nsilvered, area 40 to 80 square feet.
U nsilvered, area 80 to 100 square feet.
P utty.
Shingles, pine:
16 inches, X X X .
16 inches, extra X X X cut.
16 inches, extra X X X sawed.
Spruce boards, 1-inch.
Tar, W ilm ington.
Turpentine.
W indow glass:
American, 10 b y 14.
French, 10 b y 14, firsts, single.
French, 10 b y 14, thirds, single.
Drugs and chemicals.
A lcohol.
A lu m , lum p, crystal.
Bichromate of potash.
B lu e vitriol.
Brimstone, crude.
Calomel.
Copperas.
Flaxseed.
Glycerin, refined.
Linseed oil.
Mercuiy.
Muriatic acid.
Opium .
Quinine.
Soda ash.
Sugar of lead:
Brown.
W hite.
Sulphuric acid.
House-furnishing goods.
Furniture:
Chairs, bedroom, m aple, cane seat.
Chairs, kitchen, common, spindle.
Tables, kitchen, pine, 3f-foot.
Glassware:
Bowls, 8-inch.
Goblets, common.
Pitchers, ^-gallon.
Sets, finished.
Tum blers, J-pint.
Pails, wooden:
2-hoop (2 quotations).
3-hoop.
T ubs, wooden (4 quotations).
Miscellaneous*
Powder, rifle (2 quotations).
R ubber, Para.
Soap, castile, m ottled, imported.
Starch:
Ontario.
Ordinary laundry.
Pearl.
Pure.
Refined.
Silver gloss.
INDEX NUMBERS OF WHOLESALE PRICES.
156
SUBSTITUTIONS AND ADDITIONS.
In the compilation of the index numbers no substitution of one
grade or quality of an article for another grade or quality of the same
article or for a different article previously included was made. In
cases where quotations on a particular article could no longer be had,
or where the article had ceased to be representative, it was discon
tinued and the index number was computed on the remaining articles.
Additions to the list of articles were made from time to time as occa
sion demanded, such additions being carried into the index number
for the year.
INTERPOLATION.
Interpolation of prices was not resorted to in the preparation of
the index numbers. In cases where prices for particular periods were
lacking, the article in question was temporarily discontinued.
WEIGHTING.
The committee calculated three distinct index numbers. The first
was unweighted, while the second and third were weighted by assign
ing to each article an importance in the result equal to its importance
in family consumption. The basis selected by the committee for
determining this consumption was the Seventh Annual Report of the
Commissioner of Labor, showing the values of various articles con
sumed by a large number of families which were considered typical
of the expenditures of the mass of the people. This information is
summarized in Table 13, showing the distribution of expenditure for
2,561 normal families.
T
able
1 3 .—
DISTRIBUTION OF E X P E N D IT U R E FOR 2,561 FAMILIES.
Group.
Rent..............................
Food.............................
Fuel............. ...............
Clothing.......................
Lighting......................
All other purposes...
Per cent of
expendi
ture for
each pur
pose.
Propor
tions of
10,000.
15.06
41.03
5.00
15.31
.90
22.70
1,506
4,103
500
1,531
90
2,270
100.00
10,000
This table shows the main groups of family expenditures only, and
these were not sufficiently specific for the purpose. In order to secure
accurate and specific data as to the composition of these groups them
selves, therefore, 232 special budgets of family expenses were collected.
The table based on 2,561 families was then used to secure the propor
tion of the groups entering into consumption, while a table based on
232 families was used to secure the distribution of expenditure within
the groups themselves. The figures thus secured, showing the relative
weight in consumption of each article contained in the family budg
ets, were then applied to the various articles in the index numbers.
Few articles were found, however, with identical descriptions, hence
a method of grouping was resorted to. For example, two or more
articles contained in the index numbers were often grouped to repre-
UNITED STATES---- SENATE COMMITTEE ON FIN AN CE.
157
sent one article of the family budget, thus, “ ham,” “ bacon,” and
“ pork” were considered equivalent to the “ hog products” of the
family budgets, and an arithmetical average was made of the index
numbers of these three articles, which was made the index number
of hog products and given its appropriate weight. Briefly described,
the method as applied to the food group was as follows: The figure
denoting the importance of each article in the group in a given year
was multiplied by its simple index number (or the average where two
or more articles were grouped), and the sum of these products was
divided by the total of the figures denoting importance for the
weighted index number for the general group of food. This method
was used in securing similar weighted index numbers for the other
years of the period. It is seen that according to this method the
same weights were used for each year of the period, although they
were based on the consumption of families in 1891.
Table 14 shows the weighted relative prices or index numbers in
1891 for each of the groups considered, together with the total
weighted relative price or index number:
T able 1 4 .—R ELATIV E PRICES OF A L L ARTICLES IN 1891, M EASURED B Y CONSUMPTION.
(Average prices in 1860=100.)
Impor
tance.
Group.
Weight
ed index
number.
Result.
Rent....................................
Food...................................
Fuel....................................
Lighting.............................
Clothing.............................
All other purposes...........
1,506
4,103
500
90
1,531
2,270
100.0
103.7
98.1
48.1
75.1
95.3
1,506,000
4,254,811
490,500
43,290
1,162,029
2,164,096
Total........................
10,000
96.2
9, 620,726
In the above table the weighted index numbers for the various
groups were found in practically the same manner as has previously
been described for food, except that rent and certain items entering
into the group “ All other purposes” were considered to have re
mained unchanged. These were then multiplied by the figures de
noting importance, and the sum of the products divided by the total
of the figures denoting importance (10,000) to secure the weighted
total of 96.2. The remaining years were treated in a similar manner.
The items of budget expenditure considered as remaining un
changed (rent, taxes, insurance, etc.), constituted 31.40 per cent of
the total expenditure, leaving 68.60 per cent as affected by changes
in prices. Another set of index numbers for such articles was made
by the committee by assigning a total numerical weight of 6,860 to
such articles and working out the index numbers on that basis.
This resulted in a slightly different total index number.
TESTING.
The accuracy of the results secured was tested by comparison of
the index numbers with those of the London Economist and of Sauer
beck for England.27
27 Report from the Committee on Finance of the United States Senate on Wholesale Prices, Wages, and
Transportation, Pt. I, pp. 226,227, and 256.
1S8
INDEX NUM BERS OF WHOLESALE PRICES,
TABLES OF RESULTS.
The index numbers computed by three methods, i. e., simple
average, average of all articles weighted according to consumption,
and average of fluctuating articles only, weighted according to con
sumption, are shown in Table 15. The prices are in currency.28
T able 1 5.—R E L A T IV E PRICE SIN EACH Y E A R , 1840 TO 1891, FO E A L L ARTICLES GROUPED
B Y D IF F E R E N T METHODS.
(Average prices in 1860=100.)
Year.
1840.
1841.
1842.
1843.
1844.
1845
1846.
1847.
1848.
1849.
1850.
1851.
1852.
1853.
1854.
1855.
1856.
1857.
1858.
1859.
1860.
1861.
1862.
1863.
1864.
1865.
All articles
averaged
according
to im
All articles portance,
certain
simply
expendi
averaged.
tures
being
considered
uniform.
116.
115.
107.
101.
101.
102.
106.
106 .
101.
98.
102.
105.
102.
109.
112 .
113.
113.
112.
101.
100.
100.
100.
117.
148.
190.
216.
98.5
98.7
93.2
89.3
89.8
92.1
96.7
96.7
92.0
88.9
92.6
99.1
98.5
103.4
100.4
106.3
108.5
109.6
109.1
102.0
100.0
95.9
102.8
122.1
149.4
190.7
All articles
averaged
according
to im
portance,
comprising
68.60 per
cent of
total
expendi
ture.
97.7
98.1
90.1
84.3
85.0
88.2
95.2
95.2
88.3
83.5
89.2
98.6
97.9
105.0
105.0
109.2
112.3
114.0
113.2
102.9
100.0
94.1
104.1
132.2
172.1
232.2
Year.
1866.................
1867.................
1868.................
1869.................
1870.................
1871.................
1872.................
1873.................
1874.................
1875.................
1876.................
1877.................
1878.................
1879.................
1880.................
1881.................
1882.................
1 8 8 3 .............
1884.................
1885.................
1886.................
1887.................
1888.................
1889.................
1890.................
1891.................
All articles All articles
averaged
averaged
according
according
to im
to im
All articles portance, portance,
simply
certain
comprising
expendi
averaged.
68.60 per
tures
cent of
being
total
considered expendi
uniform.
ture.
J91- 0
172.2
160.5
153.5
142.3
136.0
138.8
137.5
133.0
127.6
118.2
110.9
101.3
96.6
106.9
105.7
108.5
106.0
99.4
93.0
91.9
92.6
94.2
94.2
92.3
92.2
160.2
145.2
150.7
135.9
130.4
124.8
122. 2
119.9
120.5 j
119.8
115.5
109.4
103.1
96.6
103.4
105,8
106.3
104.5
101.8
95.4
95.5
96.2
97.4
99.0
95.7
96.2
187.7
165.8
173.9
152.3
144.4
136.1
132.4
129.0
129.9
128.9
122.6
113.6
104.6
95.0
104.9
108.4
109.1
108.6
102.6
93.3
93.4
94.5
96.2
98.5
93.7
94.4
A convenient summary of the foregoing table, by periods of five
years, is found in Table 16:20
28 Report from the Committee on Finance of the United States Senate on Wholesale Prices, Wages, and
Transportation, Pt. I, p. 9.
«®Idem, Pt. I, p. 10.
UNITED STATES---- SENATE COMMITTEE ON FIN AN C E.
159
T able 1 6 .—R E L A TIV E PRICES, B Y 5-YEAR PERIODS, 1840 TO 1891, FOR A L L ARTICLES
GROUPED B Y D IF F E R E N T METHODS.
(Average prices in 1860=100.)
Period.
1840-1844..............
1845-1849..............
1850-1854..............
1855-1859..............
1860-1864..............
1865-1889..............
1870-1874..............
1875-1879..............
1880-1884..............
1885-1889..............
1890-1891..............
All articles AH articles
averaged
averaged
according
according
to impor
to impor
All articles tance, cer
tance,
simply
comprising
tain ex
averaged. penditures 68.60 per
being con
cent of
sidered
total ex
uniform.
penditure.
108.8
103.2
108.6
108. 2
131.5
178.8
137.5
110.9
105.3
93.2
92.3
93.9
93.3
99.4
107.1
114.0
156.5
123.6
108.9
104.4
96.7
96.0
91.0
90.1
99.1
110.3
120.5
182.4
134.4
112.9
106.3
96. 2
94.1
INDEX NUMBERS OF THE ANNALIST.
PUBLICATION.
The Annalist, a magazine of finance, commerce, and economics,
published weekly in New York City, has compiled an index number
based on the wholesale prices of 25 food commodities in the United
States. These articles are so selected as to represent a theoretical
family food budget.
HISTORY.
The publication of this index number began with the first issue
of the Annalist on January 20, 1913, and has been continued weekly
since that date in connection with the exhibit of various other items
of business activity appearing under the caption of “ Barome tries.”
SOURCE OF QUOTATIONS.
The prices used in the computation of the index number are those
prevailing in the New York and Chicago markets.
BASE PERIOD.
The 10 years 1890-1899 constitute the base period used in com
puting the index number.
PRICES: HOW SHOW N AND COMPUTED.
During the period from May 19 to September 1, 1913, the Annalist
published in each week’s issue the mean price of each selected com
modity during the preceding week, together with the relation of such
price to the price for the base period 1890-1899. The sum of these
relative prices, divided by 25 (the number of commodities) f con
stitutes the index number for the week. In all other issues of the
Annalist up to date no exhibit of wholesale prices appears to have
been made in connection with the presentation of the index number.
INDEX NUMBERS OF WHOLESALE PRICES,
160
NUMBER AND CLASS OF COMMODITIES.
As previously stated, 25 articles of food are included in the index.
These were listed in the Annalist of May 19, 1913, and in subsequent
numbers to September 1 of the same year, as follows:
Steers.
Hogs,
Sheep.
Beef, fresh.
M utton, dressed.
B eef, salt.
Pork, salt.
Bacon.
Codfish, salt.
Lard.
Potatoes.
Beans.
Flour, rye.
Flour, wheat, spring.
Flour, wheat, winter.
Corn m eal.
R ice.
Oats.
A p p les, evaporated.
Prunes.
Butter, creamery.
Butter, dairy.
Cheese.
Coffee.
Sugar, granulated.
DESCRIPTION AND GROUPING OF COMMODITIES.
The following description of the commodities included in the
index number has been supplied by the publishers of the Annalist:
New York Markets.
Codfish (Georges), corn m eal, rice, beans, evaporated apples, California pruned,
extra creamery butter, N ew Y ork State dairy butter, cheese (N ew Y ork State, whole
m ilk , held), N o. 7 R io coffee, fine granulated sugar, fresh beef, dressed m utton,
salt beef, salt pork, wheat flour (winter straights and spring p atents), M iddle W est
lard, and rye flour.
Chicago Markets.
Good to choice steers, hogs (250-300 pound packers and fair to select butcher’ s),
sheep (good to choice wethers), bacon (short, clear sides), white potatoes, and cash
oats (2 white, 3 white, and standards).
SUBSTITUTIONS AND ADDITIONS.
The statement was made in the Annalist of October 13, 1913, that
“ a substitution has been made which affects the current numbers
seven-tenths of 1 per cent.” This was occasioned by the substitu
tion of “ good to choice steers” for “ prime to fancy steers,” as quota
tions on the latter grade had become nominal in the Chicago market.
The entire index number was recast so as to conform to the change
made in this respect. No additions to the list of commodities have
been made.
INTERPOLATION.
No prices have been interpolated, as far as the published informa
tion discloses.
WEIGHTING.
The index number is unweighted and is obtained by computing the
1 rithmetic mean of the relative prices of the different comTESTING.
No test of the index number by means of comparison with other
indexes or by other means is shown in any issue of the Annalist.
UNITED STATES---- THE AN NALIST.
161
TABLE OF RESULTS.
The course of the index number by years from 1890 to 1920 and by
months from January, 1917, to December, 1920, is shown in Tables
17 and 18 compiled from various issues of the Annalist. The monthly
figures here shown have been obtained by averaging the weekly index
numbers as published.
T a b l e 1 7 .— Y E A R LY
IN D E X NUMBERS, 1890 TO 1920.
(Average prices in 1890-1899=100.)
Year.
1890................................
1891................................
1892................................
*893................................
1894................................
1895..............................
1896...............................
1897................................
1898................................
1899................................
1900.................................
Index
number.
109.252
119. 488
108. 624
116.100
102. 076
94.604
80. 096
84. 092
92. 208
93.348
99.388
Year.
Index
number.
104. 656
116. 264
107. 516
108. 664
110. 652
114. 364
117. 940
125. 756
133. 952
137. 172
131. 068
1901...................................
1902...................................
1903...................................
1904...................................
1905...................................
1906..........................
1907..........................
1908...................................
1909...................................
1910...................................
1911...................................
Index
number.
Year.
1912.
1913.
1914.
1915.
1916.
1917.
1918.
1919.
1920.
143.254
139.980
146.069
148.055
175. 720
261.796
287.080
295.607
282.757
T able 1 8 — M ONTH LY IN D E X NUMBERS, JANU AR Y, 1917, TO DECEM BER, 1920.
(Average prices in 1890-1899=100.)
Month.
January......................................................................................
February...................................................................................
March.........................................................................................
April....................................
.
...................................
May.............................................................................................
June............................................................................................
July.............................................................................................
August........................................................................................
September.................................................................................
October......................................................................................
N ovember.................................................................................
December.............................................................................
1917
1918
1919
211.633
222. 585
237. 302
263. 062
283.876
277. 769
264. 693
265. 835
273. 060
279. 587
277. 972
280. 007
279. 368
286.102
285. 838
290.138
289. 343
281. 003
283. 879
289. 829
294. 300
284. 536
286. 658
291. 365
295. 323
281. 668
292. 803
311. 243
315. 667
302. 297
306.133
307. 538
283. 325
279. 791
280. 839
287. 066
1920
368.039
292.504
297.903
314.532
320.825
321.154
303.581
283.359
267.286
247.847
234.544
208.265
INDEX NUMBERS OF BRADSTREET.
PUBLICATION.
This “ index ” 30 represents the lecord of wholesale prices of staple
articles in the primary markets of the United States and is now pub
lished every month. Formerly it was issued only every quarter.
HISTORY.
Bradstreet’s index had its beginning in the issue of September 21,
1895, of the periodical of that name, which presented a table of com
parative prices of 110 staple articles for each quarter from October 1,
1890, to July 1, 1895, under the heading, “ Five years’ prices for 110
staple products.”
The compiler evidently had in mind a record of price movements
in the United States similar to that furnished by Sauerbeck’s index
30Not an index in the true sense of the word, being rather a number representing the aggregate of per
pound prices of certain selected commodities. See also pp. 108-112 of this bulletin.
3 3 2 2 6 °— 21— B ull. 284------- 11
162
INDEX NUM BERS OF WHOLESALE PRICES.
of English prices, as he refers to it in the introductory paragraph as
follows:
l a Sauerbeck’s latest record of prices of staple products in the U nited K ingdom
during the past 30 years it is shown that quotations for 50 selected articles b y groups
averaged lower in 1894 than in any of the 16 n ext preceding years, as w ell as lower
than in the 11-year period from 1867 to 1877, which the em inent statistician selected
as representing the normal and called 100.31
In explanation of the data presented in Bradstreet’s the compiler
says: “ In the accompanying exhibit of comparative prices of staple
articles at primary markets in the United States at quarterly intervals,
beginning with the autumn of 1890, prior to the Baring crash, and
ending with July 1, 1895, is furnished what should prove an oppor
tunity for tracing the relative effects of panic and trade depression
on the prices in different lines of business/*31
In the issue of October 26, 1895, the report was extended to in
clude prices for October 1 of that year and the statement made*
that “ it will be recalled this work was first made public by Brad
street’s late in the summer with a comparison of quotations for more
than 100 articles of merchandise and produce at quarterly intervals
during the past five years.”
Again in the issue of January 11, 1896, the author has this to say
in discussing the compilation: “ Perhaps the most elaborate exhibit
which has been compiled of comparative prices of staple products,
breadstuffs, live stock, provisions, fresh and dried fruits, hides and
leather, raw and manufactured textiles, coal and coke, mineral and
vegetable oils, building materials, chemicals and drugs, and others
is presented in connection with this article. The quotations are
given for quarterly periods during the past five calendar years, and
probably few, if any, better outlines of the movement of quotations
have thus far been presented.”
The comparative prices continued to be presented on the first of
each quarter until May 8, 1897, when in connection with “ A study
of prices” an index number was published for the first time.
The index as constructed was simply the sum obtained by adding
the per pound prices of the different articles included. At first
it was not expressed in dollars and cents, but as an abstract num
ber. No attempt was made at weighting, nor was consumption
taken into account, so that the result was “ not an absolute indication
of the price movement based on the proportions in which each of the
products and articles are used, but a iair indication of the tendency. ”
The author stated that only 97 articles were included in the index,
but as actual prices were shown for 108 articles and only 10 articles
were stated to be excluded it would appear that the index comprised
98 articles.
In the issue of June 11, 1898, actual prices were shown for 107
articles, quotations for onions being dropped, and the index number
was revised to exclude the price of quicksilver. The only explana
tion gives for this was: “ It might be stated in passing that the low
ering of the index number is accounted for by the deduction of the
price of quicksilver.” The index for the period October, 1890, to
June, 1898, was thus recomputed by deducting the price of quicksil
ver; for instance, the old index for January, 1898, was 80,149 and the
new one was 75.084; that is, the price oi quicksilver on January 1,
31 Bradstreet’s, Saturday, Sept. 21, 1895, p. 594.
U NITED STATES---- BRADSTREET.
163
1898 ($0.5065 per pound), was deducted from 80.149, leaving 75.084
as the new index. There were still 10 articles, excluding quicksilver,
not included in the index, but for which comparative prices were
given.
Again on September 10, 1898, the index appeared with revised
figures. This revision was due to the quotation of a different grade
of hides. Previous to this time prices had been quoted for dry
Buenos Aires hides, but for some reason not stated the new quota
tions were for No. 1 native steer hides. The difference between
the prices of these two grades of hides in August, 1898, was $0.0925,
and this deduction from the former index for August (77.481) leaves
76.556 as the new index. This amount was deducted from every
index figure already established as far back as October, 1890.
In the issue of October 12, 1901, the first group indexes were shown
and consisted of the sum of the per pound prices for all of the articles
included in the group. The sum of the 13 groups was the index
shown for all commodities. The general index was expressed in
dollars and cents and continued to be stated this way until April 9,
1904, when it was restated in dollars, cents, and fractions thereof.
This was not a revision of the index, but simply a change in the
method of pointing off. The index numbers- for the groups had been
expressed in this way for some time before this date. The index
now began with January 1, 1892, instead of October 1, 1890, as
formerly, and was computed upon the basis of the revision of Septem
ber, 1898, until December 16, .1905, when a general index “ revised
to exclude some staples showing wide fluctuations ” in price was
published. It is not stated in connection with these figures what
articles were excluded or on how many commodities the revised
index number was based. The exhibit as published contained the
index number by quarters from January 1, 1892, to October 1, 1898,
and by months from January 1, 1899, to December 1, 1905, inclusive.
No further revision of the index number appears to have been made.
SOURCE OF QUOTATIONS.
The source of these quotations is not disclosed, but it is stated
that they are from primary markets.
BASE PERIOD.
No base period was selected in the compilation of the index num
ber, the need of such being obviated by the method employed, which
consists simply in adding together the prices per pound of the various
selected articles at the date named.
PRICES: H O W SH O W N AND COMPUTED.
Prices are published each month for a selected list of representative
commodities. These prices are shown for the first day of the cur
rent month and, for purpose of comparison,, the first day of several
preceding months and the first day of the corresponding month in
the preceding year. No range of quotations is shown in any case,
and it is evident that a single price has been used, but whether either
extreme or the mean was taken it is impossible to determine with
the source of quotations unknown. No yearly average actual prices
of commodities are published.
164
INDEX NUMBERS OF WHOLESALE PRICES.
In the issue of May 8, 1897, the price per pound of each article
was shown as quoted on the first of April, the articles being grouped
under the amount paid per pound. The list was prefaced by the fol
lowing statement: “ Bradstreet’s exhibit of 98 staple, raw and manu
factured articles, products, produce and live stock classified accord
ing to the cost of 1 pound of each on April 1, 1897. ” This exhibit was
continued at intervals for about a year and then dropped. In many
cases the figures appear to have been approximations. The list
as published in Bradstreet’s of July 10, 1897, follows. The prices are
for July 1.
Cost per pound.
$0. 0007
. 001
. 002
. 003
. 004
. 005
. 006
. 007
. 009
. 015
. 02
. 03
. 04
Connellsville coke, southern coke.
Bituminous coal, brick, iron ore.
Anthracite coal.
Salt, southern pig iron, crude petroleum, rosin, lim e, phosphate rock.
Bessemer pig iron, pine lum ber, cotton seed.
Corn, eastern pig iron, tar, spruce, hem lock.
Steel billets.
Oats, barley, rye, potatoes, h ay, sulphuric acid.
Steel rails, steel beams, refined petroleum.
W h eat, m ilk , peas, nails, alum , bicarbonate of soda.
Flour, molasses, beans, paper, caustic soda.
Hogs, lemons, hem p, jute, tin plates, cottonseed oil, turpentine, glass, flax.,
Beeves, sheep, bread, barreled beef, pork, lard, codfish* rice, linseed oil,
raisins, lead, nitric acid.
. 05 Pigs, sugar, currants, borax, bacon.
. 06 Eggs.
. 07 B eef carcasses, m utton, coffee, olive oil, hops.
. 08 Horses, mackerel, cheese, cotton.
. 10 Hams.
. 11 Copper.
. 12 Castor oil.
. 14 Standard sheetings, cotton sheetings, tin, tobacco
. 15 Butter.
. 175 Print cloths.
. 18 Tea, Buenos Aires hides, carbolic acid.
. 20 Hem lock hides, wool.
. 27 U nion leather.
. 29 Oak leather.
. 31 Ginghams.
. 34 Alcohol.
. 50 Australian wool.
. 52 Quicksilver.
. 84 R ubber.
NUMBER AND CLASS OF COMMODITIES.
In the beginning 110 articles were shown in the comparative table
of actual prices, but now only 106 are included, and of these only 06
are included in the index. Oranges, naphtha, onions, and aluminum
were the articles dropped from the table of comparative prices, but
the reason for their discontinuance is not given. Two of these, onions
and aluminum, were never included in the compilation of the index.
Two articles that at first were included in the index are no longer
included— namely, quicksilver and rubber— but these are still shown
in the table of actual prices. When these articles were dropped the
index was recomputed from that date to the beginning, necessitating
a new index figure for every previous date. The list of articles in
cludes both raw and manufactured commodities that are of general
consumption in the United States.
UNITED STATES---- BRADSTREET.
165
DESCRIPTION AND GROUPING OF COMMODITIES.
The articles on which the index is based are divided into 13 gen
eral groups, as follows: Breadstuffs, live stock, provisions and gro
ceries, fresh and dried fruits, hides and leather, raw and manufactured
textiles, metals, coal and coke, mineral and vegetable oils, naval stores,
building materials, chemicals and drugs, and miscellaneous. Since
October 12, 1901, an index has been computed usually for each of
the different groups separately. The sum of the indexes for the 13
groups is the index for the whole number of articles. Index numbers
for years are computed by averaging the 12 monthly totals.
The following list is an enumeration of the articles, under the
various groups, for which actual prices are shown in the comparative
price table. As before stated, only 96 of these 106 articles are in
cluded in the index as now compiled. This is the list and description
of articles as printed in Bradstreet’s of March 13, 1920.
Breadstuffs (6 articles).
W h eat, No. 2, red winter, elevator.
Corn, No. 2, m ixed, elevator.
Oats, No. 3, white, elevator.
Barley, No. 2 (M ilw aukee).
R ye, western.
Flour, straight winter.
Live stock (4 articles).
B eeves, best, native steers (Chicago).
Sheep, prime (Chicago).
Hogs, prime (Chicago).
Horses, average, common to best (Chi
cago).
Provisions and groceries (24 articles).
Beef, carcasses (Chicago).
Hogs, market pigs, carcasses (Chicago).
M utton, carcasses (Chicago).
M ilk (New Y ork).
Eggs, State, fresh (New Y ork).
Bread (New Y ork ).
Beef, fam ily.
Pork, new mess.
Bacon, short ribs, smoked (Chicago).
H am s, smoked.
Lard, western steam.
Butter, creamery, State, best.
Cheese, choice, east factory.
Mackerel, No. 1, bays (Boston).
Codfish, large dried.
Coffee, R io, No. 7.
Sugar, standard, granulated.
Tea, Formosa Oolong, superior.
Molasses, New Orleans, prime.
Salt, fine dom estic, sacks.
R ice, domestic, good.
Beans, choice marrow (New Y ork ).
Peas, choice (New Y ork ).
Potatoes, eastern.
Fresh and dried fruits (5 articles).
Apples (State).
Peanuts, fancy, Virginia, in hull.
Lemons, choice, box.
Raisins, layer.
Currants, new, dried.
Hides and leather (4 articles).
N ative steer hides, No. 1.
H em lock, packer, m iddle, No. 1.
Union, m iddlebacks, tannery run.
Oak, scoured backs, No. 1.
Raw and manufactured textiles (11 articles).
Cotton, m iddling upland.
W ool,
Ohio fine unwashed
delaine
(Boston).
W ool, Ohio half-blood unwashed, combed.
H em p , manila. .
Jute, average of grades.
Silk, shinshiu No. 1, filature.
F lax, N ew Zealand, spot.
Print cloths, 64s (Boston).
Standard sheetings, brown (Boston).
Ginghams, Amoskeag staple (Boston).
Cotton sheetings, southern, 3 yards.
166
INDEX NUMBERS OF WHOLESALE PRICES.
Metals (IS articles).
Iron ore, old range, Bessemer, hem atite.
Pig iron, No. 1 foundry, eastern (N ew
Y ork).
P ig iron, No. 2 foundry, southern (B ir
m ingham).
Pig iron, Bessemer (Pittsburgh).
Steel billets, Bessemer (Pittsburgh).
Steel rails, standard Bessemer (P itts
burgh).
T in plates, Am erican (Pittsburgh).
Steel beams (Pittsburgh).
Silver, commercial bars (N ew Y ork).
Copper, electrolytic (N ew Y ork).
Lead, pig, western (N ew Y ork).
T in , pig, spot (N ew Y ork).
Q uicksilver (San Francisco).
Coal and coke (4 articles),
Anthracite, stove sizes (N ew Y ork).
Bituminous (Pittsburgh), f. o. b. Chicago.
Connellsville coke, short ton, f. o. b.
Southern coke, beehive (Chattanooga).
Mineral and vegetable oils (6 articles) .
Petroleum, crude, barrels (New Y ork).
Petroleum, refined, cases.
Linseed.
Cottonseed, crude, prime (New Y ork).
Castor, No. 1.
O live, Italian, barrels.
Naval stores (3 articles).
Rosin, good, strained (Savannah).
Turpentine, machine, regular (Savannah).
Tar, regular (W ilm ington, N . C .).
Building materials (8 articles).
j Glass, window, 10 b y 15, box.
Pine, yellow , 12-inch and under.
| Tim ber, eastern spruce, wide random.
I Tim ber, hem lock, Pennsylvania, random.
Brick, Hudson River, hard.
Lim e, eastern common.
Cement, domestic, Portland, spot.
Nails, wire, store, base prices.
Chemicals and drugs (11 articles).
A lu m , potash, lum p.
Bicarbonate soda, American.
Borax, crystals.
Carbolic acid, drums.
Caustic soda, 76 per cent.
Nitric acid, 42 degrees.
Sulphuric acidf 66 degrees.
Phosphate rock, South Carolina, ground.
Alcohol, 94 per cent.
O p iu m , cases.
Quinine, sulphate, domestic.
Miscellaneous (7 articles).
Hops, Pacific, choice.
Rubber, upriver, Para, fine new.
Tobacco, m edium leaf, Burley (Louis
ville).
Paper, news, roll, transient.
Ground bone, fine, steam ed.
H a y, prime (N ew Y ork ).
Cotton seed (H ouston).
SUBSTITUTIONS AND ADDITIONS.
Numerous changes in description of the articles have occurred from
time to time, but only once, a p p a r e n t l y , has a substitution been con
sidered of enough importance to justify any change in the index.
This was in the case of dry Buenos Aires hides, for which were sub
stituted No. 1 native steer hides, when the index was recomputed
back to the beginning.
INTERPOLATION.
No method of supplying missing data is disclosed, if such has been
found necessary.
UNITED STATES— BRADSTREET.
167
WEIGHTING.
Apart from the basic plan of expressing in terms of dollars and cents
the value of 1 pound avoirdupois of each commodity, there is no
attempt at assigning varying degrees of importance to the different
articles included in the index.
TESTING.
No test has been made of the index, so far as known, other than a
comparison of the numbers with those published by the London
Economist, the Statist (Sauerbeck's), the Canadian Department of
Labor, and the Oriental Economist (Tokio) for approximately the
same dates.
TABLES OF RESULTS.
Table 19, appearing in Bradstreet's issue of March 13, 1920, illus
trates the manner in which the group index numbers for different
dates are shown.
T able 19.—B R A D ST R E E T ’ S IN D E X NUM BERS FOR SPECIFIED DATES.
Mar. 1,
1919.
Commodity group.
Jan. 1,
1920.
Feb. 1,
1920.
Mar. 1,
1920.
Breadstuffs...............................................................................
livestock.................................................................................
Provisions.................................................................................
Fruits............................. ...........................................................
Hides and leather....................................................................
Textiles.................................................................. .................
Metals........................................................................................
Coal and coke......... ........................ .......................................
Oils................................................................................ ............
Naval stores..,.........................................................................
Building materials..................................................................
Chemicals and drugs..............................................................
Miscellaneous...........................................................................
$0.1881
.6865
4.5974
.3241
2.1350
4.6317
1.1240
.0117
1.0790
.1335
. 2116
1.1363
. 9655
$0.2226
.6685
4.5243
.4506
2. 8000
6.7085
1.0534
.0130
1.0107
. 2870
.2147
1.1825
1.2280
$0.2257
.6610
4.4003
. 4806
2.7700
7.1913
' 1.0714
.0130
1.0875
.3272
.2370
1.1855
1.2185
$0.2241
. 6575
4.3263
. 4556
2.7400
7.2598
1.1207
.0130
1.0728
.3304
.2313
1.1915
1.1720
Total..............................................................................
17.2244
20.3638
20.8690
20.7950
For some years past a yearly index has been computed by averag
ing the 12 monthly indexes. The manner of presenting this in
formation is shown by the following statement, which is reproduced
from Bradstreet’s of April 9, 1921.
1921..........................
1 9 2 0 ........................
1 9 1 9 .........................
1918..........................
1 917 ..........................
1916..........................
1 915..........................
1914..........................
1913..........................
1912..................... ..
1911..........................
191C .......................
1909..........................
1908 ..........................
1907 ..........................
......................... 32 $ 12.0679
.........................
18.8095
.........................
18. 6642
.........................
18.7117
.........................
15.6385
.........................
11.8237
.........................
9 .85 3 0
.........................
8 .9034
.........................
9.21 1 5
.......................
9 .1867
8. 7129
8.9881
.........................
8 .5 1 5 3
.........................
8.00 9 6
.........................
8.90 4 5
1906.
1905,
1904
1903
1902
1901
1900
1899
1898
1897
1896
1895
1894
1893
1892
$8.4176
8.0987
7.9187
7 .9364
7.8759
7. 5740
7.8839
7. 2100
6. 5713
6.1159
5.91 2 4
6.43 4 6
6.6846
7 .5 3 2 4
7.7769
82 The index number for 1921 is an average of the numbers for the first four months of the year.
INDEX NUMBERS OF WHOLESALE PRICES.
168
The index numbers computed from the wholesale prices of 96
articles on the first day of each month from January, 1912, to April,
1921, inclusive, are shown in Table 20, also compiled from Brad
street’s issue of April 9, 1921.
T
able
2 0 .—BR AD STREET’ S IN D E X NUM BERS (FIRST OF EACH M ONTH), JAN U A R Y,
1912, TO A P R IL , 1921, INCLUSIVE.
Year.
Jan.
Feb.
Mar.
Apr.
May.
June.
July.
Aug.
Sept.
Oct.
Nov.
Dec.
1912.
1913.
1914.
1915.
1918.
1917.
1918.
1919.
1920.
1921.
$8.9493
9.4935
8.8857
9.1431
10.9163
13.7277
17.9436
18. 5348
20.3638
12.6631
$3.9578
9.4592
8.8619
9.6621
11.1415
13 9427
18.0506
17.6344
20,8690
12. 3689
88.9019
9.4052
8.8320
9.6197
11. 3760
14.1360
18. 0507
17.2244
20. 7950
11.8650
$9.0978
9.2976
8.7562
9.7753
11.7598
14.5769
18.4431
17.2795
20. 7124
11.3749
09.2696
9.1394
8.6224
9.7978
11.7485
15.1203
18. 8908
17.2376
20. 7341
$9.1017
9.0721
8.6220
9.7428
11.6887
15.4680
18.981S
18.0900
19.8752
$9.1119
8.9522
8.6566
9.8698
11.5294
16.0680
19.1624
18. 8964
19.3528
89.1595
9.0115
8.7087
9.8213
11.4414
16.3985
19.0937
20.0017
18.8273
$9.2157
9.1006
9.7572
9.8034
11.7803
16.6441
19.0260
19.4720
17.9746
$9.4515
9.1526
9.2416
9.9774
12.0399
16.9135
18.9942
19.5215
16.9094
$9.4781
9.2252
8.8620
10.3768
12.7992
17.0701
18.8885
19.9026
15.6750
£9.5462
9.2290
9.0354
10.6473
13.6628
17.5966
19.0151
20.1756
13.6263
INDEX NUMBERS OF DUN.
PUBLICATION.
An “ index” 33 number based on the wholesale prices of a large num
ber of representative commodities in general use in the United States
is published by the mercantile agency of R. G. Dun & Co., of New
York City. The information appears monthly in Dun’s Review, the
weekly journal of finance and trade issued by the above-named
company.
HISTORY.
The publication of this index number was begun in 1901 and
covered a period of time extending back to 1860. From 1901 to
1907 periodical presentation of the index in Dun’s Review appears
to have been made. With the issue of May 11, 1907, however, its
publication was discontinued and apparently was not resumed until
May 9, 1914. The issue of the latter date contained data for the first
five months of the years 1912, 1913, and 1914, respectively, but no
attempt was made in this number to supply figures for all of the period
intervening since 1907. Data for other months of 1912, 1913, and
1914 are shown in subsequent issues; and in Dun’s Review of January
9, 1915, a presentation is made of the index number on the first
of each month for the entire period from 1907 to 1914, inclusive,
thus furnishing a continuous series since the inception of the under
taking.
SOURCE OF QUOTATIONS.
The price quotations on which the index number is based are
those gathered by Dun & Co., in the principal markets of the country,
New York and Chicago prices predominating.
BASE PERIOD.
Under the method followed in the computation of the index num
ber no base period is employed, the index in the case of each article
and group being the actual amount in dollars and cents required to
purchase a year’s supply for a single individual at the date named.
83 Not an index in the true sense of the word, but a statement in dollars and cents of the per capita cost
of a year’s supply of certain commodities at each date named. See also pp. 108-112 of this bulletin.
UNITED STATES---- DUN .
169
PRICES: H O W SHOW N AND COMPUTED.
With regard to the method of calculation, the following statement
is reproduced from Dun’s Review of May 9, 1914:
Quotations of all the necessaries of life are taken and in each case the price is m ulti
plied b y the annual per capita consumption, w hich precludes any one commodity
having more than its proper weight in the aggregate. T h u s, wide fluctuations in the
price of an article little used do not materially affect the “ i n d e x /’ but changes in the
great staples have a large influence in advancing or depressing the total. * * * The
per capita consumption used to m u ltip ly each of m any hundreds of commodities
does not change. There appears to be m uch confusion on this point, but it should
be seen at a glance th at there would be no accurate record of the course of prices if
the ratio of consumption changed. I t was possible, however, to obtain figures suffi
ciently accurate to give each com m odity its proper importance in the compilation.
T his was done b y taking averages for a period of years when business conditions were
normal and every available trade record was utilized, in addition to official statistics
of agriculture, foreign commerce, and census returns of manufactures.
NUMBER AND CLASS OF COMMODITIES.
The following excerpt from the same source shows what com
modities are included:
For convenience of comparison and economy of space the prices are grouped into
seven classes: Breadstuffs include quotations of wheat, corn, oats, rye, barley, beans,
and peas; meats include live hogs, beef, sheep, and m any provisions, lard, tallow,
e tc.; dairy and garden products embrace eggs, vegetable, fru its,m ilk , butter, cheese,
e tc.; other foods include fish, liquors, condiments, sugar, rice, also tobacco, etc.;
clothing covers the raw material of each industry, as well as quotations for woolen,
cotton, silk, and rubber goods, also hides, leather and boots and shoes; metals include
various quotations for pig iron and partially manufactured and finished products,
as well as the minor metals, tin , lead, copper, etc., and coal and petroleum; miscel
laneous includes m any grades of hard and soft lumber, lath, brick, lim e, glass, turpen
tine, hemp, linseed oil, paints, fertilizers, and drugs.
The precise number of articles included in the index is not stated.
In Dun’s Review of January 9, 1915, it is said that “ about 200 pro
ducts are taken.” In a pamphlet entitled “ Commodity prices, a
record covering a period of over half a century, from Dun’s Review,
New York, January 1, 1919,” it is said that “ about 300 wholesale
quotations are taken.”
DESCRIPTION AND GROUPING OF COMMODITIES.
As previously stated,
groups: viz, breadstuffs,
foods, clothing, metals,
description of the articles
the commodities are divided into seven
meats, dairy and garden products, other
and miscellaneous articles. No further
entering into the index is given.
SUBSTITUTIONS AND ADDITIONS.
Additions to the list of commodities for which index numbers
have been computed, or substitutions of a particular grade or quality
of an article for another grade or quality of the same article, if any,
are not shown in connection with any of the published data.
INTERPOLATION.
So far as can be determined from the information at hand concern
ing the long period covered, no interpolation of prices has been made.
INDEX NUM BERS OF WHOLESALE PRICES.
170
WEIGHTING.
As stated in a preceding paragraph, weighting is accomplished by
multiplying the price of each commodity, at the date named, by its
annual per capita consumption “ for a period of years when business
conditions were norm al/7 34 as nearly as could be ascertained by
reference to reliable statistical records. It is stated in Dun’s Review
of January 9, 1915, that “ while it is obvious that the consumption of
some commodities has increased during recent years, it would defeat
the purpose of the index to change the multiplier in any instance,
because there would no longer be a comparative record of the cost of
the same quantities of the same articles back to 1860, as is now the
case.” The issue of September 7, 1901, states that “ while the figures
can not be considered exact, the approximation is sufficiently close
to attain the desired result, and the ratio being constant the com
parison with different dates shows to a cent the rise or fall in the cost
.of living.” 35
TESTING.
No comparison of the index number with those compiled by others
has been made, nor have other means of testing been employed so
far as can be determined.
TABLE OF RESULTS.
The following statistics, showing the trend of wholesale prices from
January 1, 1913, to April 1, 1921, have been compiled from Dun’s
Review of May 11, 1907, January 9, 1915, December 15, 1917, June
12, 1920, and April 9, 1921.
T able
2 1 — W H OLESALE PRICES OF SPECIFIED COMMODITIES, JAN. 1, 1913, TO A P R . I.
1921.
Date.
Breadstuffs.
Meats.
Dairy
and
garden
products.
Other
foods.
Clothing.
Metals.
Miscella
neous.
1913, Jan. 1.................
Feb. 1.................
Mar. 1.................
Apr. 1.................
May 1.................
June 1.................
July 1.................
Aug. 1.................
Sent. 1 .................
Oct. 1.................
Nov. 1.................
Dec. 1.................
1914, Jan. 1 ................
Feb . 1.................
Mar.
Apr. 1........
May 1.................
June 1.................
July 1.................
Aug. 1.................
Sept. 1.................
Oct. 1.................
Nov. 1.................
Dec. 1.................
$19.883
19.565
19.596
19.966
20.673
21.277
21.192
21.632
22.975
22.586
22. 610
23.006
21.961
20.962
22.146
21.402
21.544
23.162
21.086
22.567
26.253
24.441
25. 300
24.426
$10,912
11.522
13.047
13.478
13.183
12.963
13.090
13.080
12.786
13.053
12.211
12.059
12.150
12.625
13.168
12.868
12.813
13.068
12.979
13.427
12.839
12.093
11.907
11.324
$17,925
16.651
16.142
15.319
15.112
16.525
13.039
14.916
1^.604
17.934
19.978
20.454
20.087
18.056
16.009
15.872
16.437
16.114
17.244
16.201
17.432
17.326
18.586
19.825
$11,073
10.877
10.732
10.165
10.120
10.250
10.213
10.267
10.571
10.700
11.068
11.010
10.950
11.002
11.361
10.684
10.467
10.610
10.449
10.284
11.729
11.423
10.880
10.548
$21,015
20.835
21.143
20.938
20.807
20.705
20.534
20.250
20.507
20.947
21.074
20.815
20.664
20.241
20.434
20.641
19.969
20.686
20.834
20.975
20.398
20.259
19.970
19.883
$17,942
17.850
17.379
16.924
16.753
16.760
16.512
16.528
16.742
16.760
16.758
16.596
16.170
16.185
15.881
15.784
15.559
15.695
15.691
15.764
16.126
15.974
15.849
16.134
$22.082
22.428
22.422
22.427
21.676
21.570
21.739
21»842
21.868
21.922
21.804
21.794
22.546
22.570
22.772
22.540
21.441
21.761
21.425
21.522
22.198
22.015
21.848
22.043
Total.
$120.832
119.728
120.461
119.217
118.324
120.050
116.319
118.515
122.053
123.902
125.503
125.734
124.528
121.641
121.771
119.791
118.230
121.096
119.708
120.740
126.975
123.531
124.340
124.183
34 Dun’s Review, May 9,1914, p. 23.
The issue of May 9,1914, contains the statement that “ Dun's index number does not propose to show
the cost of living, because wholesale prices are taken and all luxuries omitted. Its economic value lies m
showing the percentage of advance or decline from month to m onth."
«
U NITED STATES---- DUN.
171
T able 2 1 .—W H O L E S A L E P R IC E S O F SPE C IF IE D C O M M ODITIES, JAN. 1, 1913, TO A P R . 1,
1921—Concluded.
Date.
1915, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1916, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1917, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1918, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1919, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1920, Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
1921, Jan.
Feb.
Mar.
Apr.
Breadstuffs.
Meats.
$25.891
29.052
28.606
28.867
29.807
28.357
26.467
25.999
24.978
23.540
24.024
25.164
27. 318
28. 78i
26.278
26. 703
26.773
25 631
26.378
28.660
31. 061
31.821
36. 772
36.090
36.152
37.865
40.955
43.813
55.360
53.504
53.918
64.071
54.688
55.518
55.680
53.996
54.276
54.001
55.498
57.036
51.328
48.360
51.420
51.620
50.314
49.196
47.472
47.947
48.599
44.999
44.633
49.039
48.873
51.237
51.728
54.757
53.233
48 009
47 529
48.281
48.943
50 626
49.874
52.684
56 965
58.504
57.170
49.871
51.570
42.713
39. 017
32.969
32.697
29.602
31.059
27. 914
$10.705
10.601
10.731
11.073
11.668
12.513
12.134
11.388
11.440
11.469
11.392
10. 551
11. 494
12.233
13.222
14.166
14.611
15.045
14.400
13. 655
14.690
13. 691
14.238
14.248
15.020
16.124
17.031
18-894
19.385
19.810
18.824
17. 746
19.355
19.127
18.168
19.008
19.292
20.577
20-917
22.246
22. 467
22. 362
23.719
23.085
23.664
22.901
21.930
21.556
22.192
21.530
22.027
22.892
24.362
24. 712
25.660
25.105
23.790
20.084
19.144
20.007
19.955
20.937
19.937
20.588
21.384
21.536
22.019
22.124
19.899
19.896
18. 889
16.935
15.240
15. 274
16. 451
15.709
Dairy
and
garden
products.
$19,289
17.464
15.580
15.585
15.464
15.132
15.563
16.030
16.256
18.769
20.616
20. 971
20.509
20.400
20.812
21. 256
20.633
19.267
19.435
17.366
21.541
20.702
24.273
25.403
25.167
27.372
31.509
29.301
30.722
33.606
26.449
21.247
22.751
25.802
25.886
27.021
27.416
28.768
27.123
24.155
23.706
23.826
24.750
24.681
25.009
26.439
27.334
27.631
27.138
24. 705
22.937
24.440
26.120
26. 901
26.160
26.877
26.293
27.983
28.731
30.094
29.077
28.843
28. 727
28.331
28.963
27.944
28.044
26.450
26.039
26.721
26.343
27.205
25.176
22.634
20.121
19. 049
Other
foods.
$10.002
10.478
10.822
10.761
10.705
10.597
10.724
10.970
10.850
10.717
10.956
11.224
11.212
11. 401
11 527
11. 932
12.070
12.231
12.156
12.016
11.962
12.616
13.021
12.923
12.928
12.988
13.166
13.289
13.717
13.865
14.225
15.213
15.552
16.086
18.720
18.767
18. 744
18.848
19.194
20.326
21.414
21.096
21.929
22.307
22.491
23.010
23.367
23.407
23.962
23.400
23.847
23.829
22.727
22.808
23.342
23.695
23.470
23.382
24.157
24.630
24.944
25.447
25.364
25.384
25.246
24.977
25.521
25. 593
24.911
23.589
23.158
21.651
20.690
19.198
19.013
19.044
Clothing.
$19.724
20.117
20.221
20.480
20.786
20.748
20.902
21.400
21.462
21.926
22.325
22.808
23.420
23 601
23.783
24.947
25.139
25. 392
25.800
25.899
26.516
26.826
29.099
30.234
30.082
30.380
30.389
30.678
32.081
33.025
36.527
36.917
38.615
39.436
40.444
40.745
40.880
42.384
42.213
43.322
43.450
44.707
45.238
44.285
44.739
44.533
43.670
43.157
43.194
42.249
40.464
39.173
39.565
41.798
45.623
48.558
47.926
49.852
51.408
52.285
52.778
54.415
54.102
54.752
53.696
51.804
50.268
49.538
46.643
44.838
41.566
38.471
34.108
32.095
29.541
28.814
Metals.
$16,163
16.296
16.343
15.942
15.834
16.138
16.607
16.616
16.956
17.065
17.276
18.328
18.893
19.819
20. 387
20.643
20.889
21.656
21.174
21.057
21. 224
21. 326
21.798
23.390
24.451
25.029
25.977
26.683
28.443
29.888
32.390
32.575
32.657
31.159
29.843
28.413
29.273
29.584
29.914
29.508
29.880
29.936
30.170
30.345
30.609
30.677
30.554
30.394
28.762
28.587
28.217
25.637
25.796
25.559
25.759
26.606
26.533
26.578
26.711
27.727
28.963
29.761
30.400
30.723
30.994
31.017
31.172
32.046
32.846
33.381
32.473
29.871
28.149
26.079
25.109
24.803
Miscella
neous.
$21.794
21.654
21.855
22.383
22.385
22.507
22.561
22.676
22.742
23.177
23.878
24.100
24.820
26.025
26.101
26.043
26.082
26.175
25. 799
25.277
25.024
25. 373
25.639
25.802
25.762
26.515
27.217
27.354
28.727
28.887
29.617
31.010
31.392
32.551
32.009
32.222
32.294
32.858
33.118
33. 720
34.420
34.556
35.349
35.735
36.056
36.471
36.302
34.580
34.912
34.963
34.750
34.958
35.435
36.052
37.097
39.979
40. 893
41.615
42.734
43.719
44.612
45.439
46.084
46.367
46.220
46.666
46.349
46.203
45. 742
44. 526
42. 549
40.940
40.627
39. 071
Total.
$124.168
125.662
124.158
125.090
126.649
125.992
124.958
125.079
124.684
126.663
130.467
133.146
137.666
142.260
142.110
145. 690
146.197
145.397
145.142
143.930
152. 018
152. 355
164.840
168.093
169.562
176.273
186.244
190-012
208.435
212.585
211.950
218.779
215.010
219. 679
220.753
220.172
222.175
227.020
227.977
230.313
226.665
224.843
232.575
232.058
232.882
233.227
230.529
230.375
230.146
220.050
217.037
219.973
222.193
227. 973
233.707
241.651)
238.342
235.867
238.573
244.639
247.394
253.748
253.016
257.901
263.332
262.149
260.414
262.288
248.257
237.311
227.188
211.628
198.609
185. 822
181.921
174.404
172
INDEX NUMBERS OF WHOLESALE PRICES.
IN D E X N U M B E R S O F G IB S O N .
PUBLICATION.
This index of wholesale prices in the United States is published
by Thomas Gibson, New York, every Saturday, in his weekly market
letter.
HISTORY.
In March, 1910, Prof. J. Pease Norton published a “ report on a
new method of compiling index numbers on the Sauerbeck selection
of commodities, modified with the Dun system of weighting,” which
was prepared for use in the weekly market report of Thomas Gibson.36
The work was undertaken as a continuation of the Dun index, which
had been suspended in May, 1907.
In this compilation 50 articles, divided into four general, groups,
were used instead of the much larger number included in Dun’s index.
The general food group was in turn divided into vegetable foods and
animal foods. The descriptions of the 50 articles whose prices formed
the index were the same as those used for these 50 articles in Bulletin
of the United States Bureau of Labor, No. 75. The actual and
relative prices for 1907 of these 50 articles appear to have been taken
from the latter source. The plan followed in the compilation of this
index was intended to be that used by Sauerbeck. It is claimed
that no manufactured or derivative products are included, but that
only primary commodities have been used.
Since November, 1912, only 22 articles, all of which belong to the
food group alone, h&ve been included in the index number.
SOURCE OF QUOTATIONS.
As previously stated, the quotations used to join this index number
with the one compiled by Dun were those published for January,
1907, in Bulletin of the United States Bureau of Labor, No. 75.
The source of later quotations is not given.37
BASE PERIOD.
The years 1890 to 1899 are used as the base period in the compu
tation of the index number.
PRICES: H OW SHOW N AND COMPUTED.
The actual prices of the articles are not shown for any period, the
only data published in Gibson’s weekly market report being the
index for all commodities.
NUMBER AND CLASS OF COMMODITIES.
As has been stated, when this index was first published it covered
50 articles from the farm, mines, and other sources, and included
such as had been subjected only to an initial manufacturing process.
Since November, 1912, it has been calculated on the food group
alone, including 22 articles. It is stated that the articles covered
are those essentially primary in their nature.
36 See also article by Prof. Norton in Quarterly Journal of Economics, August, 1910, pp. 750-759. Pub
lished by Harvard University, Cambridge, Mass.
37 It is stated, however, in the Quarterly Journal of Economics, August, 1910, p. 758 (footnote) that “ Sta
tistics collected from trade journals were used from January, 1909, to compute relative prices.”
UNITED STATES— GIBSON.
173
DESCRIPTION AND GROUPING OF COM M ODITIES.
The present list of articles is divided into two groups, as follows:
Vegetable foods (13 articles).
W heat, contract price.
W h eat flour, spring patents.
W h ea t flour, winter patents.
Barley, b y sample.
Oats, cash.
Com , No. 2, cash.
Corn m eal, fine yellow .
Potatoes, white.
R y e , No. 2.
Sugar, 8 9 °, fair refining.
Sugar, 96°, centrifugal.
Coffee, Rio, No. 7.
Tea, Formosa, fine.
Animal foods (9 articles).
Beef, steers (average of quotations for two grades).
Beef, fresh native sides.
Beef, salt.
M utton, sheep (average of quotations for two grades).
M utton, dressed,
Pork, hogs (average of quotations for two grades).
Bacon, short rib sides.
H am s.
Butter (average of quotations for three grades).
SUBSTITUTIONS AND ADDITIONS.
Since the adoption of the present list there have been no additions
of new articles nor substitutions in the place of those carried, so far
as can be ascertained from the material published.
INTERPOLATION.
Methods of supplying lacking statistical data, if resorted to, are
not disclosed.
WEIGHTING.
The weights assigned to the four groups formerly included in the
index number were 50 for foods, 18 for textiles, 16 for minerals, and
16 for other commodities.
The weighting was accomplished by using a combination of
figures from Dun’s report and the material published by the
United States Bureau of Labor. The first operation was to secure
an average of Dun’s general index numbers for the years 1890 to
1899, which was found to be 0.843. The sum of the relative prices
for all the articles in a group as published for January, 1907, in
Bulletin No. 75 of the Bureau of Labor was then found. These
relative prices were based upon the average for 1890 to 1899 as
100. The sum of these relatives was then divided by the average
of the Dun number, 0.843. The quotient thus obtained was termed
a multiplier. The total of the relatives of a group was then multi
plied by this multiplier. The result divided by 100 was the index
for that group, and the sum of the indexes for the four groups was the
general index number.
174
INDEX NUM BEBS OF WHOLESALE PRICES.
The following statement explains the process of calculating the
index by the above m ethod:
T
able
2 2 .—IL L U S T R A T IV E E X A M P L E OF CALCU LATIN G GIBSON’S IN D E X NUM BERS.
Commodity.
Foods...........................
Textiles.......................
Minerals......................
Other............................
Total
relatives
of the
groups.
Multi
pliers.
Weighted
product.
2422.1
1264.0
1324.8
1408. 7
1.9159
1.6860
1.4987
1.3488
46.4050
21.3010
19.8548
19.0005
Index number
106.5613
Since the reduction of the number of articles on which the index
number is calculated from 50 articles of all classes to 22 food com
modities, no explanation has been given concerning the method of
weighting employed, so far as can be ascertained, ft is stated, how
ever, in Gibson’s weekly market letter of January 11, 1913, and in
subsequent issues that the index number is weighted according to
Dun’s method.
TESTING.
The compiler of Gibson’s index compares the result obtained under
his method, 106.5613, with 107.2640, Dun’s number for the same
period. As a further test to show that figures compiled b y this
method would take the course of Dun’s index number, the following
figures are shown:
Date.
January, 1907............................
February, 1907..........................
March, 1907................................
April, 1907..................................
May, 1907...................................
1896..............................................
Gibson
numbers.
Dun
numbers.
106.56
108.01
109.38
110. 56
113. 41
72. 22
107.26
107.37
109. 91
(107.90)
(109.00)
74.32
The compiler expresses the opinion that in April and May of 1907
the comparison of the two index numbers can not justly be made,
inasmuch as it appears probable the Dun calculator changed his
system of weighting for those months.
TABLES OF RESULTS.
The average yearly index numbers for the cost of foodstuffs, the
only part of the original series now published, as computed b y this
process from 1890 down to the present time, are shown in Table 23
from figures appearing in Gibson’s weekly market letter of April 15,
1921.
U NITED STATES— GIBSON.
175
T able 2 3 .—AVER AG E Y E A R L Y IN D E X NUM BERS, 1890 TO 1920.
Average
yearly
index
number.
Year.
Average
yearly
'index
number.
Year.
Average
yearly
index
number.
Year.
1890...............................
1891...............................
1892...............................
1893...............................
1894...............................
43.4
50.8
45.3
46.0
43.4
1901..............................
1902.............................
1903.............................
1904.............................
1905..............................
44.5
58.5
49.0
48.3
47.3
1912..............................
1913................
1914..............................
1915..............................
1916... .
62.6
58.1
60.8
64.0
74.9
1895...............................
1896...............................
1897...............................
1898...............................
1899...............................
1900...............................
42.0
34.0
34.7
38.7
41.6
44.2
1906.............................
1907..............................
1908.............................
1909..............................
1910............................
1911............................
49.8
50.9
51.2
59.2
59.3
56.9
1917..............................
1918..............................
1919..............................
1920..............................
110.8
122.8
121.4
127.0
Monthly averages from January, 1910, to March, 1921, also shown
in the publication referred to above, are as follows:
T able
2 4 .—AVER AGE M O NTH LY IN D E X NUM BERS, 1910 TO M ARCH, 1921.
January...........................
February.........................
Marcli...............................
April................................
May..................................
June.................................
July..................................
August.............................
September.......................
October............................
November.......................
December........................
1910
1911
1912
1913
1914
1915
1916
60.8
61.2
63.6
62.6
60.8
59.0
59.6
59.8
59.6
56.9
54.2
53.8
54.3
52.9
53.3
53.1
53.5
52.9
57.5
€0.1
61.2
62.0
61.3
60.3
62.3
61.7
63.8
67.8
69.4
66.6
62.6
61.1
60.4
59.8
59,1
56.8
55.5
57.0
57.8
59.0
57.8
57.3
58.6
59.3
60.0
58.4
58.4
58.2
58.-2 ; 64.7
58.2 68.0
57.8 66. 7
57.7 67.8
57.9 68.3
59.7 64.3
58.9 64.4
64.9 63.1
68.6 58.5
62.9 60.0
63.1 60.6
62.3 62.1
65.6
68.2
69.5
71.3
72.3
70.8
71.9
76.1
78.4
82.2
87.1
85.1
; 1918
1919
1920
87.4 118.9
90.5 121.9
96.4 ; 126.1
109.2 130.5
118.5 125.7
114.2 122.7
116.4 123.3
117.9 122.2
119.4 123.2
120.1 119.0
119.9 118.8
121.4 120.9
119.7
116.7
122.9
129.8
128.0
122.9
127.9
126.7
116.9
111.2
114.6
119.4
130.4
127.3
133.8
147.4
155.4
154.7
141.9
125. 8
118.5
106.9
95.7
86.0
1 9 1 7
1921
81.9
78.8
80.8
AUSTRALIA.
INDEX NUMBERS OF THE COMMONWEALTH BUREAU OF CENSUS AND
STATISTICS.
PUBLICATION.
In December, 1912, a report entitled “ Prices, Price Indexes, and
Cost of Living in Australia,” compiled by G. H. Knibbs, Common
wealth statistician, was published by the then recently organized
Labor and Industrial Branch of the Commonwealth Bureau of Census
and Statistics at Melbourne as its Report No. 1. Both wholesale and
retail prices, together with import and export index numbers for
Australia, were among the subjects considered in the report.
This publication was the first of a series designed to include topics
covering general industrial conditions as well as prices. Its main
object, as stated in the preface, was “ to furnish information as to
prices in past years in such a form as to be fully comparable with
that which it is proposed to publish periodically in the future.”
The continuation of these index numbers is to be found in Report
No. 2, entitled “ Trade-Unionism, Unemployment, Wages, Prices,
and Cost of Living in Australia, 1891, to 1912,” under date of April,
1913, and in subsequent reports, numbered 5, 6, 7, 8, 9, and 10, issued
in December, 1914, May# 1916, June, 1917, July, 1918, July, 1919, and
176
INDEX NUMBERS OF WHOLESALE PRICES.
October, 1920, respectively. Quarterly index numbers, also, are
shown in the Labor Bulletin, published quarterly up to June, 1917,
by the Labor and Industrial Branch of the Australian Bureau of
Census and Statistics.
HISTORY.
The author of these reports, having studied the various systems
of index numbers published in other countries, became convinced
that the methods ordinarily followed were so defective as to be mis
leading. He believed that an accurate system of inquiry should be
determined upon to secure reliable and satisfactory data on which
to base index numbers, and further that a uniform method for the
international study of prices as a basis for world index numbers
should likewise be worked out by those economists interested in the
subject. These conditions he undertook to meet.
SOURCE OF QUOTATIONS.
A t first an attempt was made to secure from market reports whole
sale prices of a representative list of commodities for the capital
towns of each State. This plan was abandoned when it was found
to be impracticable owing to the lack of complete records and to the
difficulty in obtaining comparable returns. Moreover, the compila
tion of figures for Melbourne alone involved so much labor that no
attempt was made to complete any other city.
The figures for Melbourne were obtained mainly from market prices
published in the ordinary press and in special trade reviews. Where
there was any question as to the reliability of the quotations the
figures were verified by “ reference to well-known and important
business firms dealing in the articles in question.”
BASE PERIOD.
The basic period selected for the computations of the wholesale
price indexes was the year 1911, the aggregate expenditure on all arti
cles and on each group of articles in this year being taken as 1,000.
To quote: “ The index numbers show the amount which would have
had to be expended in each of the years specified in order to purchase
what would have cost £1,000 in 1911, distributed in purchasing the
relative quantities (indicated by the mass units) of the several com
modities included in each group and in all gtoups respectively.”
In the reports for the years 1916 and after, as well as in the quar
terly Labor Bulletin after 1914, index numbers are also shown 6n the
July, 1914, base as 1,000 in order that the variations in price levels
since the outbreak of the war may be clearly seen.
PRICES: H O W SH O W N AND COMPUTED.
A table in the appendix to the first report shows the average annual
wholesale prices in Melbourne of all commodities included in the
investigation except meats from 1871 to 1912 (first 9 months only),
inclusive. Data for all of the 92 commodities, now included in the
index numbers, are contained in the appendix to later reports.
Prices for meat were not obtained for the years prior to 1884 and
1885 nor for the years 1886 to 1889, inclusive. The unit of meas
urement is given for each article and the price stated in shillings
AUSTRALIA— BUREAU OF CENSUS AND STATISTICS.
177
and pence. The articles for which prices are quoted are divided
into eight groups. In most cases monthly prices were obtained, and
from these the yearly avarages were computed. For tea, coal, cotton,
wool, and silk, however, monthly prices were not available, so that
yearly averages based in each case on expert opinion 38 were secured.
The monthly quotations, from which the yearly average is com
puted, are not shown.
The group and general index figures for 1861 and'1866 are shown
elsewhere in the initial report, but no actual prices are given for these
years.
NUMBER AND CLASS OF COMMODITIES.
In the computation of index numbers for the years prior to 1911
(the base year) the aggregate expenditure on 80 commodities w^as
used, while for 1912 and subsequent years the number of commodities
included was increased to 92. The author states that the commodi
ties included are generally in the nature of raw materials— that is,
materials in which the labor cost is relatively low.
There are no articles of clothing, boots or shoes, or furniture
included. The reason assigned for their omission was the imprac
ticability of obtaining periodic prices for predominant grades and
qualities and of satisfactorily determining the relative importance in
consumption of the various items, the author contending that the
character of clothing and of furniture includes the element of change
due to the influence of fashion, and that where incomes are limited
economy strikes first at these articles.
DESCRIPTION AND GROUPING OF COMMODITIES.
The 92 commodities used in computing the index numbers for 1912
and subsequent years are divided into 8 groups, as follows:
I. Metals and coal, 14 commodities.
II. Textiles, leather, etc., 10 commodities.
III. Agricultural produce, 16 commodities.
IV. Dairy produce, 9 commodities.
V. Groceries, 21 commodities.
VI. Meat, 5 commodities.
VII. Building materials, 10 commodities.
VIII. Chemicals, 7 commodities.
The list of articles, with their description or brand, the unit upon
which quoted, and the “ mass unit” — i. e., the extent to which a com
modity is used— are shown in Table 25.
38 Source of this expert opinion not stated.
33226°— 21— Bull. 284------- 12
178
1NTMK KtTMBEBS OF WHOLESALE PRICES.
TA-BLE 2 5 . — M ELB O U RN E
W H O L E SA L E PRICES, COMMODITIES INCLUDED, UNITS OF
M EASU R EM ENT, AND MASS UNITS.
(Commonwealth Bureau of Census and Statistics, Labor and Industrial Branch, October, 1920, Report
No. 10, p. 67.F
Group I.— Metal# and coal (1$ commodities).
Commodity.
Iron:
Pig ...................................................................................
Rod and bar.......................................................................
Angle and T .......................................................................
plate....................................................................................
Hoop...................................................................................
Galvanized..........................................................................
Tinned plates............................................................................
Fencing wire ............................................................................
Zinc, sheet..................................................................................
Lead sh°-ct.................................................................................
Lead pipes.................................................................................
Copper^ sheet............................................................................
Quicksilver.................................................................................
Coal............................................................................................
Brand.
Unit.
Ton......................
........do..................
........do..................
........do..................
........do..................
........ do..................
100 pounds..........
Ton.......................
........do..................
........do..................
........do........... .......
Pound.................
........do..................
N e w c a s t l e , on Tons....................
wharf.
Mixed numbers__
Stafford..................
........do......................
........do......................
........do......................
26-gauge................
I. C. coke...............
No. 8 .......................
Mass
unit*
6|
3*
3J
3
5
60
6
1
f
i
2,000
12
600-
2,7021
Total................................................................................
Group I I .— Textiles, leather, etc. (10 commodities).
Bran bags...................................................................................
Com sacks..................................................................................
W oolpacks................... . . ......................... ................................
Leather:
Waxed kip..........................................................................
Waxed split................................................................ .
Medium crop......................................................................
Cotton ....................................................................................... Raw ........................
Wool
................................................................................... Greasy....................
Twine ...................................................................................... Reaper and binder
Tallow......................................................................................... Mutton, prime___
Dozen..................
........do..................
Each....................
Pound.................
........do..................
........d o..................
........do..................
........do..................
........do..................
Ton......................
Total............................. ..................................................
110
250
200
600
600
600
24,000
12,200
150
11
38,71 If
Group III.—Agricultural produce (16 commodities).
Bushel.................
Wheat.........................................................................................
Ton......................
Flour...........................................................................................
........do..................
Bran............................................................................................
........do..................
Pollard........................................................................................
Oats__ .'...................................................................................... Feed........................ Bushel.................
Oatmeal...................................................................................... Colonial................... Ton......................
Barley......................................................................................... Malting................... Bushel.................
D o......................................................................................... Feed........................ ........do..................
........do..................
Maize...........................................................................................
Best manger.......... Ton......................
Hav ............................................................................................
Chaff........................................................................................... Good oaten............ ........do..................
Straw............................................................. - ........................... Victorian................ ........do..................
Peas.............................................................................................
B u sh el...............
Potatoes......................................................................................
Ton......................
Malt........................... *................................................................ Victorian................ Bushel.................
Ton......................
Onions.........................................................................................
Total.................................................................................
500
48
14
14
1,200
1*
150
100
1,000
135
135
25
55
40
140
3
3,560^
Group I V .—Dairy produce (9 commodities).
H am ............................................................................................
Bacon..........................................................................................
Cheese..........................................................................................
B'utter......................................................................................... Best fresh...............
Lard............................................................................................. In bladders............
Eggs............................................................................................. Ordinary................
Honev.........................................................................................
Beeswax......................................................................................
Condensed m ilk........................................................................
Bacchus Marsh...
Total.................................................................................
Pound.................
........do..................
........do..................
........do..................
. . . do..................
Dozen..................
Pound.................
........do..................
Dozen pounds...
800
3,200
1.500
9.500
200
1,800
600
40
160
17,809
AUSTRALIA— BUREAU OF CENSUS AND STATISTICS.
179
t a b l e 2 5 .—M E L B O U R N E W H O L E S A L E PR IC E S, C O M M O D IT IE S IN C L U D E D , U N IT S OF
M E A S U R E M E N T , A N D MASS U N IT S —Concluded.
Group V — Groceries (21 commodities).
Commodity.
Currants.
Raisins...
Herrings.
Brand.
Sultanas..........
1-pound, fresh..
Salmon...
1 - pound, tall,
Alaska.
Halves...................
Plantation...........
Taylor’s ................
No. 1A...................
Sardines..
Coffee___
Cocoa___
Sugar----Macaroni..
Sago........
Rice........
Patna................
Liverpool fine..
Rock..................
Coleman’s....... .
Salt.........
Do ..
Mustard..
Starch...
Blue----Matches.
Candles..
Tobacco.
Coleman’ s white...
Keen’s ....................
Australian safety..
Gouda................... .
T w o S e a s , in
pocket pieces.
Tea..........
Kerosene.
Unit.
Pound.................
____do..................
Dozen, 1-pound
tins.
------do..................
unit.
1.400
1.400
50
Dozen halves___
Pound.................
____d o ................
Ton......................
Pound.................
100 pound...........
Ton......................
____do..................
------do...................
Dozen, 1-pound
tins.
Pound.................
____do..................
Gross. ...............
Pound.................
____do................ .
50
90
1,600
1,300
____do.
Gallon.
3,000
1,700
Total.
100
200
100
22
200
7
2
7
1
6
100
11,385
Group VI.— Meat (5 commodities).
Beef..............................................................................................
Mutton........................................................................................
Veal..............................................................................................
Lam b............. ...........................................................................
Pork............................................................................................
Average quality...
........do......................
........do.................... .
........do......................
........do......................
100 pounds.........
Pound.................
........do..................
........do..................
........do..................
Total.................................................................................
390
33,000
2,000
5,600
3,700
44,690
Group VII.—Building materials (10 commodities).
Flooring:
6 x l£ ...........
6 x |...........
Timber.
6x f .......
6 x J...........
W eatherboards..
Oregon................
Shelving.
Portland.
Cement___
White lead.
Slates..........
Welsh, 20 x 10..
100 feet linear...
....... do................
....... do................
....... do................
....... do................
1.000 feet super
ficial.
....... do................
Cask...................
Ton....................
1.00 0
30
30
30
30
200
20
10
30
f
*
381J
Total.
Growp V III.— Chemicals (7 commodities).
Cream of tartar.......................................................................... In kegs....................
Carbonate of soda.....................................................................
Saltpeter............................
.
....................................... Refined...................
Sulphur......................................................................................
Caustic soda...............................................................................
Alum ........................................................................................... Lump.....................
Cyanide potassium...................................................................
Total.................................................................................
Pound.................
Ton......................
........do..................
........do..................
100 pounds.........
Ton......................
Pound.................
400
*
?
\
570
OT8*
180
INDEX NUMBERS OF WHOLESALE PRICES.
In all reports index numbers are given for all groups taken as a
whole and for each group. Under each group in the first report is
shown the index number for a few individual articles of importance
computed on the price in 1911 as the base, but no table is given
showing an index for each of the articles separately.
SUBSTITUTIONS AND ADDITIONS.
Cases of substitution of a particular grade or quality of an article
for another grade or quality of the same article, if any, are not ap
parent in the tables, owing to the manner in which the information
is presented. The author states, however, that “ every care was
taken to insure that the prices quoted for each article refer to a uni
form quality” and that “ special precautions were taken to insure
substantial continuity of quality or grade.”
In the computation of the index numbers for 1912 in the second
report, as previously stated, the author added 13 articles and dropped
raw silk, so that the index for 1912 and later years covers 92 articles
instead of 80, as formerly. The aggregate expenditure on these 92
articles in 1911 formed the base for the 1912 index. In group 3
the mass unit for hay was changed from 270 to 135, and oaten chaff,
a new article in this group, was assigned a mass unit of 135, thus
making the sum of the mass units used for hay and oaten chaff in
1912 equal the mass unit for hay in the earlier years.
INTERPOLATION.
As already stated, prices for meats were not secured prior to 1890
except for 1884 and 1885. For the full period since 1871 the index
numbers were accordingly worked out for the seven groups, excluding
meats, and also for the period since 1890 for the eight groups, including
meats. The figures for the general index for 1871 to 1889 (except
1884 and 1885) were then adjusted, on the basis of the results for
succeeding years, so as to include meats. The exact procedure has
not been disclosed.
WEIGHTING.
The system of weighting used differs materially from the system
generally employed by compilers of index numbers. The author
bases his index numbers on what he terms the “ aggregate expendi
ture method. ” By this method the cost of an unvarying bill of
goods is calculated at the varying prices prevailing during differ
ent years. The extent to which a commodity is used is expressed
by a number termed the “ mass unit.” The mass unit is devel
oped from the figures which denote the quantity used or consumed,
which latter amount has in general been obtained by adding to
the production of each commodity in Australia the amount of im
ports and from this sum subtracting the amount of exports. The
figures have, in general, been based on the average production and
the average export and import returns for the five years 1906 to
1910, inclusive. No further explanation as to the source of his con
sumption figures has been given by the author. The mass unit is
obtained in each case by dividing the figure denoting the quantity
consumed by 10 and approximating the quotient. For instance,
the average annual consumption of pig iron is stated to be 64 thousand
tons, which is restated as a mass unit of 6 J.
AUSTRALIA---- BUREAU OF CENSUS AND STATISTICS.
181
The mass unit having been established and the average yearly
price for the year determined, the process then was to multiply the
mass unit by the price. Thus, the price of pig iron was 81s. 2 d. ($19.48)
for a ton in 1911. This price, 81s. 2 d. ($19.48), multiplied by the mass
unit (6 £) gives the aggregate expenditure on pig iron in 1911. This
process was applied to each article of the metals group in 1911 and
each of the years during the entire period (1890-1912), the sum of
such products producing the yearly aggregate group expenditure.
The same mass unit was used for all the years of the period. The
index for a single group and for all groups was obtained by dividing
the total expenditures of a given year by the total expenditures of
the basic year, i. e., 1911, and the result multiplied by 1,000.
Table 26 illustrates the system of weighted index numbers used:
T able
2 6 .—COMPUTATION OF IN D E X NUM BERS—IL L U STR A T IV E E X A M P L E OF AG
GREG ATE E X P E N D IT U R E S M ETH OD.
Article.
Butter...........................................
Bread .........................................
Mutton.........................................
Milk
........................................
Total..................................
Quantities
consumed
(in mil
lions).
Unit.
90
470
330
300
1911
1901
a.
Pound..............
2-pound loaf...
Pound.............
Quart...............
Total expenditures
(in millions).
Prices.
15
3
3
4
a.
1901
18
4
5
5
d.
1,350
1,410
990
1,200
4,950
!
1911
d.
1,620
1,880
1,650
1,530
G, 650
Thus 6,650 millions is the total expenditure for this group in 1911,
which is the base or 1 ,000. To secure the index figure for 1901, the
total aggregate number 4,950 millions for 1901 is divided by 6,650
millions, the base, which quotient multiplied by 1;000 equals 744 as
the index for 1901. A similar process was used for each of the
groups represented in the report. The author lays particular em
phasis on the fact that his index numbers are reversible, by which
he means that they may be recomputed with any other year than
1911 as the base and the results be as accurate as if that year had
been taken as the base originally.
TESTING.
The author tests his system of weighting by mass units instead of
actual quantities consumed by a comparison of results obtained under
the two methods by taking a list of prices from 1871 to 1911 for 73
articles, the year 1911 being used as a base. In the first instance the
actual figures w^ere used and in the second the mass units. The
result in the first case was 1,194 and in the second 1,193, the slight
difference thus shown appearing not to warrant the extra arithmetical
labor required by the use of the actual figures instead of the rounded
numbers or mass units.
182
INDEX NUM BERS OF WHOLESALE PRICES.
He further tests his index numbers by a comparison with index
numbers obtained by other methods. Table 27 illustrates this
comparison:
T able
2 7 .—PRICE IN D E X E S :
E X A M IN A T IO N AS TO R E L IA B IL IT Y OF V AR IO U S
M ETH O DS.
Weighted Weighted
Un
price-ratio price-ratio weighted
method
method
(geometric (arithmetic price-ratio
method.
mean).
mean).
Aggregate
expendi
ture
method.
Index number for 1871, with 1911 (equaling 1,000) as
base year.................................................................................
1,194
1,195
1,289
1,310
He considers the first two as valid but the last two as invalid for
his use.
TABLES OF RESULTS.
Table 28 contains the index numbers for each group and for all
the groups as a whole for specified years from 1861 to 1919, inclusive.39
No index numbers for meats are given in the original sources for
years before 1884 or for the years 1886 to 1889.
T a b l e 2 8 . — M ELB OU RNE
W H O LESALE-PRICE IN D E X NUM BERS FOR SPECIFIED Y E A R S ,
1861 TO 1919.
(Average prices in 1911=1000.)
Year.
1861............
1871............
1881............ :
1891............
1901............
1911............
1912. . . . . . .
1913............
1914............
1915............
1916............
1917............
1918............
1919............
Agricul
Metals Textiles, tural
pro Dairy Groceries.
and coal. leather,
etc.
duce, etc. produce.
1,438
1,096
1,178
895
1,061
1,060
1,021
1,046
1,099
1, 284
1,695
2,129
2,416
2,125
1,381
1,257
1,115
847
774
1,000
991
1,070
1,032
1,017
1,423
2,008
2,360
2,363
1, 583
1,236
1, 012
1,024
928
1,000
1, 370
1,097
1,207
2,162
1,208
1,157
1,444
1, 985
1,008
864
935
995
1,029
1, 000
1,206
1,054
1,137
1,530
1, 485
1,423
1, 454
1,651
1,963
1, 586
1,421
1,032
1,048
1,000
1,052
1,024
1,021
1,133
1,322
1, 343
1,422
1, 516
Meat.
888
1, 345
1,000
1, 357
1,252
1,507
2,435
2,515
2,403
2,385
2, 348
Building
material.
1,070
1,044
1,091
780
841
1,000
1,057
1,128
1,081
1, 275
1,491
1,884
2,686
2,851
Chemi
cals.
2,030
1,409
1, 587
1,194
917
1,000
978
995
1,253
1, 528
1,760
2,171
3,225
2,898
All com
modities.
1,538
1,229
1,121
945
974
1,000
1,170
1,088
1,149
1,604
1,504
1,662
1,934
2,055
The variations in the index numbers of the separate commodity
groups for the years 1915 to 1918, and for each month from January,
1918, to March, 1919, are shown in Table 29, taking July, 1914, the
last month before the outbreak of the war, as the base (or 1 ,000) for
each group.
39 From Commonwealth Bureau of Census and Statistics, Labor and Industrial Branch, Report No.
10, p. 68.
AUSTRALIA— BUREAU OE CENSUS AND STATISTICS.
183
TAEiiE *>9.—V.El-BOUR NE W H O L E SALE-PRICE IN D E X NUM BERS, B Y Y E A R S , 1915 TO
1918, AN D B Y M ONTHS, JAN U A R Y, 1918, TO JUNE, 1920.
[Commonwealth Bureau of Census and Statistics, Labor and Industrial Branch, Report No. 9, July,
1919, p. 75, and No. 10, October, 1920, p. 69.]
(Average prices of July, 1914== 1000.)
Year
and
month.
July, 1914..
1915............
1916............
1917............
1918...........
Jan----F e b ...
M a r.. .
A p r .. .
M ay. . .
June...
J u ly ...
A u g. . .
Sept...
O c t....
N o v ...
D e c .. .
1919:
Jan___
F e b ...
Mar. . .
Apr. . .
May. . .
June...
July...;
A u g ...
Sept...
Oct___
N o v ...
Dec. . .
1920:
Jan___
F e b ...
M ar.. .
A p r .. .
M ay...
June...
Agricul
Metals Textiles, tural pro Dairy Groceries.
and coal. leather,
duce, etc. produce.
etc.
1, 000
1,168
1, 539
1, 919
2,197
2,132
2,157
2,161
2,161
2,192
2,195
2,206
2,225
2,232
2,234
2,229
2,237
1,000
934
1,307
1, 841
2,324
2,432
2,365
2,463
2,430
2,309
2,331
2,249
2,392
2,410
2,310
2,074
2,122
1, 000
2, 024
1,130
1,084
1, 351
1,185
1,191
1,232
1,261
1, 382
1, 359
1, 356
1,358
1, 357
1, 417
1,573
1, 539
1,000
1,272
1,235
1,181
1,210
1,144
1,155
1,188
1,209
1,288
1,272
1,250
1,217
1,180
1,163
1,182
1,271
1,000
1,098
1,266
1,302
1,378
1,331
1,336
1,324
1,385
1,366
1,355
1,409
1,414
1,408
1, 402
1,394
1,397
2,178
2, 042
2,000
1,950
1,950
1,135
1,858
1,819
1, 816
1, 864
1,841
1,863
1,942
1,838
1,850
1,974
2,001
2,139
2,259
2,290
2,250
2,432
2,537
2,591
1,653
1,661
1,611
1,619
1,618
1,597
1, 684
1,903
2,002
2,356
2,375
2,242
1,263
1,290
1,355
1,361
1,402
1,401
1, 413
1,353
1, 376
1,412
1, 419
1, 423
1,889
1, 919
2,051
2,051
2,135
2,135
2,729
2,832
2,805
2,733
2,648
2,597
2,265
2,269
2,256
2,339
2,519
2,613
1, 432
1,490
1,615
1,690
1, 765
1, 870
Meat.
Building
materials.
Chemi
cals.
All com
modities.
1,000
1,164
1,361
1,722
2,448
2,046
2,069
2,107
2,156
2,596
2,593
2,636
2,656
2,616
2,616
2,624
2,663
1,000
1,490
1, 716
2,141
3,085
2,685
3,180
3,275
3,275
.3,001
3,207
3,462
3,363
3,271
3,188
2, 995
2,916
1, 000
1,406
1, 318
1,450
1,695
1,635
1,633
1,668
1, 677
1, 711
1, 709
1,700
1,721
1, 720
1,727
1, 716
1, 722
1,415
1,407
1,415
1,423
1,433
1,438
1, 477
1,500
1, 494
1, 524
1, 513
1, 564
1,000
1,502
1, 551
1, 480
1,469
1,540
1,532
1, 556
1,517
1,452
1,469
1,463
1,408
1,426
1, 490
1,403
1,370
*
1,370
1,316
1, 465
1,559
1,483
1, 510
1,475
1,452
1,522
1,536
1,324
1, 321
2,652
2,637
2,606
2,476
2,568
2,500
2,431
2,486
2,588
2,711
2,782
2,807
3,126
3,139
3,089
2,900
2,681
2,729
2,771
2,771
2,633
2,716
2,674
2,662
1, 718
1,678
1,688
1, 711
1,715
1,730
1,760
1, 815
1,850
1,997
1,987
1,972
1, 555
1,608
1,602
1, 924
1, 965
1,951
1,466
1,486
1,456
1,600
1,695
2,075
2, 820
2,867
2 ,977
2,977
3,074
3,074
2,678
2,717
2,797
2,797
2,972
2,972
2, 026
2,063
2 ,089
2,173
2,250
2,330
In the initial report issued in 1912 the author presents the follow
ing table of index numbers by quinquennial periods to show the aver
age level of prices over periods of several years. The average for
each 5-year period is 1,000 and that for 1911 and 1912 is compared
with this base. For instance, taking the average prive level of
1871-1875 as 1,000, that for 1911-1912 has fallen to 806, and taking
that of 1876-1880 as 1,000 that for 1911-1912 in comparison is 877.
The other figures are to be read in the same manner.
T
able
3 0 .—IN D E X NUM BERS FOR 1911-12, W IT H A V E R A G E E X P E N D IT U R E IN EACH
SUCCESSIVE Q U IN Q U E N N IA L PERIO D AS B ASE.
Base period
(prices=
1,000).
1871-1875. .
1876-1880..
1881-1885..
1886-1890..
1891-1895..
1896-1900..
1901-1905..
1906-1910..
I.
Metals
and
coal.
672
746
821
854
1, 225
1,134
1,098
lj 013
II.
Jute,
leather,
etc.
792
926
1, 012
1,172
1, 340
1, 383
1,224
1, 059
III.
IV .
Agri
V.
Dairy
cultural produce.
Groceries.
produce.
841
895
946
934
1, 345
1, 265
1,137
1, 094
1,037
957
957
974
1, 299
1, 249
1, 089
1,039
678
731
775
928
1, 015
1,028
1,089
1,082
VI.
Meat.
1,438
1,122
833
981
V II.
Building
materials.
889
1,0Q7
1, 071
1,193
3,362
1, 244
1, 214
1,084
VIII.
Chemi
cals.
621
641
651
730
917
1, 019
1,091
1,097
All
groups
together.
805
877
932
999
1, 288
1, 222
1,115
1,070
INDEX NUMBERS OF WHOLESALE PRICES.
184
A table contained in the first of the two reports compares index
numbers of wholesale prices in Australia with those of the United
Kingdom, Belgium, Germany, Italy, France, Canada, the United
States, and New Zealand. Complete data for all countries are
shown for the years 1890 to 1911, inclusive. In the cases of the
United Kingdom and the United States the comparison is extended
back to 1840. This comparative table does not show, of course,
the data as published originally in the different countries, but as
recomputed for each country on the base 1911 equals 1 ,000. The
last column of the table contains figures computed by weighting
the index number for each country by its relative population, thus
supplying what the author says may be termed the world’s index
number. Comparison of the Australian index numbers with those
for other countries is continued in somewhat different form in later
reports.
INDEX NUMBERS OF THE BUREAU OF STATISTICS OF NEW SOUTH
WALES.
The Bureau of Statistics of New South Wales is publishing a yearly
index number of prices of the principal articles of domestic produce
exported from New South Wales, calculated on the average f. o. b.
prices at Sidney. The average prices prevailing in 1901 were adopted
as the base, or 1 ,000.®
T
able
3 1 .—
IN D E X NUM BERS OF PRINCIPAL E X P O R TS FROM N E W SOUTH W A L E S , 1901
TO 1918.
[Official year book of New South Wales, 1915, p. 810, and 1918, p. 482.]
(Average prices in 1901=1,000.)
Year.
1901.............
1902.............
1903.............
1904.............
19Q5.............
1906.............
1907.............
1908.............
1909.............
Pastoral pro Metals: Sil
ducts: wool,
ver, lead,
All articles.;
tallow, hides, copper, tin.
leather, etc.
1, 000
1,096
1,125
1,112
1,192
1, 316
1, 354
1,122
1,137
1, 000
935
992
1,011
1,149
1, 432
1,461
1, 073
1,066
1,000
1, 065
1, 065
1, 071
1,150
1, 277
1,343
1,164
1,188
Year.
1910...........
1911...........
1912...........
1913...........
1914...........
1915...........
1916...........
1917...........
1918...........
Pastoral pro
ducts: Wool, Metals: Sil
All articles.
tallow, hides, ver, lead,
leather, etc. copper, tin.
1, 214
1,194
1,263
1, 408
1,451
1, 686
1, 988
2,212
2,333
1, 111
1,189
1, 454
1, 451
1, 302
1, 464
1, 948
2, 241
2,387
1, 205
1,194
1, 327
1,367
1, 365
1, 620
1, 878
2,116
2,207
a Since 1918, index numbers of wholesale prices in Sydney, including 100 commodities classified into
8 groups and based on prices in 1911 as 1,000, have been published by the Bureau of Statistics. See
Official Yearbook of New South Wales, 1919, pp. 482-489.
AUSTRIA— DR. BELA YON JAN KOVIC H .
185
AUSTRIA.
INDEX NUMBERS OF DR. BELA VON JANKOVICH.40
PUBLICATION AND HISTORY.
This index for the years 1867-1897 appeared under the title “ Die
Fluktuation der Waarenpreise im Grosshandel und die Schwankungen der Wechselkurse der oesterreich-ungarisehen Papiervaluta in
den Jahren 1867-1897” (The fluctuation of wholesale prices and
the variation in the rate of exchange of the Austro-Hugarian paper
values). It was printed in the Hungarian economic review entitled
“ Kozgazdasagi Szemle 1899.”
It was continued to 1909 and reprinted in the Bulletin of the
International Institute of Statistics, volume 19, Part III, page 136
et seq., in 1912 (.Bulletin de L ’ Institut International de Statistique) ,
under the title “ Index Nummer von 45 Waaren in der oesterreichungarischen Monarchie, 1867-1909; System Sauerbeck, zum Teil vom
Verfasser korrigiert” (Index number of forty-five articles in the
Austro-Hungarian monarchy, 1867-1909, according to the system
of Sauerbeck, with some corrections by the author).
Sauerbeck’s method of computation was followed as closely as
possible in order that the Austrian index might be comparable
with Sauerbeck’s index, since England throughout the period had
a gold standard. Articles were also selected to correspond as closely
as practicable with those entering into Sauerbeck’s index.
SOURCE OF QUOTATIONS.
The sources of the actual prices vary, being mostly official and
semiofficial publications of Austro-Hungarian cities and Provinces.
BASE PERIOD.
The years 1867-1877 were used as a base for all articles except
petroleum, for which 1873-1877 constitutes the base, and flax, for
which 1874-1877 constitutes the base.
NUMBER AND CLASS OF COMMODITIES.
The index includes 45 articles. Actual as well as relative prices
are shown for all articles except tea, but in a few cases the actual
prices are not complete. Relatives are also shown for each of six
groups into which tne 45 articles are separated, and for the entire 45
taken together. The six groups are: Grains (1-8), animal products
(9-15), colonial goods (16-19), minerals (20-26), textiles, raw mate
rials (27-34), and miscellaneous (35-45).
40 Dr. von Jankovich was in 1911 professor of the theory of finance and credit in the University of Buda
pest (according to the Minerva Yearbook of the Learned World) and vice president of the Hungarian
Chamber of Deputies.
186
INDEX NUM BERS OF WHOLESALE PRICES.
TABLE OF RESULTS.
Table 32 summarizes the results of Dr. von Jankovich’s compila
tions :
T able
3 2 .—SUM MARY TABLE SHOW ING IN D E X NUM BERS FOR THE PRICES OF 45
ARTICLES IN THE W H O L E SA L E M AR K ETS OF AU ST R 1A -H U N G A R Y . 1
[Bulletin de L’Institut International de Statistique, vol. 19, Pt. I l l, p. 156.]
Year.
1867............
1868............
1869............
1870............
1871............
1872.'..........
1873............
1874............
1875............
1876............
1877............
1878............
1879............
1880............
1881............
1882............
1883............
1884............
1885............
1886............
1887............
1888............
1889............
1890............
1891............
1892............
1893............
1894.......... .
1895............
1890............
1897............
1898............
1899............
1900............
1901............
1902............
1903............
1904............
1905............
1906............
1907............
1908............
1909............
Grains
(1 - 8 ).
110
95
88
95
101
101
108
114
91
96
103
91
89
103
99
99
93
93
84
82
79
79
78
83
89
80
75
74
76
71
81
89
79
76
81
83
83
91
96
89
95
109
115
Animal
products
(9-15).
90
96
98
Colonial
goods
(16-19).
101
101
102
99
104
103
104
108
102
111
99
98
95
94
103
103
97
97
97
94
90
96
97
101
103
101
97
93
92
94
90
92
96
90
91
91
97
94
94
99
97
98
97
103
112
109
121
129
131
125
130
AH
foods
(1-19).
101
96
95
99
102
104
107
106
94
96
101
86
91
82
82
79
77
72
64
62
61
88
66
67
70
64
64
64
68
62
58
54
47
45
46
47
41
37
39
45
46
44
46
46
48
Index
Miscella
Mineral Textiles
(raw
neous raw All raw number
products materials) materials materials
for 45
(20-26).
(20-45). articles.
27-34).
35-45).
96
94
95
92
90
84
82
81
82
81
83
86
80
79
78
80
76
78
83
78
78
78
81
84
88
95
94
98
100
107
101
93
99
104
116
106
107
98
86
97
93
89
81
86
86
89
82
77
74
74
75
80
78
81
82
79
78
73
72
74
76
81
96
109
106
92
93
90
96
112
117
105
103
117
104
109
106
105
110
95
87
87
86
94
85
82
87
82
81
78
78
74
73
74
71
72
70
67
65
69
65
64
63
62
59
67
76
71
71
76
78
76
85
88
79
79
104
99
103
106
106
103
107
99
103
106
109
106
100
101
95
92
96
98
89
82
82
79
78
83
85
80
74
73
71
73
70
70
67
71
82
84
82
80
82
81
77
74
74
74
74
73
72
70
72
66
66
67
65
66
68
68
74
73
69
67
68
69
74
79
76
79
94
89
93
96
88
68
66
68
68
75
84
81
76
77
77
79
88
92
84
86
104
98
100
102
106
105
104
99
91
95
98
89
89
85
87
86
86
85
8°
77
77
77
77
77
78
74
75
72
72
71
72
75
76
82
80
78
80
82
85
91
95
91
94
1 Numbers appearing in the box headings refer to the column numbers of the articles making up the six
groups as they appear in the detailed table.
BELGIUM.
INDEX NUMBERS OF HECTOR DENIS.
HISTORY.
Hector Denis, professor at the University of Brussels, is believed to
be the author of the only series of index numbers of wholesale prices in
Belgium that has been presented with any measure of continuity and
completeness. This series has not, however, been published in uniform
manner, but has appeared as a gradual development in various pub
lications credited to its author. The only exception seems to be a
BELGIUM-----HECTOR PENIS.
187
table of index numbers computed for 28 articles of export, which was
given in 1911 in practically the same form in which it first appeared
in 1895.
PUBLICATION.
Among the publications of Prof. Denis in which his various tables
of indexes appear are the Economic and Social Depression and the
History of Prices (La depression economique et sociale et Vhistoire des
prix), 1895,41 and the Index Numbers of Moral Phenomena (.Les index
numbers— nombres indices— des phenomenes moraux), 1911.42
The most nearly complete examples of his indexes, however,
are published in the Bulletin of the International Institute of Statis
tics, volume 19, Part III, pages 157-19-5, and are computed to
include the years 1909, 1910, or 1911, as the case may be. Only
two of the earlier tables appear in the bulletin. Of these the index
numbers based on the 28 exports are continued to include the year
1910, as is likewise a comparative table that had appeared in his
Index Numbers of Moral Phenomena. This comparative table is
used by Prof. Denis to test his own general index for the 28 articles of
export of Belgium. He reduced the general index numbers of France
(Palgrave), Germany (Soetbeer), England (Sauerbeck), and the
United States (Falkner-Hardy) to the common base 1867-1877, and
presented them in parallel columns with his own for each year and for
five-year periods from 1860 to 1893.43 In the same publication the
table appears a second time, but with data for each year from 1850 to
1910,44 and without the reduction of the indexes to a common base.
A second table of index numbers of exports, similar to but not
identical with the one published in 1895 on the base period 1865,
also appears in the bulletin. Separate tables of index numbers are
also given for vegetable products, meats and butter, cereals, and
other groups of articles.
SOURCE OF QUOTATIONS.
Prof. Denis based his computations for the index numbers appear
ing in his study of the economic and social depression on the quota
tions given in the tables of foreign commerce of Belgium (tableaux du
commerce exterieur de la Belgique), i. e., on the prices used for fixing
the customs values. Of these Mr. Armand Julm 45 remarks that he
finds that the averages adopted by the commission for the official
values correspond but remotely with the real fluctuations in prices.
Therefore the accuracy of Prof. Denis’s index numbers may appear
rather doubtful. However, this criticism can apply only to the
single original table of 28 exports that is continued to include 1910,
41 In this appear four tables of indexes.
The first is an index for 28 exports yearly from 1850 to 1890, com
puted on the baseperiod 1867-1877. The second index covers the same period of years and almost the samo
articles, except that some are grouped and one or two are added, the base period being the single year 1865.
The third index is similar to the first, but is based on 2 2 imports. The fourth table presents the general
index number for imports and also for exports for every year from 1865 to 1890, the period 1867-1877 again
being used as the base.
42 Published by the Royal Academy of Belgium in its Memoires, second series, 4, Brussels, 1908-1911.
Separate tables of price index numbers are given for cereals, beef, wheat, coal, and metals, all computed on
the base period 1857-1877. A comparative table shows index numbers for the United States, Belgium,
Germany, and England.
43 This table is given on p. 189 of this bulletin.
44 The general index computed for Belgium by Prof. Denis, which appears in this table, is printed on p.
189 of this bulletin for the years 1891-1910.
^ “ The economic progress of Belgium from 1880 to 1908/’ in the Journal of the Royal Statistical Society,
1911, p. 268.
188
INDEX NUMBERS OF WHOLESALE PRICES.
since in the additional tables published in the bulletin of the Inter
national Institute of Statistics referred to above he states in a foot
note that prices up to 1852 were taken from the secular almanac of
the observatory (VAlmanack seculaire de VObservatoire) and those
for later years from the statistical yearbooks (Annuaires statistiques).
BASE PERIOD.
For the base period Prof. Denis selected the years 1867-1877, as
did Sauerbeck for England. His reason for such a choice appears to
be the fact that during those years there were periods of both rise and
depression in prices, while a later period would not include the eco
nomic depression that followed 1873 and an earlier one would cover a
time of rising prices only.
DESCRIPTION OF COM M ODITIES.
The articles for which index numbers have been computed are not
described specifically in the publications mentioned above, but it is
stated in the one dealing with the economic and social depression
that the 28 exports (whose indexes are continued to 1910) and the
22 imports (not computed after 1890) were selected with a view to
including those most prominent, and that the total of 50 articles so
selected comprises two-thirds of the total exports and imports of
Belgium.
WEIGHTING.
The indexes are not weighted, and in his Economic and Social De
pression and the History of Prices Prof. Denis justifies himself by
stating that in spite of the use of weighted averages the variation in
the curve of prices as shown in the diagrams accompanying that
publication remains essentially the same and that therefore he has
not abandoned the simpler method, but presents his indexes with the
caution that the results are to be taken only as an approximation of
the truth.
TABLES OF RESULTS.
Below follows the comparative table of general index numbers for
several countries as it appears in his treatise on the Index Numbers
of Moral Phenomena mentioned above.
189
BELGIUM---- HECTOR DENIS.
3 3 .—COMPARISON OF GENERAL IN D E X NUMBERS FOR FRANCE (PALG R AVE),
GERM ANY (SOETBEER ), ENGLAND (SAUERBECK), UNITED STATES (F A L K N E R H A R D Y ), AND BELGIUM (DENIS), REDUCED TO A COMMON BASE.
T able
(Average prices in 1867-1877=100.)
A continuation of the general index numbers for Belgium as com
puted by Prof. Denis in the above table appears on page 158 of volume
19, Part III, of the Bulletin of the International Institute of Statis
tics, as follows:
Index
numbers,
Year.
189
189
189
189
189
189
4
5
6
7
8
9
59. 0
61. 5
6 1 .0
5 6 .0
5 8 .5
6 1 .8
j
Year.
!
190
190
190
190
190
190
Index
numbers.
0
1
2
3
4
5
6 3 .4
63. 1
6 4 .0
60. 6
61. 8
60. 9
Index
numbers.
Year.
190
190
190
190
191
6
7
8
9
0
64. 8
6 9 .2
6 7.7
6 5 .9
6 4.7
CANADA.
INDEX NUMBERS OF THE DEPARTMENT OF LABOR.
PUBLICATION.
This compilation of wholesale prices for the Dominion of Canada is
published yearly by the Department of Labor of Canada at Ottawa.
Index numbers for each group of commodities and for its main sub
divisions are also published monthly in the Labor Gazette, the official
organ of the department of labor.
190
INDEX NUM BERS OF WHOLESALE PRICES.
HISTORY.
The first report on wholesale prices made by the Canadian Depart
ment of Labor was published in 1910 and covered the years 1890 to
1909, inclusive. The object in undertaking this work was to deter
mine as accurately as possible the nature and extent of the general
rise in wholesale prices which had occurred in Canada during recent
years. Prior to the beginning of this work the Labor Gazette, the
official organ of the department, had for some time published each
month certain data regarding prices in connection with its review of
industrial and labor conditions. The importance of the subject and
the unsatisfactoriness of general statements in a matter of this kind
led the department in 1910 to adopt a more comprehensive and sys
tematic method of treating the subject of prices in the monthly sum
mary and also to extend the inquiry into the wholesale prices of a
selected list of representative staple commodities back over the pre
ceding 20 years.46 In subsequent annual and monthly reports the
price data have been brought down to the present time.
SOURCE OF QUOTATIONS.
It is stated that the practice followed throughout the investigation
was “ to collect and collate the best available published information
and to submit the result for verification to long-established firms at
the wholesale center in question.” The daily press and weekly trade
journals of Canada and the printed reports of exchanges, boards of
trade, etc., are mentioned as the principal sources of data. When
reliable printed matter failed, information was obtained from books
of manufacturers and wholesalers.
A source used for verification purposes in the case of a few impor
tant raw materials imported by manufacturers direct from the primary
markets of the world, and in which there is no wholesale trade in
Canada, was the declared import values, which were divided by total
quantities to show the average prices. Toronto and Montreal mar
kets furnish the great mass of the quotations published in the
reports .47
BASE PERIOD.
The base period selected for the computation of index numbers for
practically all commodities is the decade 1890-1899. Two reasons
are given for this selection: ( 1 ) The period was considered as repre
sentative of normal conditions as any available, containing a time of
falling and a time of rising prices, and (2 ) direct comparison with the
similar study of the United States Department of Labor was con
sidered very desirable, and this was made possible by choosing the
same base period .48 In a few instances, owing to special reasons, a
period other than the decade 1890-1899 has been chosen as the base.
46 Wholesale Prices in Canada, 1890-1909, p. 2.
47 Idem, p. 9.
48 Wholesale Prices in Canada, 1890-1909, p. 440. The base period used by the United States Department
of Labor (Bureau of Labor Statistics) has since been changed.
CANADA---- DEPARTMENT OF LABOR.
191
PRICES: H O W COMPUTED AND SH O W N .
In the first report the prices quoted are stated to be “ for the most
part those prevailing on the opening day of each month, though if,
in particular cases, these were found to be abnormal, an average of
the week was taken.” 49 In the report for 1912 it is stated that the
manner of quoting prices is the same as in the earlier reports except
that for certain articles subject to rapid fluctuations (grains, live
animals, certain meats, butter, eggs, potatoes, and fresh fruits— 40
in all) weekly instead of monthly quotations were obtained .50 This
plan was continued in the preparation of the wholesale-price data for
1913 and subsequent years.
Difficulty was encountered in obtaining quotations of a uniform
quality of certain articles, particularly of manufactured articles,
through a series of years. It is stated in the reports that wherever
such articles are quoted care has been taken to see that changes in
quality are accounted for in the prices given .51 In a few cases— as,
for example, in the case of cotton goods— the prices published are not
simple quotations on a single variety, but averages of a large number
of varieties.
In the annual reports the actual prices are published for each com
modity by months, or, in some cases, by weeks, and the average of
these quotations is given as the price for the year. Index numbers
are published in the annual reports for each commodity by years and
in the Labor Gazette for each group and subgroup (56 items in all)
by months currently. Index numbers do not seem to be published
for single commodities by months. Many of the actual prices are
stated in the form of a range of price, and apparently the mean is
used for computations based on these figures.
Some commodities whose price is largely governed by seasonal
conditions are quoted for only those months of the year when they are
in season— as, for example, blue grapes, for which quotations are
given only for September and October.
NUMBER AND CLASS OF COMMODITIES.
The index number for the 20 -year period 1890-1909 is based on 230
commodities so-called, some of these quotations being, however, as in
the case of cotton fabrics, the average of a large number of varieties
of the articles. In the first annual report (covering the year 1910)
one quotation was dropped and six new quotations were added, thus
making the number of commodities 235. In the second annual report
(covering the year 1911) one quotation was dropped, but the entire
number covered by the index was increased to 261. In the latter
report the statement is made that it is hoped ultimately to include
about 280 commodities. The total was increased to 272 in 1912 by
the addition of nine new articles and additional series of quotations
in the case of two articles previously included. No change in the list
was made in 1913. The new articles have been included in the index
numbers since 1910, so as to assist immediate comparisons, but no
recalculation of the entire series of index numbers back to 1890 is to
49 Wholesale Prices in Canada, 1890-1909, p. 439.
50 Idem, 1912, p. 2.
si Idem, 1890-1909, p. 439.
192
INDEX NUMBERS 0E WHOLESALE PRICES.
be made on the enlarged basis until the number of commodites is
completed .53
In recent reports, prices for a number of articles which it is hoped
ultimately to include in the index number are published in connec
tion with prices for the 271 commodities included at present. Both
raw materials and manufactured articles are included in the com
modities used in computing the index number. Difficulties attend
ing the employment of manufactured articles were recognized, but
their inclusion on a conservative basis seemed imperative in selecting
a sufficiently large number of representative commodities. With
respect to the original number, 230, it was said that “ the effect of
tendencies incidental to the manufacturing processes are present in
about 40 per cent of the quotations.” 53
DESCRIPTION AND GROUPING OF COMMODITIES.
The commodities for which index numbers have been computed
are shown in the following list, which is taken from the report for 1913
(pp. 218-240) and the Eighth Annual Report of Wholesale Prices in
Canada, 1917 (pp. 87-92).
I. Grains and fodder.
Oats, No. 2, white, Ontario.
Peas, No. 2, Ontario.
R ye, N o. 2, Ontario.
Shorts.
Straw.
W heat, N o. 1, northern, Manitoba.
W heat, No. 2, w hite,'O ntario.
Barley, western.
Barley, N o. 2, Ontario.
Bran.
Corn, N o. 3, yellow .
Flaxseed, N o. 1, northwestern.
H a y , Montreal.
H a y, Toronto.
Oats, No. 2, white, western.
II. Animals and meats.
Bacon, English boneless breakfast.
B eef, plate.
Beef, dressed, hind quarters.
Beef, dressed, forequarters.
Cattle, western prime.
Cattle, choice steers, Toronto.
Fowls.
Hams, city cured, m edium .
Hogs, select, live.
Hogs, dressed.
Lard, pure.
M utton, dressed.
Lam b.
Pork, Canada, heavy short-cut mess.
Sheep, export ewes.
Turkeys.
V eal, dressed.
Ill* Dairy products.
Butter, creamery, Montreal.
Butter, creamery solids, Toronto.
Butter, dairy, prints, Toronto.
Cheese, western colored.
Eggs, fresh, Montreal.
Eggs,
M ilk ,
M ilk,
M ilk,
storage, Toronto.
at Montreal.
at Toronto.
at Victoria, B . C.
IV. Fish.
Codfish, dry, f. o. b.
H addock, d ry, f. o. b.
H a lib u t, fresh.
Herring, salted.
Lobster, fresh.
Wholesale Prices in Canada, 1912, p. 2.
53 Idem, 1890-1909, p. 438.
52
Lobster, canned.
Mackerel, salted.
Salmon, B . C . , canned.
Salmon trout, fresh.
W hitefish, fresh.
CANADA---- DEPARTMENT OF LABOR.
193
V . Other foods.
(a) Fruits and vegetables.
1. Fresh fruits.
N ative—
Apples, good seasonable.
Cherries.
Grapes, blue.
Peaches, Leno covers, No. 1
fruit.
Pears, early, Bartletts and
winter.
Plums, early, Lombards., green
gages.
Raspberries, red.
Strawberries.
Foreign—
Bananas, yellow.
Lemons, Messinas and V erdellis.
Oranges, navels and Valencias.
2. Dried fruits.
A pples, evaporated.
Currants, Patras.
Prunes, Bosnia.
Raisins, Sultanas.
3. Fresh vegetables.
Beans, hand-picked.
Onions, Canadian R ed.
Potatoes, Montreal.
Potatoes, Toronto.
Turnips.
Tomatoes.
4. Canned vegetables.
Com , standards, 2 ’s.
Peas, standards, 2 ’s.
Tomatoes, 3 ’s.
(6) Miscellaneous groceries and provisions.
1. Breadstuffs.
Flour, straight rollers.
Flour, strong bakers.
Flour, winter w heat patents.
Flour, M anitoba 1st patents.
Bread, Toronto.
Bread, Victoria, B . C .
Biscuits, soda.
Oatmeal, standard.
R ice, Patna.
Tapioca, m edium pearl.
2 . Tea, coffee, and chocolate.
Chocolate, D iam ond.
Coffee, R io, No. 7.
Coffee, Santos.
Tea, good common Japan.
3. Sugar, etc.
Glucose.
H oney, strained.
M aple sugar.
Molasses, New Orleans.
Sugar, Montreal granulated.
Sugar, M ontreal yellow.
4. Condim ents, etc.
Pepper, black , pure.
Cream of tartar.
Salt, fine, dairy, cheese, and table.
Soda, bicarbonate of.
Vinegar, w hite wine, proof strength.
V I. Textiles.
(a) W oolens:
(d) F la x products:
W ool (Ontario), washed.
W ool (Ontario), unwashed.
Y arn.
W oolen underwear.
Beaver cloth.
( b) Cotton:
Cotton, upland m iddlin g.
G ray cottons.
W oven colored fabrics.
Prints.
;:•) S ilk :
Silk, raw, Italian, classical.
S ilk , raw, Japan.
S ilk, spool.
S ilk, machine.
F lax sewing twine.
L inen rope, w hite.
F lax fiber.
Tow, fine.
( e) Ju te:
Jute, first marks.
Hessian, 10b ounce, 40-inch.
( / ) O il cloths:
Floor oilcloth, No. 3 quality.
Table oilcloth, assorted patterns
(5 -4 w ide).
3 3 2 2 6 °— 21— B u ll. 2 8 4 -
-1 3
XjSTDEX LU M BERS OF WHOLESALE PfilCES.
194
VII. Hides and tallow, leathers, and boots and shoes.
Hides and tallow :
No. 1 beef hides.
No. 1 green calfskins.
Horsehides.
Tallow rendered. N o. 1 stock, in
barrel.
Leather:
No. 1 Spanish sole, for jobbing.
No. 1 slaughter sole, heavy.
Leather— C oncluded.
Harness, No. 1, U . O .
H e a v y upper.
Boots and shoes:
M e n ’s split blucher b a l., pegged.
M e n ’s box calf blucher b a l., G. W .
W o m e n ’s Dongola blucher b a l.,
F . S.
VIZI. Metals and implements.
(a) Metals— Concluded.
(a) M etals:
A n tim on y.
Copper.
Iron, pig, Summerlee.
Iron, p ig, N b. 1 foundry, N . S.
Iron, common bar.
Iron, black sheets.
Iron, galvanized sheets.
Iron, tin -p late, charcoal.
Iron, tin -p la te, coke.
Iron, boiler-plate.
W rought iron, N o. 1.
L ead , im ported.
L ead, dom estic, Trail.
N ick el.
Q uicksilver.
Silver.
Solder.
Spelter.
Steel billets.
Steel, cast.
T in, ingots.
Z inc, sheets.
( b) Im plem ents:
Anvils, W rights’ , 80 pounds and
over.
A xes, standard.
Chains, coil.
Crowbars.
Grindstones, 40-200 pounds.
Hammers, sledge.
Horseshoes.
Mallets, carpenters* hickory.
Screws, bench wood.
Soldering irons.
Vises, Wrights’ .
EE. Fuel and lighting.
Coal, N . S ., run of mines.
Coal, Crow’s Nest Pass.
Coal, Pennsylvania anthracite.
Coke, Connellsviile furnace.
Coke, Crow’s Nest Pass.
Gasoline.
Coal oil, prime white.
Coal oil, water white.
Calcium carbide.
Matches.
X . Building materials.
(a) Lumber.
Pine, all grades, Ottawa.
Pine, good sidings, Ottawa.
Pine, No. 1 cuts, Toronto.
Laths, pine, Ottawa.
Pine, shipping, culls, Ottawa,
Pine, box: boards, Ottawa.
H em lock, Ottawa.
Spruce, N ew Brunswick.
Shingles, N ew Brunswick.
Birch, Toronto.
Maple, Toronto.
Oak, Toronto.
British Colum bia fir.
British Columbia shingles.
(6) Miscellaneous building materials.
Brick, fire.
Brick, common building.
Cement, Canadian Portland.
H inges.
Iron pipe.
Lead pipe.
L im e.
N ails, cut.
Nails, wire.
P itch.
Plaster of Paris.
R ed lead, dry.
Sash cord.
Sash weights.
Soil pipe, m edium .
Tar.
W ire, copper.
W ire, iron.
W ire cloth.
W ire fencing.
CAM ADA---- DEPARTMENT OF LABOR.
1%
(c) Paints , oilt. and glass.
Rosin, white.
Shellac.
Turpentine.
Varnish,
Venetian red (dry color).
W h ite lead..
W indow glass.
Benzine.
Glue.
Linseed oil (raw);
Linseed oil (boiled).
Paris green.
Prepared paints.
P utty.
X I . House furnishings.
(a) Furniture:
K itchen chairs (common spindle).
K itchen
tables
(4-foot,
w ith
drawer).
D ining tables (hardwood, exten
sion, 8-foot).
Sideboards (hardwood, with mirror
16 b y 28).
Bedroom sets (dresser and stand,
hardwood).
Iron beds, continuous pillars, 4-foot
(b) Crockery and glassware:
Tumblers^ tank glass-, J-pint.
Earthenware: W hite
cups
and
saucers.
( b) Crockery and glassware— Concluded,
Earthenware: 10-piece printed toi
let sets.
Earthenware: 97-piece printed d in
ner sets.
(c) Table cutlery:
K nives, celluloid handle, m edium
size.
Silver-plated knives and forks, 6
pennyweight, m edium.
(d) K itchen furnishings:
Pails.
T ubs.
Brooms.
Sadirons*. Mrs-. Potts’ .
X II. Drugs and chemicals.
Copperas.
Glycerin.
Indigo.
M uriatic acid.
O pium .
Quinine.
Soda ash.
Sulphuric acid.
A lcohol, 65 Q . P.
A lcohol,, wood.
A lu m .
Bleaching powder.
Borax.
Brim stone.
Carbolic aeid.
Caustic soda.
X III. Miscellaneous.
(a) Furs.
M in k , dark.
M uskrat, best fall and winter.
Raccoon.
Skunk, black Canadian.
( b) Liquors and tobacco.
Hops.
M alt.
W h isky (Can. Club 2 0 -8 up ).
A le and porter (draft).
Tobacco, smoking.
Tobacco, raw leaf..
(<?) Sundries.
Binder twine.
Gunpowder.
Paper, news print.
Rope.
R ubber, Para Island;
Soap.
Starch.
SUBSTITUTIONS AND ADDITIONS.
In cases where new articles have been included in the index number
an effort was made to secure a series of quotations back to 1890. In
some cases this was not found possible, owing to “ poverty of records,
changes in industrial methods or consumption standards, etc.” It is
stated that no satisfactory solution of this problem has been found,
196
INDEX NUM BERS OF WHOLESALE PRICES
but that “ the method followed was to assign to the new commodity
the index number of the commodity displaced or most nearly repre
sented in the year in question. Thus the index number of the lowest
grade of pine lumber was assigned to hemlock in the year in which
that article first makes its appearance in the quotations. In this way
the new^ commodity creates a minimum of disturbance in the index
number of the year in which it first occurs, whilst subsequent varia
tions make themselves duly felt .” 54 In the case of calcium carbide, an
entirely new commodity introduced into the index in 1894, and the
case of Crow’s Nest Pass coal and Crow’s Nest Pass coke, both intro
duced in 1899, the price first quoted was taken as 100, or the base,
while in the case of cotton prints, introduced in 1893, the average
price for the years 1893-1899 was taken as the base.
INTERPOLATION.
So far as can be determined, no price quotations have been inter
polated. In the case of a few commodities, however, the statement
is made that, owing to incomplete records, associated data have been
drawn upon in calculating the base prices. Thus the price of flax
seed for the base period 1890-1899 was estimated from the price at
Chicago from 1890-1910, as published in the reports of the United
States Bureau of Labor Statistics, and . the price at Winnipeg from
1906-1910. The base price of plate beef was in like manner “ cal
culated from the percentages of cattle and beef prices from 1906 to
1911,” while the base price of dressed veal is “ based on the average
prices of other meat products, 1890-1899.” 55
W EIGHTING.
The general index number is the simple average or arithmetic
mean of the index numbers of the several commodities; i. e., the sum
of the relative prices of the different commodities, divided by the
number of commodities. Certain commodities are represented by
more than one quotation and, as would happen in any extensive list
including both raw materials and manufactured products, some com
modities are represented indirectly more than once, as, for example,
lumber, which is also represented by furniture. In the opinion of the
compiler “ an extended list of articles tends to weight itself.”
TESTING.
With the object of testing the results obtained by ,the use of the
simple arithmetical average of the index numbers representing the
several commodities, a weighted index has been computed. This is
based on the table prepared by the British Association for the A d
vancement of Science, a committee of which dealt exhaustively with
the whole subject of index numbers in 1887-1890. The table follows .56
Wholesale Prices in Canada, 1890-1909, p. 447.
1913, pp. 129, 130.
1890-1909, p. 442.
m Idem,
m Idem,
CANADA---- DEPARTMENT OF LABOR.
T able
197
3 4 .—BR ITISH ASSOCIATION FOR THE ADVANCEM EN T OF SCIENCE TABLE OF
W EIGH TS FOR THE CONSTRUCTION OF AN IN D E X NU M BER .
[The following explanation of the table is given in the Canadian report:
The second column gives in round numbers (000,000’s being omitted) the average national expenditure
(in pounds) on each class of article at present and for the last few years, and presumably also for the imme
diate future the proportion at least, if not the absolute amounts, of expenditure. In the estimated amount
of consumption allowance is made for the addition to the value made before the articles are in the form in.
which they are finally consumed.
In the third column these amounts (or proportions) are reduced to percentages (of the total amount
expended 011 such articles).
In the last column the relative importance proposed to be assigned to each article in the index number
is stated, mainly 011 the basis of the percentages in the third column, but with modifications so as to substi
tute even figures for the convenience of handling.]
Articles consumed or used up.
Relative
Estimated
importance
expendi
proposed
Percentage
ture per
for each
of each
annum on amount
to article in
each article
index
total.
(0 0 0 , 0 0 0
number
omitted).
reduced to
percentage.
Breadstuffs:
Wheat........................................................................................................
Barley.................. .....................................................................................
Oats............................................................................................................
Potatoes, rice, etc....................................................................................
~
ry
6 .5
£60
30
50
50
1
3.3
5.4
5.4
5
5
20
Meat and dairy food:
Meat....................................
............................
.
__
Fish............................................................................................................
Cheese............................... : ................ .......................................................
Butter........................................................................................................
Milk.
.....................................................................................................
100
20
10.9
1
10
2}
2 .2
60
n
6.5
20
Mass luxuries:
Sugar..........................................................................................................
Tea..............................................................................................................
Beer............................................................................................................
Spirits.........................................................................................................
Wine.........................................................................................................
Tobacco.....................................................................................................
30
20
100
n
2h
3.3
2 .2
1
40
10
9
10.9
4.3
f
1 1 .1
1 1 .1
10
2i
23
Clothing:
1
Cotton........................................................................................................ 1
W ool.......................................................................................................... !
Silk.............................................................................................................
Leather..........
...................................................................................
30
3.3
20
2 .2
1 .1
i
10
Metals and minerals:
Coal............................................................................................................. !
Iron............................................................................................................. !
Copper....................................................................................................... !
Lead, tin, zinc, etc.................................................................................. ;
*
Miscellaneous:
Timber.......................................................................................................
Petroleum.................................................................................................
Indigo......................................................................................................... 1
Flax and linseed...................................................................................... I
Palm oil.....................................................................................................
Caoutchouc........................... ...................................................................
20
100
50
25
25
2£
2 .2
24
24
24
I
1
10.9
5.4
2.7
2.7
10
10
5
24
24
20
30
5
^
3.3
*. 5
l. 5
3
10
1 .1
1 .5
3
5
5
1 .5
1
1
1
1
10
Total......................................................................................................
920
100.0 I
100
1 This percentage does not agree with that found in the Canadian report, but is correct according to the
expenditure given in the preceding column.
198
I jSTDEX LU M BER S OF WHOLESALE FRICES.
Slight modifications were made in the above table to meet the
groupings adopted in the Canadian investigation and the different
standards of consumption in that country. In the abscnce of statistics
directly bearing on consumption standards in Canada, apart from
statistics of import and export trade and of production, use was.
made of the special studies of family expenditures of the British,
United States, and Massachusetts labor departments.
The following figures show the weightmg used in the Canadian
study : 57
Group.
|
j
Weight,
Grains and fodder.............................. ...........
Animals and meats.........................................
Fish....................................................................
Dairy produce..................................................
Other foods................................... ..................
Textiles.............................................................
Hides, leather, boots and shoes...................
Metals and implements:
(a) Metals..................................................
( 6 ) Implements.......................................
Fuel and lighting............................................
Building materials:
(a) Lumber...............................................
(b) Miscellaneous building materials..
(c) Paints, etc...........................................
House furnishings...........................................
Drugs and chemicals......................................
Miscellaneous:
(a) Furs.....................................................
(6 ) Liquors and tobacco........................
(c) Sundry................................................
Total.......................................................
18
10
2*
7h
15
8
2
8
1
ia
6
2
1
2
2
1
2
2
100
The results of the testing may be seen in Table 35 which gives
both the weighted index numbers {or each year from 1890 to 1917,
inclusive,58 and the unweighted index number, from 1890 to 1919,
inclusive.
T able 3 5 .— W E IG H T E D A N D U N W E IG H T E D IN D E X N U M BER S, 1890 TO 1919.
Year.
1890................................
1891................................
1892................................
1893................................
1894................................
1895 ............................
2896................................
1897................................
1898................................
1899................................
1900 ............................
1901................................
1902................................
ig©3................................
1904................................
Weighted
number.
1 1 2 .0
111.3
104.9
103.9
97.2
95. 6
90.6
89.9
95.5
99.0
105.8
106.0
109.6
109.7
1 1 0 .6
Unweighted
number.
110.3
108.5
1 0 2 .8
1 0 2 .5
97.2
95.6
92.5
92.2
96.1
1 0 0 .1
108.2
107.0
109.fr
110.5
111.4
Weighted
number.
Year.
1905................................
1908................................
1907................................
1908................................
1909................................
1910................................
1912................................
1913................................
1914................................
1915................................
1916................................
1917................................
1918................................
1919................................
113.8
1 2 0 .1
129.2
125.1
126.3
128.0
131.1
143.9
139.6
139.1
154.2
182. a
241.4
Unweighted
number.
113.8
120. 0 1
126.2
1 2 0 .8
1 2 1 .8
124.2
127.4
134. 4
135. 5
136.1
148.0
182.0
237.0
278.3
293.2
TABLE OF RESULTS.
Table 36, reproduced from various reports, shows by groups of
commodities the index numbers for the 30 years, 1890-1919, inclusive.
57 W holesale Prices in Canada, 1890-1909, p . 12.
58 Idem , 1918, p . 11.
CANADA---- DEPARTMENT OF LABOR,
T able
199
3 6 — IN D E X NUM BER S, B Y GROUPS OF COMMODITIES, 1890 TO 1919.1
(Average prices in 1890-18^9=100.)
Group.
1.
Grains and fodder...............
1890
1891
1892
1893
1894 I 1895
1896
1897 | 1898
116.7
123 9
101.7
106 2
97 3
106,7
108.5
105.8
90.6
104. 7
^9.1
117.7
110.4
99.7
94.3
98. 7
104.6
96.4
95.0
97.3
89.9
98.8
92.2
94. 8
101.4
95. 2
93.6
98.6
85.2
82. 4
90.1
102. 6
87.1
96.9
92.9
80.6
90.4
90.1
98.6
87.5
98.5
98.9
85.7
93.1
96.4
1 1 1 .2
4. Fish........................................
103.0
103 3
120.3
111.4
1899
98. 8
97. 9
92.9
99.6
94. 3
95.2
105.0
93.6
99. 8
109.4
87.6
94.3
93.5
111. 9
98.0
96.9
90.8
87.4
95.8
97.2
107.6
96.7
95.1
101.4
1 1 0 .0
1 0 0 .6
191 3
104 2
109 6
125.4
103.8
107.4
114 4
103 2
106 7
107.6
102.9
106.6
1 0 2 .1
91.1
87.0
1 0 2 .6
1 0 2 .2
1 0 1 .0
102.9
97.5
97.0
103.5
117.6
109.5
109
110
104.4
106.8
98.2
100.9
104.4
103.7
103.7
98.6
104.4
104.6
98.7
95.5
101.3
103.1
1 0 2 .8
4
10H 8
100 5
110 3
95.2
96.1
97.9
100.3
97.1
93.9
96.2
97.5
99. 8
93.9
87.7
95.5
99.8
103.7
99.7
98.8
123.6
99.4
100.3
113.5
98.7
93.7
80.5
99.4
91.3
80.7
98.0
92.6
8 8 .0
1 1 1 .1
1 1 1 .8
1 1 2 .0
99 7
99.0
106.7
103.9
91.2
103.9
103.3
102.3
109.5
110.3
108.5
1 0 2 .8
102.5
97.2
95.6
•92.5
92.2
96.1
Group -
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1. Grains and fodder___ _____
99.9
103.4
109.0
106.4
96.4
116.1
106.5
117.9
108.9
116.2
98.1
105.9
115.7
115.5
111.3
107.2
119.5
110.4
113.6
116.4
120.7
115.1
115. 7
100.7
114.6
119.6
118.5
130.1
103.1
123.4
128.1
140.2
133.8
131.5
129.5
112.5
126.1
125.5
148.3
129.6
136.3
120. 5
110.3
113.8
107.3
111 3
109 5
113 2
98.6
103.6
112. 8
149.9
148.6
133.6
134.0
107.6
108.3
135. 4
1 2 1 .2
110.4
1 0 2 .8
1 0 0 .1
1 0 0 .8
1 0 2 .2
104.7
104.9
105.5
105.7
99.7
106.2
103.0
108.4
106.1
104.1
128.6
106.0
106.4
134.8
107.1
108.8
106.3
104.2
1 1 1 .0
114.0
1 0 2 .2
128.8
107.7
126.3
109.6
105.5
131.3
107.2
122.4
112.7
109.6
134.1
106.8
125.3
107.3
106.4
152.7
104.7
135.3
113.0
106.3
165.2
108.7.
141.2
112.7
108.5
162.6
107.5
136.8
101.5
114.6
106.0
121.9
107.9
99.8
147.3
103.3
113:0
140.9
103.3
110.9
145.2
103.7
116.8
168.1
107.0
115.9
171.3
107.8
119.1
217.4
108.1
239.2
125.5
123.0
231.8
118.0
117.6
227.2
117.5
1 2 1 .1
229.2
108.1
120.9
Total............................... 108.2
107.0
109.0
110.5
111.4
113.8
1 2 0 .0
126.2
1 2 0 .8
1 2 1 .2
1911
1912
1913
1914
1915
1916
1917
31918
U919
316.2
354. 5
259. 4
247. 0
250. 5
356.9
276.5
310.9
357.8
296. 5
236,6
251.5
373.1
349.5
8
1 0 2 .2
1 0 2 .1
1 0 1 .2
99. 8
1 0 1 .8
8 6 .0
98.0
1 0 0 .1
. Metals and implements:
Im p le m e n ts .
__
9. Fuel and lighting................
10. Building materials:
Lumber.............................
Miscellaneous.
Paints, oils, glass.............
11. House furnishings...............
12. D r u g s q.nd r/hp.rrnr.ftls
_
13. Miscellaneous:
Furs...................................
Liquors and tobacco___
Sundries..... .........................
Total...............................
2.
3.
4.
5.
6 .
7.
8.
9.
10.
11.
12.
13.
A n im a ls a n d m e a t s . _.
Dairy products.....................
Fish...............I.......................
Other foods...........................
Textiles.................................
Hides, leather, boots.........
Metals and implements:
Metals................................
Implements....................
Fuel and lighting.................
Building materials:
Lumber.............................
Miscellaneous.............
Paints, oils, glass.............
House furnishings...............
Drugs and chemicals..........
Miscellaneous:
Furs...................................
Liquors and tobacco___
Sundries............................
1 0 0 .2
110.5
8 6 .5
94.9
1 0 0 .0
1 1 1 .8
125.9
1 1 0 .2
7
98.1
1 2 2 .2
106.9
1 1 0 .2
98.4
1 0 1 .0
118.2
1 2 2 .0
104.6
128.1
109.2
1 0 1 .1
1 0 1 .8
1 2 0 .2
1 2 1 .8
9 6 .5
1 0 0 .0
99.0
96.8
1 1 1 .0
1 2 0 .0
1 0 2 .2
1 1 2 .8
107.1
1 0 0 .2
93.3
1 0 0 .1
101.9
102. 4
103.8
154.6
105.7
135.2
110.4
103.9
1 2 1 .6
Group.
1910
Grains and fodder...............
Animals and meats.............
Dairy products...... .............
Fish___ ..................................
Other foods...........................
Textiles..................................
Hides, leather, boots...........
Metals and implements:
Metals . . . .........................
Implements......................
9. Fuel and lighting.................
10. Building materials:
Lumber.............................
Miscellaneous...................
Paints, oils, glass.......
11. H o u s e f u r n is h in g s ..................
1 2 . Drugs and chemicals..........
13. Miscellaneous:
Furs...................................
Liquors and tobacco----Sundries............................
140.7
163.6
135.7
145.1
111.3
114.6
135.4
148.4 167.3
146.6 ‘ 160.8
136.2 159.0
143.6 155.7
118.7 126.0
119.2 120.7
139.6 152.4
136.8
180.8
154. 7
158. 0
117.4
130.8
163.9
156.5
192.3
154.4
156.0
118.8
133.5
171.8
186.9
187.2
161.4
149.7
125.5
149.2
180.5
195.2
217.7
183. 5
184. 8
156.2
193.4
233.4
281.5
288.1
230.5
205. 8
97.6
1015
103.0
108.3
104.5
100.5
117.4
104.7
113.3
119.1
105.6
118.2
113.9
103.8
110.9
152.4
198.9
135.2
132.6
259.1
181.6
193.0
273.1
241.4
108.8
2 2 1 .8
206.0
240.4
237.1
165.4
181.3
112.7
144.8
126.2
113.3
182.1
111.4
140. 7
129.5
175.7
115.9
157.1
136.5
181.3
182.1
154.9
200. 5
157.1
252.2
214.5
203.5
222.4
203. 7
267.9
267.1
223.4
310. 8
256.9
283.4
310.8
224.5
379.0
321.2
232.0
1.
2.
3.
4.
5.
6.
7.
8.
158.5
109.2
145.5
1 1 2 .1
2 2 0 .6
263.4
275.1
109.5
1 1 2 .1
166.5
105.4
148.6
114.5
115.5
234.5
132.9
118.0
252.9
151.2
100.3
297.3 307.9
155.2 134.7
104.3 i 113.1
205.4 161.9
136.9 135.6
108.5 j 116.6
299. 8
142.4
143.0
411.6
167.8
186.8
602.2
216.6
219.5
1009.2
275.8
134.4
136.1
182.0
237.0
278.3
293.2
1 1 0 .6
1 0 2 .6
154.5
1 1 0 .4
Total............................... 124.2 ; 127. 4
135.5
1 2 1 .6
148.0
2 1 1 .6
1 230 commodities, 1890-1909; 272 commodities, 1910-1914; 271 commodities, 1915-1919, one Iin« oC spelter
having been dropped in 1915.
3 Preliminary figures.
T able
200
Table 37 showing the course of wholesale prices, classified into commodity groups, in Canada during the years 1914
to 1920, has been reproduced from the January, 1921, Labor Gazette, published by the Department of Labor of
Canada. The figures are in certain instances preliminary and are therefore to be regarded as subject to possible
revision.
3 7 .—IN D E X NUMBERS OF WHOLESALE PRICES B Y GROUPS OF COMMODITIES, 1914 TO 1920.
(Average prices, 1890-1899=100.)
1915
1916
1917
1918
1919
INDEX
1914
1920
Commodity group.
238.1
249.2
242.6
183.8
296.3
293.4
210.3
201.5
319.8
325.0
259.0
236.3
314.2
369.4
251.0
240.9
272.3
343.7
294.4
268.3
330.0
389.7
286.4
.223.3
368.4
350.0
352.3
245.1
377.7
350.7
333.1
244.7
384.0
356.5
317.9
240.6
399.0
359.9
302.6
239.2
412.6
371.8
292.2
286.6
413.9
378.8
282.0
248.2
401.1
378.9
299.6
244.4
372.4
366.2
305.5
241.3
348.1 313.9 275.3 261.1
363.4 348.4 331.0 320.8
311.1 318.7 322.1 340. 0
243.2 236.5
249.5
170.5
153.6
192.7
240.2
234.9
177.7
215.7
285.9
308.2
215.9
268.5
277.4
258.4
225.3
326.9
261.8
280.2
246.0
370.4
283.5
246.1
257.7
383.6
280.7
239.7
251.2
364.3
385.3
317.0
282.3
414.0
387.6
347.1
288.6
419.4
400.8
352.7
293.1
420.9
363.2
377.8
304.6
428.7
391.8
428.5
316.6
422.0
344.6
404.3
316.2
410.5
297.9
352.9
325.9
398.3
292.2
258.8
319.3
392.8
282.8
227.6 2 1 1 . 2
300.8 287. a
387.4 382.4
264. 1 255.8
242.5
271.3
357.5
250.7
226.1
256.5
328.6
231.8
191.9 2 1 0 . 6 274.4 259.7 283.8 242.8 197.6 227.4 236.7 1 250.2 254.2 253.2 247.0 248.7 249.0 246.1 242.2 231.9 2 1 1 . 8
136.7 161.4 194.5 199.0 223.8 241.4 238.4 248.4 243.7 250.3 250.3 250.3 251.7 255.4 256.8 273.4 273.2 273.2 273.2
125.6 180.9 229.7 188.1 242.4 246.8 229.8 251.1 254.4 254.7 289.3 304.1 330.1 348.0 352.7 349.9 349.2 319.3 317.6
189.6
178.4
227.5
179.5
258.5
222.3 238.6 273.0 279.9
215.3 209.4 227.7 233.8
261.7 263.2 312.6 339.5
2 1 2 . 8 222.4 250.9 298.1
263.1 289.2 293,1 272.3
304.6
218.5
404.9
325.0
2 2 1 .2
419.9
232.4
433.3
363.5
215.3
439.9
235. 8
447.9
363.4
214.2
485.0
250.5
471.1
364.5
219.5
485.0
256.7
473.5
371.8
227.2
533.0 533.9
250.6j 251.6
489. 8 i 472.9
389.2 389.2
230.2 233.0
512.0
267.7
467.4
389.2
247.1
516.2
273.3
443.3
383.2
247.1
494.5
273.3
437.3
387.6
245.4
480.5
273.8
415.3
390.2
238.6
480.5
271.5
395.4
390.2
232.5
457.0
260.8
382.7
390.2
228.1
292.3 399.5 396.7 511.5 583.1 ! 742.3 945.6 1785.3 1851.4 1851.4 1779.7 1072.1 900.3 915.9 8 6 8 . 2 8 6 8 . 2 8 6 8 . 2 459.4 364.6
136.7 161.3 164.1 2 0 2 . 2 222.9i 258.7 274.1 317.3 314.0: 316.3 316.3 316.8 320.8 315.1 315.1 307.8 303.7 303.7 298.0
142.1 165.8 194.5 2 0 0 . 8 218.9 207.4 214.2 248.1 2 1 2 . 0 205.3 207.5 . 2 1 2 . 2 216.5 215.8 215.2 214.4 215.7 208.9 210.3
All commodities................... 136.5 134.6 138.9 150.2 172.1 180.9 212.7 248.7 258.7 284.0 286.5 294.0 336.4 343.5 349.0 353.1 350.6 349.3 346.8 330.2 320.6 317.6 304.2 290.5
i
PRICES,
182.2
153.6
197.4
152.6
249.0
WHOLESALE
179.0
231.7
160.5
155.9
OF
Grains and fodder................ 140.9 150.4 191.7 191.2 181.0
Animals and meats.............. 191.2 195.7 177.9 195.0 196.3
Dairy products..................... 179.9 131.3 177.5 141.2 186.7
Fish......................................... 153.9 148.9 160.0 137.9 163.7
Other foods:
Fruits and vegetables.. 125.2 131.2 115.1 103.5 169.6
Miscellaneous................. 112.9 112.5 133.4 138.8 143.2
Textiles.................................. 135.2 132.8 126.1 153.4 1*74.2
Hides, leather, and boots... 168.1 173.6 178.1 176.3 193.5
Metals and implements:
Metals.............................. 114.7 109.2 1 1 2 . 6 176.5 198.4
Implements.................... 106.6 106.6 107.5 113.2 116.6
Fuel and lighting.................. 113.6 109.0 108.9 106.2 1 2 2 . 0
Building materials:
Lumber.......................... 183.5 183.2 178.0 174.1 178.1
Miscellaneous.. ............. !114.0 1 1 0 . 8 108.2 120.3! 132.4
Paints, oils, and glass.. 140.2 140.6 142.9 162.1 ! 193.9
House'furnishings............... 128.8 128.8 131.9 138.7, 146.7
Drugs and chemicals........... 1 1 1 . 1 1 1 1 . 6 135.0 174.2 250.4
Miscellaneous:
Raw furs...................... : 226.5 235.0 1 2 1 . 8 144.0! 269.6
Liquors and tobacco.. . 138.8 128.3 137.9 134.7 136.7
Sundries.......................... 109.3 106.2 113.6 116.31 135.1
NUMBERS
Jan. July. Jan. July. Jan. July. Jan. July. Jan. July. Jan. July. Jan. Feb. Mar. Apr. May. June. July. Aug. Sept. Oct. Nov. Dec.
CANADA---- DEPARTMENT 0E LABOR.
201
Table 38, taken from the Labor Gazette, June, 1920 (p. 738), and
March, 1921 (p. 439), gives the general index number, by months,
from January, 1919, to February, 1921, inclusive.
T a b le
3 8 .— G ENER AL IN D E X NUMBERS OF W H O LESALE PRICES B Y
TO F E B R U A R Y , 1921.
Year and month.
1919.
January...........
February.........
March...............
April.................
M a y .................
June..................
July...................
August............ .
September____
October...........
November____
December........
Index j
num- i
ber.
286.5
279 8
277.6
279.6
284.1
284 1
294.0
301.1
301.5
299.6
307.7
322.7
Year and month.
1920.
January.........
February.___
March............
April..............
May................
June...............
July.................
August...........
September. . .
October.........
November___
December___
Index
num
ber.
338.4
343.5
349.0
353.1
356.6
349.3
346.8
330.2
326.6
317.6
304.2
290.5
MONTHS. 1)19
Year and month.
1921.
January...
February.
Index
num
ber.
281.3
270.1
DENMARK.
INDEX NUMBERS OF THE STATE STATISTICAL BUREAU.
HISTORY AND PUBLICATION.
This series of index numbers is based on the values of Danish
imports and exports. It was first compiled in 1907 by Michael
Koefoed, chief statistician of the State Statistical Bureau of Denmark,
and was published annually in the trade statistics of that country ,59
with a brief advance statement of it appearing in the journal of the
statistical office.60 Data for late years have been published in the
Statistical Yearbook (StatistisJc Aarbog). It covers a period extend-,
ing from 1876 down to recent years.
SOURCE OF QUOTATIONS.
The index numbers are computed from average annual prices
reported by various corporations, public authorities, and a consid
erable number of private business houses, upon the basis of which
the customs officials determine the value of imported and exported
commodities. For the four varieties of grains included in the index,
the figures taken are the official Government prices.
BASE PERIOD.
The decade 1891-1900 constitutes the base period used in the com
putation of the annual index numbers.
NUMBER AND GROUPING OF COMMODITIES.
From a list of about 100 commodities entering into the import and
export trade of Denmark, there were selected for inclusion in the
Danmarks vareindf^rselog-udf^rsel i aaret 1906-1912. Udgivet af det Statistiske departementet.
Copenhagen, 1907-1913. (Danmarks Statistisk Tabelveerk. 5. raekke, Litra D.)
60 Statistiske efterretninger, udgivet af det Statistiske departementet.
Copenhagen, 1909-1914.
202
INDEX NUM BERS OF WHOLESALE PRICES.
index 38 of the more important ones classified into three groups,
as follows:
Group L — Fats, oleomargarine, wheat flour, flaxseed, copra and palm kernel, rice,
coffee (green), cocoa bean, tobacco (raw), cotton, cotton yarn (undyed ), tallow,
copper (including brass, copperplates, and tin ), and petroleum— in all, 14 commodities,
Group I I . — Horses, eggs, salt herring, w heat, potatoes, wool, wool yarn (undyed),
hides (raw ), boots and shoes (not including those with silk tops), paper (writing and
print), firewood, and bricks— in all, 12 commodities.
Group I I I — Salt pork, meat (other than pork), butter, rye, barley, oats, m aize, oilmeal cakes, sugar (3 articles or more— granulated, whiter than D utch standard No. 18;
rock sugar, e tc.; also granulated, whiter than No. 9), lumber (rough, for ships, e tc.),
coal, bar and hoop iron— in all 12 commodities.
It is not possible to ascertain with any certainty the number and
variety of articles included. The classification used in the adminis
tration of the customs laws determines the nature of the commodi
ties which enter into this index number. A somewhat arbitrary
method of combining articles has, therefore, been adopted. Thus*
the articles coming under the single designation of “ boots and shoes”
apparently include all boots and shoes except those with silk tops;
“ sugar” includes two separate items in the tariff schedule and forms
in reality three or more articles, while “ paper ” includes two kinds
(writing and print) made up of various qualities combined for the
purposes of collecting the customs duty.
WEIGHTING.
The system of weighting is unique. The commodities have been
placed in three distinct groups, as already noted, and these three
groups in their numerical order have been given the relative impor
tance in the total index of 1 , 2 , and 3, respectively. No state
ment is made as to the reason why certain commodities were thrown
into any particular group. An examination of the grouping of the
commodities, however, leads to the inference that they were thrown
into one group or the other on the basis of their relative importance
in consumption.
TABLE OF RESULTS.
No group index numbers are given, only a general index for all 38
Commodities being published. This table follows:
3 9 .—IN D E X NUM BER OF W H O L E SA L E PRICES OF 38 IM PO RTA NT ARTICLES
E N T E R IN G INTO TH E IMJPORT AND E X P O R T TR A D E OF D EN M A R K , 1876 TO 1917.
T able
{Danmarks varemdf^rselog-udf^rsel i aaret 1912. Udgivetaf det Statistiske departementet. Copen
hagen, 1913: Pt. 2 , p. 7*, and Statistisk Aarbog 1920.J
(Average prices for 1891-1900= 100.)
Year.
Index
number.
1876....................
1877.....................
1878 .................
1879.....................
1880
. . ..
1881
.........
1882 .................
1883 .................
1884....................
1885 ...................
1886 ..........
145
135
122
120
128
129
127
126
120
109
101
Year.
1887.............
1888.................
1889.................
1890.................
1891.................
1892................
1893.................
1894.................
1895.................
1895.................
1897.................
1
Index
number.
99
105
109
109
-112
101
100
94
92
93
95
Year.
1898.................
1899.................
1900.................
1901.................
1902.................
1903.................
1904.................
1905.................
1905.................
1907.................
1908.................
Index
number.
99
105
110
106
108
105
107
110
111
118
113
Figures taken from the Statistisk Aarbog, 1920.
Year.
1909.................
1910.................
1911.............
1912.................
1913.............
1914...........
1915.................
1916.................
1917.................
1918.................
1919.................
Index
number.
115
120
123
130
1129
1145
1184
1244
1322
1392
1421
FRANCE-----AN m TAIRE STATISTIQUE BE DA FRANCE.
203
FRANCE.
INBEX NUMBERS OF ANNUAIRE STATISTIQUE DE LA FRANCE.
PUBLICATION.
The Statistical Annual (Annuaire Statistique de la France) pub
lished until 1918 by the General Statistical Office of France (Statistique
Generate de la France) in Paris, under the direction of the Ministry of
Labor and Social Welfare, contains index numbers for a group com
posed of certain food commodities and for a second group comprising
other commodities, such as mineral products, textiles, hides, oils,
etc., for all years since 1857.
In addition to these there are shown in certain of the reports, for
purpose of comparison, Sauerbeck's inde^ number for the United
Kingdom, as published in the Journal of the Royal Statistical Society,
the former index number for Hamburg, Germany, based on import
values, the index numbers for the United States, published by the
Senate Committee on Finance in 1893 and by the Bureau of Labor
Statistics in later years, and the one compiled by Jules Domergue for
France and published in La Reforme Eeonomique.
HISTORY.
Index numbers were first published in the Annuaire Statistique of
1904 (p. 151*). Previous to this date average wholesale prices for
certain food commodities, for fodder, and for fuels had been shown,
but no totals were made for these average prices, each of which represented data for one year. The prefatory note to the table ap
pearing in the 1904 report states that the index numbers for France
are based on the import values of 43 articles, the period 1867 to 1877
being taken as the Ibase. In this table all commodities are divided
into two classes—foodstuffs and miscellaneous materials. An index
number is also given for all articles combined.
In the preparation of the 1907 report the period 1891-1900 was
adopted as the base instead of the years 1867-1877 and, in accord
ance with this change, new index numbers were computed for all
preceding years.61 In 1912 a further change was made by the sub
stitution for the years since 1905 of index numbers based on the aver
age annual prices of 45 articles in interior markets of the country
instead of the import values of 43 articles, as in former reports.62
SOURCE OF QUOTATIONS.
The index numbers for years prior to 1906 are based on data
published by the customs administration showing the values of
.imports. These values were fixed by the board of appraisers (com
mission des valeurs en douane)» Beginning with 1906, as has been
stated, the index numbers are computed from the average yearly
prices of the different articles in interior markets of France. These
average prices are compiled mainly from records of transactions on
the Paris Bourse and from periodicals.63
** Annuaire Statistique.de la France, 1907, p. vii.
» Idem, 1912, p. 223*.
63 Mem, pp. 8 8 *, 89*.
204
INDEX NUMBERS OF WHOLESALE PRICES..
BASE PERIOD.
Prior to the 1907 report the base period used was that of 1867-1877.
In the 1907 report, as already stated, the base period was changed to
1891-1900. In the report for the years 1916, 1917, and 1918 the base
was again changed to 1901-1910 and recomputations for the previous
years made.
PRICES: H O W SH O W N AND COMPUTED.
The prices shown in the reports are in all cases averages of those
prevailing during the year. For years prior to 1906 these averages
are based on values assigned to imported articles by officials of the
customs service. The figures for years beginning with 1906 repre
sent in each case the average of 12 monthly quotations in selected
interior markets.
NUMBER AND CLASS OF COMMODITIES.
The index numbers for the years from 1857 to 1905, inclusive, are
computed on the import values of 43 articles, while those for years
since 1905 are based on the market prices of 45 articles. Both raw
and manufactured commodities are included, the former predomi
nating.
DESCRIPTION AND GROUPING OF CO M M ODITIES.
The commodities for which average yearly prices are published in
the Annuaire Statistique are shown in the following list appearing in
the report for 1913 (pp. 88* and 89*):
10.
n.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
WTheat.
W h ea t flour.
R ye.
Barley.
Oats.
Maize (corn).
Potatoes.
R ice.
Beef (V ille tte ).
V eal (V ille tte ).
M utton (V illette).
Pork (V illette).
Beef (H ailes Centrales).
V eal (H ailes Centrales).
M utton (Hailes Centrales).
Pork (H ailes Centrales).
Salt meats.
Butter.
Cheese (soft).
Cheese (dry).
Sugar (white, No. 3).
Sugar (refined, good quality).
Coffee.
Cocoa.
Bar iron (N o. 2).
Cast iron (pipes).
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Cast iron (colum ns).
Cast iron (plates).
Copper (bars).
T in .
Lead.
Zinc.
Coal.
Cotton.
F lax (raw).
H em p .
Jute.
W ool.
Silk (raw).
H ides (cattle).
Hides (horses).
Tallow .
Rapeseed oil.
Linseed oil.
A lcohol.
Petroleum (refined).
Soda (carbonate).
Soda (nitrate).
Indigo.
T im ber (Russian fir).
T im ber (Austrian oak).
R ubber (Para,.fine).
It is stated on page 223* of the 1913 report that since 1905 the index
numbers are computed on 45 of the above-named articles. It is not
shown which articles are not included .64 As previously stated, the
See, however, page 206 of this bulletin for list of articles published in the Bulletin de la Statistique
G£nerale de la France, October, 1912.
FRANCE---- AN NU AIitE
STATISTIQUE DE LA FRANCE.
205
commodities are arranged in three groups: Foodstuffs, miscellaneous
articles other than food, and all commodities combined. No descrip
tion of the articles appears in direct connection with the index num
bers as published in the Annuaire Statistique.
SUBSTITUTIONS AND ADDITIONS.
Except for the changes made in the preparation of the 1912 report,
no additions to the list of articles or substitutions of one grade or
quality of an article for another have been made, so far as the printed
information discloses.
INTERPOLATION.
No prices have been interpolated, as far as can be ascertained from
the reports.
W EIGHTING.
The index numbers are unweighted.
TESTING.
Other than the arrangement by which the index numbers are ex
hibited in comparison with other index numbers, by years, no testing
as to accuracy of results is apparent from the information at hand.
TABLES OF RESULTS.
Table 40, showing the variation in the index number by years from
1857 to 1918, inclusive, is reproduced in condensed form from the
Annuaire Statistique of 1916, 1917, and 1918 (Vol. X X X V , p. 312*).
T able
4 0 .—FLUCTUATIONS IN W H O LESALE PRICES, B Y Y E A R S, 1857 TO 1918.
(Average prices in 1901-1910=100.)
Year.
1857
1858........
1859 . .
1860
1861
1 8 6 2 ....
1863
1864
1865
1866 .
1867......
1 8 6 8 ....
1 8 6 9 ....
1870------1871........
1872. ..
1873.
1874.
1875. . .
1876
1877
Food.
Miscel
laneous
com
Total.
modi
ties.
133
114
118
127
137
127
122
119
115
124
128
133
129
133
144
135
138
133
127
131
140
166
153
150
155
145
152
157
156
144
142
133
131
130
133
134
149
147
131
130
128
124
152
137
137
144
142
142
143
141
132
134
131
132
130
133
138
144
144
132
129
130
131
Year.
1878....
1879....
1880....
1881....
1882....
1883....
1884....
1885....
1886....
1887----1888....
1889....
1890....
1891....
1892....
1893....
1894....
1895....
1896....
1897....
1898....
Food.
132
134
133
130
126
119
106
107
102
98
101
105
103
106
106
108
101
94
88
91
95
Miscel
laneous
Total.
com
modi
ties.
112
120
106
117
110
120
108
106
104
97
93
90
117
114
88
93
96
97
93
88
84
77
79
77
76
79
110
101
99
95
92
96
100
100
98
95
94
87
85
82
83
86
Year.
1899....
lyOO...,
la O l....
1902....
1903....
1904....
1905....
1906....
1907....
1908....
1909....
1910....
1911....
1912....
1913....
1914....
1915....
1916....
1917....
1918....
Food.
92
93
96
95
96
94
99
98
105
104
104
109
119
124
116
120
151
193
261
325
Miscel
laneous
Total.
com
modi
ties.
93
103
94
92
95
95
97
109
112
98
99
108
109
114
115
116
167
238
336
446
93
99
95
94
96
94
98
104
109
101
101
108
113
118
116
118
162
218
302
392
INDEX NUM BERS OF WHOLESALE PRICES,
INDEX NUMBERS OF THE STATISTIQUE GENERALE DE LA FRANCE.
HISTORY.
An index number apparently based on the same 45 articles that are
included in the Annuaire Statistique index, is published in the quar
terly bulletin which is issued by the General Statistical Office of
France. The publication of this index number began with the April,
1912, issue of the bulletin and has been continued m each subsequent
issue.
SOURCE OF QUOTATIONS.
The prices used in computing these index numbers are compiled
from various sources, among which are reports of trade bodies, priceregulating agencies, and commercial periodicals.
BASE PERIOD.
The base period from which price changes are measured is the aver
age of the years 1901-1910, represented by 100.
PRICES: H O W SHOW N AND COMPUTED.
Each number of the Bulletin de la Statistique Generate contains &
considerable amount of price data. Market quotations for most of
the articles included in the index, and covering a period of several
months, are shown for Paris or some other city of France in comparison
with similar information for leading cities of other countries. Current
market prices, by weeks or months, also are published for a large
number of commodities. In most instances the prices relate to a
particular day o f the week or month. Relative prices for each of
the 45 articles in the index, dating back to 1901 and computed on
prices in 1901-1910 as the base, are published in the issues of January,
1919, and January, 1920.
NUMBER AND CLASS OF COMMODITIES.
The index numbers of the Statistique Generale are based on prices
of 45 commodities, divided into two groups, as follows:
Twenty food products.— Wheat, wheat flour, rye, barley, oats*
maize, potatoes, rice, beef (2 kinds), mutton (2 kinds), pork, salt
meats, butter, cheese, raw sugar, refined sugar, coffee, and cocoa.
Twenty-five industrial products.— Cast iron, wrought iron, copper,
tin, lead, zinc, coal, cotton, flax, hemp, jute, wool, silk, salted hides,
skins, tallow, rapeseed oil, linseed oil, alcohol (90°), petroleum,
soda carbonate, soda nitrate, indigo, lumber, and rubber.
W EIGHTING.
No system of weighting the commodity prices is employed in the
construction of these index numbers.
TABLES OF RESULTS.
Tables 41 and 42 taken from the January, 1921, issue of the Bulletin
de la Statistique Generale de la France (pages 109, 1 1 0 ), give the
index numbers since 1913 and also offer a comparison with certain
index numbers of Great Britain, Italy, and the United States.
201
FRAN OB-----STATISTIQUE GENERALE BE LA FRANCE.
T able 4:1.—IN D E X N U M B E R S, B Y G R O U PS OF C O M M ODITIES, B Y Y E A R S , 1913 TO 192a
A N D B Y M O N TH S, O C T O B E R T O D E C E M B E R , 1920.
(A verage prices in 1901-1910=100.)
Date.
191 3
191 4
191 5
191 6
194.7............................. ..............
191 8
191 9
..........................
192 0
October...............................
November..........................
December........ ..................
Food products.
Gen
eral
index
num
bers
(45).
Vege
table
food
(8).
120
115
118
162
218
302
332
412
589
580532. 6
502.6
123
151'
204
291
358
376
512
504.8
468.5
431. 4
Ani
mal
focd
(8).
118
122
149
191
254
338
463
593
629.4
612.3
607.8
Sugar,
coffee,
cocoa
Industrial products.
Miner-1
ais and! Tex
metals ! tiles
(6).
Total
(20).
(4 ).
106
112
160
174
213
235
26S
446
447: 8
400
376. 4
116
120
1 152
193
2G
0
325
389
532
543.2
512.3
490.9
Miscel
lane
ous
(7 ).
(12).
120
117
117
I 127
154
197
210
278
354
325340
538
520
326
863
538
544.1
745
597.1
509.1
462.4
556.3
110
109
160
219
332
461
445
577
579.6
547. 8
518.6
Totai
(25).
115
114
169
1234
336
446
430
635
609.3
548. £
512
1
1
This number does not agree with that given in Table 42, but both are as shown in the Bulletin dela
Statistique Generate.
T able 4 3 .—COMPARISON OF IN D E X NUMBERS FOR FRANCE, EN G LAN D , IT A L Y , AN D
THE U NITED STAT1TS.
(Average prices in 1901-1910= 100.)
France,
England.
Italy.
Unite#
States.
Prof.
Bachi
(44).
(200).
Statistique geneiale.
Date.
Food
products
Indus
trial
products
(25).
116.4
120.2
150.9
192.7
260.6
325.2
389.2
531.7
545
545.2
52L2
515.5
114.7
116.1
166.9
237.6
335.8
445. 6
429.9
634-6
648.2
694.7
639.1
556.7
609.3
548. S
512
(20).
USA ........................
19*4.................
191 5
191 6
191 7
1918,.........................
1 9 1 9 ......................
192a.................
First quarter. *
Second quarter
Third quarter..
Fourth quarter
October__
N ovember.
December..
5*3.2
512.3
490. 9
Total
(45).
115.4
117.9
161.6
217.6
302.4
392.1
411.8
588.9
602.3
628. 2
586; 7
538.4
my
532.6
502.6
Sauer
beck
(45).
Econo
mist
(44).
115.9
116.3
145.1
184.7
238
262.1
280
337.8
348,2
354. 8
343.6
304.5
326.6
301. 7
282.1
114.8
113.3
141. 2
184.3
234. 2
258.1
269.9
325
345.1
344. 9
330.5
279.7
305.7
281
252.5
119.9
114
159.1
239.3
367.1
m
438.5
671
780.1
761.3
781.4
793.8
788.8
761.7
Dun
111.4
114.8
119
ms
191.4
215.4
216.9
232.9
236.2
245. 4
238. 4
211. a
223
213. 5
199
In 1911 the General Statistical Office published a volume devoted
exclusively to the subject of wages and the cost of living in France at
different epochs.65 Under the cost of living topic is included a study
of wholesale prices, contract prices, and retail prices, with index num
bers for each class. Index numbers are also given for wages and for
family budgets.
In Table 43, which has been compiled from data published on pages
44 and 45 of the report, are shown index numbers covering the period
from 1880 to 1909, inclusive, computed from the wholesale or import
Salaires et Colit de 1’ Existence a diverses Epoques, jusqu’en 1910.
Paris.
Imprimerie Nationale.
208
I3STDEX NUM BERS OF WHOLESALE PRICES.
prices of 10 commodities in common use. The figures for refined
sugar and illuminating oil are based on wholesale prices of the Bourse
de Commerce of Paris. Those for wine are computed from prices
published annually by the minister of finance in the Bulletin de
Statistique et de Legislation. The index numbers for the remaining
articles are based on import prices. The average of the prices for
1891-1900; taken as 100, constitutes the base.
T a b le
4 3 .—R EL ATIV E PRICES OF COMMODITIES, 1S80 TO 1909.
(Average prices in 1891-1900=100.)
Year.
1880..............................
1881..............................
1882.............................
1883..............................
1884..............................
1885.............................
1886..............................
1887..............................
1888..............................
1889..............................
1890..............................
1891.............................
1892..............................
1893..............................
1894............................. 1
1895.............................
1896.............................
1897..............................
1898..............................1I
1899..............................
1900..............................
1901..............................
1902..............................
1903..............................
1904..............................
1905..............................
1906..............................
1907..............................
1908.............................
1909.............................
Bread. Butter. Cheese.
137
135
134
114
105
101
107
99
112
109
100
106
10 1
84
78
78
100
120
121
105
105
105
106
107
100
100
98
98
95
91
89
82
82
82
82
82
100
107
98
96
107
96
102
104
107
111
107
100
102
110
120
120
104
129
129
130
107
107
105
111
117
117
121
121
121
110
102
99
>02
110
113
110
100
104
97
98
98
98
99
99
101
110
113
104
106
110
88
120
120
128
Pota
toes.
150 1
142
150
145
133
150
100
92
125
92
118
117
83
100
100
67
83
117
100
117
117
167
150
167
150
167
183
167
199
199
Rice.
132
129
118
122
122
125
125
1-25
107
125
125
111
125
129
118
100
82
93
90
79
82
82
82
75
80
84
86
89
104
100
Wine,
ordi
nary.
Oil,
edible.
169
169
150
166
184
183
183
175
161
149
146
134
138
124
115
105
148
156
157
144
155
155
156
140
118
122
138
116
111
90
89
122
98
92
88
78
73
73
73
112
99
70
56
78
109
65
62
71
101
121
128
128
131
142
149
181
192
66
!
62
70
Sugar,
refined.
133
114
106
101
100
100
92
94
102
111
102
103
101
Coal.
98
100
100
82
82
77
71
71
82
Oil,
illumi
nating.
132
133
130
156
121
109
94
97
107
120
122
120
122
93
93
123
97
98
109
88
101
90
88
88
91
103
95
96
92
99
101
100
96
91
78
50
63
56
55
58
60
90
95
105
113
147
131
111
101
93
93
112
128
117
128
100
93
89
117
110
104
90
83
88
115
141
109
106
Table 44 , reproduced from page 45 of the report, contains three
series of index numbers representing in each case the average of index
numbers computed for the 10 articles included in the preceding table.
In addition to index numbers based on wholesale or import prices,
similar data for contract prices paid by the Department of Public Aid
and for retail prices are given in this table.
FRANCE---- STATISTIQUE GENERALE DE LA FRANCE.
209
T able 4 4 .—THREE SERIES OF IN D E X NUMBEJRS COMPUTED FOR 10 ARTICLES, 1880 TO
1909.
(A verage prices in 1891-1900=1000
Whole
sale or
import
prices.
Year.
132
129
126
125
1880
1881
1882.
1883
1884.
1885.
1886.
1887.
1888.
1889.
1890.
1891.
1892
1893
1894
122
120
112
108
110
114
117
110
105
102
Contract
prices
paid
by the
Depart
ment of
Public
Aid.
Retail
prices.
126
121
125
116
109
Ill
117
105
106
110
101
104
106
104
105
108
106
101
100
94
104
105
103
104
108
103
105
120
Year.
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
Whole
sale or
import
prices.
Contract
prices
paid
by the
Depart
ment of
Public
Aid.
94
94
98
99
98
102
107
107
106
99
102
109
115
122
122
Retail.
prices.
86
93
93
99
100
97
99
95
93
94
M
95
95
99
101
INDEX NUMBERS OF LA REFORME ECONOMIQUE.
PUBLICATION.
This series of index numbers of wholesale prices in France formerly
was published weekly in La Reforme Economique, a journal of social
and political economy, Jules Domergue, editor. Sincexthe beginning
of the war period, only certain information for a limited number of
individual commodities has been continued, no general index number
being calculated.
HISTORY.
For a number of years prior to 1900 comparative prices for a
varying number of commodities were published in this journal. In
the earlier numbers the increase or decrease in prices of seven com
modities as compared with the primary period (1890) was shown by
means of chart. In 1894 prices for corresponding dates in 1892,
1893, and 1894 were given, and for some articles an average monthly
price was computed. In 1896 a weekly table of prices was presented
for the first time, and on May 9, 1897, a series of tables was begun
showing the average monthly prices of all the commodities used in
the computation. Beginning with January, 1899, an annual average
price was computed and published for the years 1890 to 1898, inclu
sive, and in addition current prices were compared with these by
means of annual average prices. No comparison was made by means
of index numbers, however, until 1900, when the method of presenting
the variation in prices for stated periods was changed by showing in
addition to the average price of each commodity the simple percent
ages of increase or decrease in the various commodities. Owing to
the situation brought about by war, only fragmentary information
has been published since 1913.
3 3 2 2 6 °— 21— B ull. 284--------14
210
INDEX NUMBERS OF WHOLESALE PRICES.
SOURCE OF QUOTATIONS.
Actual commercial transactions are sought for in the collection of
price data. The prices quoted are said to be those obtained from the
records of licensed brokers (courtiers assermentes) and private
brokers (courtiers Hires) in different parts of Paris and in the depart
ments; official quotations of the Department of Agriculture; prices
obtained in the municipal markets of La Villette and Hailes Cen
trales; quotations on importations as stated in the records of the
Government warehouses in Paris and elsewhere; and official quota
tions of the price of bread furnished by the prefecture of the Seine.
BASE PERIOD.
The year 1890 is used as the base period. Nothing is stated in
connection with the figures, so far as can be ascertained, as to why
this year was chosen in preference to any other year or period of
years.
PRICES: H O W SHOW N AND COMPUTED.
As has been stated, no comparison of prices was made by means of
percentages until 1900. In that year, after noting the principal
objections to an index number representative of the combined prices
of all commodities for which prices are quoted, it was decided to
construct an index for each article and an index for each group of
articles, in addition to the index for the entire list. The prices used
in the calculations were the average annual prices already published
in La Reforme ficonomique for the period 1890-1895, the quarterly
average prices for the years 1896-1898, and the average monthly
quotations for ythe year 1899.
A special presentation of the price of cotton is made in the issue of
September 23, 1900, showing for five grades the range of prices o f this
commodity from 1875 to September, 1900, with an index based on
the price in 1875 (equal to 100 ). In the supplement to the issue of
November 24, 1901, the averege prices of wheat, sugar, wine, and
alcohol for the periods 1884-1891 and 1893-1900 have been computed
and the divergence in price between the two periods is shown. The
prices for the year 1892 are not a factor in either period.
NUMBER AND CLASS OF COMMODITIES.
In tracing this index number through the period elapsing since its
inception, there is great difficulty in determining the number and
class of commodities which have been included in the compilation.
Between 1890 and 1899 prices and index numbers were shown for
from 40 to 56 or even more articles, according to the system of count
ing and classification adopted. Thus the compiler evidently com
bined four kinds of meat (beef, pork, mutton, and veal) into one
commodity, while from two to four kinds of wool, silk, or cotton were
each sometimes given a separate index number. From 1900 to date
it would appear that approximately 48 articles make up the series of
index percentages, although average annual wholesale prices are
given for several additional articles.66 After 1904 a statement printed
on the cover page and called 44La Thermometre des Affaires en
France ” shows a separate index number for only 21 leading commodi
ties, although the general percentage index number includes addi
tional articles, as may be verified by actual arithmetical test.
68
Annuaire Statisque de la France, 1912, p. 223*.
FBAKCE— LA RE-FORME ECONOMIQUE.
211
DESCRIPTION AND GROUPING OF COM M ODITIES.
The classification of the various commodities has differed from time
to time. The first summary table, presented in the issue of January
14, 1900, page 67, shows index percentages for 56 commodities (43 if
certain grades of textiles and bar and structural iron are not considered
separate commodities), classified in five main groups as follows:
Food produces.— Wheat, flour,, rye, barley, meats, wine, sugar,
alcohol, coffee, butter, sirup, edible starches (fecula),67 oleomargarine,
tallow, lard, cocoa, rice, and bread.
Textiles.— Silk (2 grades), wool (raw and yarn), linen (raw and
thread), cotton (raw and spun), hemp, jute.
Agricultural products.—Hides (raw), leather, oats, maize, fodder,
fatty acids (3), rapeseed oil, linseed oil.
Minerals and metals,— Coal, petroleum, copper, tin, zinc, lead, steel,
iron (2 grades), sheet iron.
Miscellaneous.— Rubber, sulphuric acid, hydrochloric acid, chloride
of lime, carbonate of soda, sal soda, sulphate of ammonia, super
phosphates.
In the issue of April 14, 1901, average annual wholesale prices are
presented for all of these 56 commodities, while relative prices are
shown for only 43 of them, divided into five groups as above, except
that the fifth group is termed 11Chemicals and fertilizers ” but contains
the same commodities as the group which is designated above as
“ Miscellaneous” commodities. Certain interchanges were also made
as between “ Food products” and “ Agricultural products.” This
classification was continued until January, 1902, so as to include
indexes for the }^ear 1901.^
In 1904 the list of commodities for which separate relative prices
were presented was reduced to 21 commodities as follows: Wheat,
meat, wine, sugar, alcohol, coffee, coal, petroleum, copper, tin, zinc,
lead, steel, iron, silk, wool, flax, cotton^ nitrate of soda, superphos
phates, and sulphuric acid.
Beginning with the issue of December 10 , 1905, a change was made
in the form of classification, the articles beii^g grouped as ( 1 ) agricul
tural products and (2 ) industrial products. Under this classification
average annual wholesale prices are presented for 21 articles in the
first group, if forage is counted as one commodity and not as two, and
wines as one instead of three, meats as one instead of four {beef r pork,
veal, and mutton) Tand fatty acids as a single commodity instead of
three; whereas if all these subdivisions are counted as separate com
modities,. the number would appe&r as 29. In the second group,, that
of industrial products, there are 28 or 40 commodities, according to
the system o f counting adopted.
SUBSTITUTIONS AN D ADDITIONS.
In the issue of La Reforme Ecanomique for January 15, 1899, the
average prices of raw and spun silk of several grades were included for
the period of 1890-1898, and on February 12 of the same year average
prices of iron (two grades), steel rails, sheet iron, bar copper, and
tin (Banka) were added for the years 1890 to that date.
Two further groups were added on April 2 , 1899. The first of these
included wine, coffee (Santos, good average), hops (Burgundy),
67
The French term “ f6 cule” includes such articles as potato flour, tapioca, sago, arrowroot flour, etc.
212
INDEX NUMBERS OF WHOLESALE PRICES.
nitrate of soda, sulphate of ammonia, and superphosphates; the sec
ond, sulphuric acid 66 °, hydrochloric acid 20- 2 1 °, chloride of lime
105-110°, carbonate of soda 90-92°, and caustic soda 80°, for manu
facturing and laundry purposes. Quotations for-hides, raw (three
quotations), and tanned (four quotations) were added on April 23,
1899. The index for tanned leather does not appear after 1900, nor
do those for cocoa, rice, rubber, oleomargarine, bread, lard, and
fatty acids.
During the year 1902 percentage relatives for two articles were
dropped from the list, ana in the first issue of 1903 two more indexes
were dropped. After 1905 no quotations appear for the following
classes of wine: Alicante, Huelva, Aragon, Valence, and Haw.
Apparently no adjustment of previous percentages have been made.
No further change in the list of articles or method of presenting the
variation of prices appears to have been made since 1905.
Oleomargarine did not figure in the index series until 1896, and it
was dropped in 1900. Up to 1900 separate relative prices were cal
culated for two kinds of sugar, raw and refined; after that date only
one index is shown, although wholesale prices are quoted for botn
kinds. Apparently, however, the index percentage is based on the
average of the prices shown for each kind. Similar changes have
taken place in the coal index. Thus, four index percentages were
resented up to 1900, but after that date only one is shown, which is
ased on the average wholesale prices of all four kinds.
In general, when any changes were made in the number or classes
of commodities the index percentages were recomputed back to the
base year, 1890.
P
INTERPOLATION.
No method of price interpolation has been resorted to, so far as
can be determined from the information published in La Reforme
Economique.
WEIGHTING.
No scientific method of weighting has been used, the arithmetic
average alone being employed m the construction of the index num
bers. The method of calculating the yearly general index for groups
of commodities and for all commodities seems to be as follows: For
the years already covered by the reports on wholesale prices, the sum
of the average prices for the year of the different articles was divided
by that of the basic year (1890). Thereafter, average monthly prices
were obtained by getting the average of the weekly quotations made
during the month as published in La Reforme Economique, and from
these average monthly prices the yearly average price was computed,
the yearly general index then being computed as before. The rela
tives for each article (when given), the index for each group, and the
general index are in all cases simple percentages based on the prices
of corresponding items for the corresponding period in 1890.
TESTING.
In the earlier years Sauerbeck’s table of index numbers for Eng
land was occasionally given for comparison, and in the later issues it
has appeared quite regularly. No other comparisons are made.
TRANCE— LA REFORME ECONOMIQUE.
213
TABLE OF RESULTS.
The following table of index numbers has been compiled from the
numbers of La Reforme Economique published during the years
1892-1913:
T able
4 5 .—IN D E X NUM BERS COMPUTED FROM A V E R A G E A N N U A L PRICES, 1891 TO 1913.
[Data from La Reforme Economique.]
(Average prices in 1890=100.)
Index
num
ber.
Year.
3891 ..................................
1892..................................
1893 ..................................
1894..................................
1895 ..................................
1896..................................
1897 ..................................
1898 ..................................
99.6
94.2
97.6
89.4
84.4
82.2
83.4
87.6
Year.
Index
num
ber.
1899 ......................................
1900 ......................................
1901......................................
1902 ......................................
1903 ......................................
1904 ......................................
1905 ......................................
1906 ......................................
95.6
102.4
95.8
94.2
95.8
95.2
95.8
105.4
Year.
1907
...................
1908
.................
1909
..........................
1910 ....................................
1911 ....................................
1912......................................
1913......................................
Index
num
ber.
1 1 2 .2
1 0 1 .2
1 0 1 .8
108.2
113.8
117.8
116.0
INDEX NUMBERS OF EMILE LEVASSEUR.
PUBLICATION.
This “ Inquiry into the price of food commodities for a period of
25 years in 70 high schools of France’ ’ was published in the Revue
Economique Internationale, Brussels, ixi May, 1909. Later in the
same year, under the title of “ Le coiit de la vie,” it appeared as a sep
arate pamphlet, which also was published by the Revue .68
HISTORY.
Toward the close of 1908 the minister of public instruction, at the
request of Mr. E. Levasseur, addressed to the principals of 70 high
schools of Paris and the Departments an inquiry concerning the
prices of certain food products and of coal as paid by the schools
since 1880.
The inquiry covered the years 1880, 1885, 1890, 1895, and each
year from 1900 to 1908, inclusive.
The schools selected, exclusive of those of Paris and its suburbs,
included some of the large and some of the small institutions in each
of the nine agricultural sections of France.
Mr. Lucien March, chief of the general statistical office (statistique
generale) of France, assisted in the work by assuming the responsi
bility for the calculation of the index numbers from the figures
secured.
The author states that from the data received it was possible to
secure a sufficiently exact report of the variation which the prices of
commodities had undergone for a period of about 25 years.
SOURCE OF QUOTATIONS.
The prices considered are contract or semiwholesale prices (yrix
dJadjudication) obtained from 70 high schools. The articles are 21
in number— 20 food commodities and coal.
68 Le eoftt, de la vie. Enquete sur le prix des denrees alimentaires depuis un quart de si&cle dans 70
lycees, par E. Levasseur, membre de PInstitut, administrateur du College de France.
214
INDEX NUM BERS OF WHOLESALE PRICES.
BASE PERIOD.
The average price for the two years 1895 and 1900, taken as 100,
is used as the base. The index numbers for the period 1880 to 1908
derived from the use of this base are shown in the following table,
appearing on page 7 of the pamphlet:
Index.
number.
Year.
1880.
1885.
1890.
1895.
1900.
1901.
1902.
111. 9
104. 2
1 0 1 .4
1 0 0 .2
99. 3
99. 8
9 8. 8
Year.
1903
1904
1905
1906
1907
1908
Index.
number.
..
..
..
..
..
..
99. 9
99. 9
98. 0
98. 8
1 03 .1
1 0 6 .5
DESCRIPTION OF COMMODITIES.
The articles selected for which Mr. E. Levasseur computed .index
numbers are the following:
Bread.
Fresh meats (other than pork).
Fresh pork.
Smoked pork (charcuterie).
Poultry and game.
R ed wine.
W h ite wine.
B eer.69
Cider.70
Butter.
Drippings and lard.
O il (table).
Eggs.
M ilk.
Cheese.
Sugar.
Fresh fish.
Salted fish and canned fish.
Codfish.
Potatoes.
Coal.
TESTING.
Mr. Levasseur verified his index numbers by comparison with index
numbers for France, England, Germany, and the United States.
The variations as shown by the index numbers of prices in 70 high
schools were verified by comparisons with the following:
(a) Index numbers relating to France only—
1 . Statistics prepared by Mr. Lucien March, chief of the general
statistical office of France, and published in the Annuaire Statistique
de la France. These show a greater increase for all merchandise in
general than for food commodities alone (except in 1895). The index
numbers show a rapid increase since 1905.
2 . Index numbers calculated by Mr. de Foville. These numbers
were based on the difference between the' rates of duty, and show
prices to have been low in 1900, with an increase in all the following
years, especially since 1903.
3. Index numbers calculated by Mr. Levasseur from the duties on
32 food commodities, corresponding closely to those of the high
schools. These figures show a rapid rise in prices since 1903.
4. Index numbers calculated on the basis of current prices since
1882 by the purchasing agent of the southern railways. These index
numbers verify (or confirm) the index numbers of the high schools of
the southwest.
(i) Index numbers relating to foreign countries—
5. England— Sauerbeck’s index numbers for 45 articles of general
merchandise. These index numbers are published in the Journal of
© Included in the calculation of index numbers in 1 instance only.
70 Included in the calculation of index numbers in 3 instances only
FRANCE— EM ILE LEVASSEUR.
215
the Royal Statistical Society. Sauerbeck’s index numbers for food
commodities show, like those of the high schools, a slight fall in prices
from 1900 to 1902 and also a marked rise in 1907 and 1908.
6 . Index numbers computed in Germany for the city of Hamburg.
7. Index numbers computed b}^ the United States Bureau of Labor
Statistics. These index numbers are for wholesale and retail prices,
respectively.
In conclusion the author states that u these diverse statistics,
despite the differences of detail, confirm the statistics of the 70 high
schools and show clearly that the great changes in prices are not due
to special or local causes, but to general causes, the results of which
are felt at the time in all the gfeat markets which are in constant
commercial communication with each other. ” 71
TABLES OF RESULTS.
Table 46, reproduced from page 15 of the publication, shows the
variations in the index numbers for food articles in the 70 high
schools of Paris and suburbs and of the 9 agricultural sections of
France, by years, division into large and small schools being m ade:
TABLE 4 6 .—IN D E X NUM BERS OF PRICES OF 20 FOOD COMMODITIES AND COAL IN 70
HIGH SCHOOLS OF PARIS A N D ITS ENVIRONS AN D OF TH E 9 AG R ICU LTU RA L SEC
TIONS OF FRANCE, W IT H DIVISION INTO LAR GE AN D SM ALL SCHOOLS.
(Average prices in 1895 and 1900=100.)
Sections.
1880
1885
1890
1895
1900
1901
1902
1903
1904
1905
1906
1907
Paris............................
Environs of Paris___
Northwest:
Large schools___
Small schools___
North:
Large schools___
Small schools___
Northeast:
Large schools___
Small schools.. . .
East:
Large schools___
Small schools....
Southeast:
Large schools___
Small schools
South:
Large schools___
Small schools___
Southwest:
Large schools___
Small schools___
West:
Large schools___
Small schools___
Central:
Large schools.. . .
Small schools___
115
114
112
106
99
100
101
98
99
102
99
99
97
99
100
105
99
99
95
97
96
99
98
102
99
105
117
111
111
100
105
106
100
100
100
100
96
98
96
100
99
101
103
102
93
99
95
102
95 1
104 !
104
109
111
114
98
99
102
97
100
105
100
95
100
98
98
98
99
99
98
99
97
100
96
107
99 1
110 ;!
102
109
111
122
106
106
103
98
101
100
99
100
99
101
98
98
97
97
97
9S
95
99
96
100
103
102 :t
107
109
108
109
103
108
101
101
100
99
100
101
97
99
99
100
99
100
101
100
99
99
99
99
102
104
106
106
105
104
100
106
100
102
100
98
100
99
100
94
99
95
101
100
99
97
98
101
101
106
108
107
108
116
102
102
95
98
98
101
95
103
93
99
96
101
90
103
90
100
88
100
99
107
98
112
106
115
102
112
98
104
101
102
99
98
101
98
100
101
98
101
98
103
96
100
98
102
104
107
105
109
108
112
99
103
102
109
99
100
101
100
105
104
102
101
104
99
102
96
98
96
97
97
101
102
108
107
1
98 ; 98
102 ! 102
95
103
102
104
104
113
117
111
113
101
108
101
100
94
99*
104
100
96
100
97
103
93
100
1908
104
Paris............................
Other schools:
Large schools (37)
Small schools (30)
114
109
99
100
100
101
97
100 !
99
94
96
98
98
110
113
103
105
100
102
100
101
99
99
99
101
98
99
100
100
99
101
96
99
97
101
101
105
105
109
Total.................
111
104
101
100
99
100
99
100
100
98
99
103
107
Index numbers computed on the prices of 20 food commodities
and coal in 1908 are shown in Table 47, the figures being given
separately for Paris, its suburbs, and the large and small schools,
respectively, of the 9 agricultural sections of France.72
Le cout de La vie, p. 11.
72
Idem, pp. 22 and 23.
INDEX NUMBERS OF WHOLESALE PRICES.
216
4 7 .—IN D E X NUM BERS OF PRICES OF 20 FOOD COMMODITIES AN D COAL IN 20
GROUPS (PARIS, ENVIR O NS, LAR GE AN D SM ALL SCHOOLS OF T H E 9 A G R IC U L T U R A L
SECTIONS OF FRANCE) IN 1908.
T able
(Average prices in 1895 and 1900=100.)
Sections.
Fresh Smoked
Bread. pork. pork.
Paris........................
Environs of Paris..
Northwest:
Large schools..
Small schools..
North:
Large schools..
Small schools..
Northeast:
Large schools.,
Small schools..
East:
Large schools..
Small schools.,
Southeast:
Large schools.,
Small schools.,
South:
Large schools.
Small schools.
Southwest:
Large schools.
Small schools.
West: '
Large schools.
Small schools.
Central:
Large schools.
Small schools.
Poul
try
Red White
and wine. wine.
game.
115.3
120.2
117.8
174.4
116.4
135.6
106.9
47.2
62.2
73.1
102.1
117.6
121.6
166.0
114.2
164.1
113.2
111.7
124.0
•65.1
57.9
64.7
113.3
119.2
113.1
106.3
189.6
97.1
105.8
113.0
53.3
73.3
80.0
97.8
114.5
119.8
106.2
106.1
136. J
116. £
105.7
106.2
59.3
69.3
119.9
116.6
116.5
113.7
121.1
116.3
117.6
112.9
122.0
120.7
123.8
113.5
115.7
128.1
106.7
118.7
97.8
116.0
122.1
Beer. Cider.
But
ter.
Drip
pings
and
lard.
Fresh
meats
(exclu
sive of
pork).
104.5
121.9
131.2
123.3
104.4
105.3
110.3
112.1
169.5
154.2
110.5
109.9
105.7
104.0
107.5
81.1
76.3
72.9
111.1
138.5
70.4
70.1
71.6
88.1
109.4
101.3
97.9
65.2
81.2
77.7
84.5
116.1
102.8
135.2
134.6
95.7
107.7
110.7
134.7
68.1
63.7
103.6
111.5
104.4
150.6
109.3
128.6
101.3
117.1
108.7
74.3
72.6
112.1
118.4
112.8
116.6
116.1
110.5
103.1
122.0
128.3
110.1
136.5
116.7
130.2
110.3
106.5
135.5
130. £
131. <
61.2
61.8
72.0
82.2
119.3
95.2
106.7
109.5
105.2
121.5
132.5
111.0
110.5
140.1
118.6
108.2
65.0
66.7
56.4
84.6
112.6
Generalaverage. 119.0
113.0
63.3
9.2
Sections.
Paris........................
Environs of Paris.
Northwest:
Large schools..
Small schools..
North:
Large schools..
Small schools..
Northeast:
Large schools..
Small schools.,
East:
Large schools.,
Small schools.,
Southeast:
Large schools.
Small schools.
South:
Large schools.
Small schools.
Southwest:
Large schools.
Small schools.
West:
Large schools.
Small schools.
Central:
Large schools.
Small schools.
General average.
113.1
Oil
(table).I Eggs.
Milk.
91.5
113.8
Salted
fish.
Cod
fish.
104.1
Pota
toes.
91.3
112.2
110.3
107.5
100.0
108.0
108.9
59.3
60.7
71.0
98.4
108.3
87.9
116.1
127.9
112.4
96.9
116.2
111.2
61.0
61.0
98.5
108.6
93.4
114.1
108.2
104.7
106.5
104.5
110.6
130.5
64.7
64.1
114.8
114. 5
117.2
125.0
107.2
99.0
105.4
125.5
152.8
61.2
79.2
81.4
125.2
150.7
107.4
105.1
120.6
123.9
98.0
105. G
59.2
116.3
86.7
59.1
86.7
91.4
118.7
130.8
136.1
102.9
109.3
115.3
123.7
100.4
125.0
120.6
125.0
62.5
58.2
93.9
100.0
106.0
117.0
125.2
74.2
111.6
115.6
134.0
121.7
111.6
104.8
113.0
56.8
62.6
75.9
78.2
109.6
143.2
105.8
106.5
119.7
100.2
84.3
126.3
123.4
99.1
94.3
117.0
109.8
60.1
60.6
74.8
75.7
128.4
97.9
119.6
131.4
114.0
123.1
103.9
114.6
123.5
100.0
123.5
155.8
61.5
61.0
84.0
85.7
84.3
81.7
9.5
82.6
95.5
138.9
122.3
112.2
87.7
118.4
112.7
151.8
63.3
64.9
110.3
103.8
109.1
115.3
105.1
117.6
116.2
62.6
89.0
107.5
110.5
81.6
127.2
103.8
123.1
112.6
126.2
78.2
109.2
112.0
120.3
Fresh
fish.
105.2
110.9
101.1
116.7
Cheese. Sugar.
105.5
109.0
112.5
118.8
110.1
111.1
103.7
109.1
108.1
125.8
110.7
Coal.
113.5
110.3
122.3
117.6
103.7
130.3
135.7
124.6
95.0
117.7
128.5
109.7
180.4
124.7
119.9
121.2
111.1
107.1
148.8
156.0
128.3
144.9
129; 5
GERMANY---- FORMER IMPERIAL STATISTICAL OFFICE.
217
GERMANY.
INDEX NUMBERS OF THE FORMER IMPERIAL STATISTICAL OFFICE.
PUBLICATION.
This series of index numbers, which represents wholesale prices of
commodities in German markets, formerly was published yearly in
the Yierteljahrshefte zur Statistik des Deutschen Reichs, a publica
tion of the Imperial Statistical Office, and appeared regularly for
each year in the first quarter of the succeeding year.
The first report, including index numbers, was published in 1905
and covered the years 1899 to 1904. The table of index numbers
in later reports regularly covered the 10 -year period ending with the
date of the publication of the report. Since the dissolution of the
German Imperial Government in 1918 the publication of the Vierteljahrshefte has been continued by the Statistical Office (Statistischen Reichsamt) but only fragmentary price data have been in
cluded.
HISTORY.
Beginning with 1879 the German Imperial Statistical Office pub
lished monthly average wholesale prices of commodities of importance
in German markets. These were shown in detail in the Monatshefte
zur Statistik des Deutschen Reichs up to the year 1891, and for later
years in the Vierteljahrshefte zur Statistik des Deutschen Reichs.
The object of the price study, as stated at the outset, was the col
lection of reasonably accurate and adequate average prices repre
senting fixed grades of important articles of the wholesale trade,
with a view to the gradual assembling of really useful data for the
observation of the movement of prices. It was not until the year
1905 that the publication of relative prices was begun. The official
series of index numbers was extended back only as far as the year
1899.
SOURCES OF QUOTATIONS.
The number of markets represented in this study was limited
to those with permanent arrangements for furnishing reasonably
accurate and representative quotations. The following sources of
information, representing 30 wholesale markets, are acknowledged
in the report for the year 1917: Chambers of commerce or boards, of
trade in Augsburg, Berlin, Bielefeld, Brunswick, Bremen, Breslau,
Danzig, Dortmund, Frankfort on the Main, Halberstadt, Hamburg,
Cologne, Konigsberg (in Prussia), Krefeld, Landeshut (in Silesia),
Lubeck, Magdeburg, Mannheim, Miihlhausen (in Alsace), Munich,
Miinchen-Gladbach, Nuremberg, Posen, and Stettin; the adminis
trations of municipal stockyards and slaughterhouses; the mill
administration in Bromberg (for wheat flour f. o. b. Berlin), the
stock exchange in Dusseldorf, the board of directors of the stock
exchange in the city of Essen, the United German Jute Manufac
turers in Harburg on the Elbe (for the raw jute f. o. b. Hamburg),
the Merchants’ Association at Lindau, the Royal Administration
of Mines at Saarbrucken, the Imperial Grain Office at Berlin, the
Imperial Meat Office at Berlin, the German Green Hides Corporation
at Berlin, the Statistical Office of the city of Leipzig, and the Bureau
of Trade Statistics at Hamburg.
218
INDEX NUM BERS OF WHOLESALE PRICES.
From the beginning ordinary published market quotations were
avoided as representing fluctuations in quality and as not being
scientifically constructed. The study was restricted to prices se
cured currently from the above-named or similar sources.
BASE PERIOD.
The 10 -year period 1889 to 1898 was taken as the base period.
reasons were assigned for this selection.
No
PRICES: H O W SH O W N AND COMPUTED.
Three tables show the prices involved in the computation of index
numbers. The first shows average monthly prices for the current
year, the second shows average yearly prices for the 20 -year period
ending with the current year, and the third shows relative prices
for each year of the 10 -year period ending with the current year.
All actual prices shown are averages. A tabular statement in the
first report (February, 1879) gave for each of the 26 markets then
included in the study the intervals at which prices for the Imperial
Statistical Office were determined and the methods of determining
the quotations. According to this statement the average actual
prices represent great variations from market to market in the
number of original quotations involved in the computation, some
being based on daily determinations while others are based on weekly
or even monthly determinations, and some representing a medium
price or quality while others are averages of the prices o f the highest
and lowest or of the highest, medium, and lowest grades of the com
modities reported.
A few series of index numbers represent interrupted series of actual
prices and a few others represent series of actual prices whose com
parability is broken within the period involved in the table.
NUMBER AND CLASS OF COM M ODITIES.
From the beginning (1899) 44 series of index numbers were given.
Of this number three series represented iron and two each represented
coal and petroleum. Each oi the other commodities was represented
by a single series. The actual number of commodities was therefore
40. No index number for the total of the 44 series of index numbers
was published, nor were index numbers shown for groups of com
modities.
The index numbers for the 40 articles represented 235 series of
actual prices. The number was originally 238, the two Stuttgart
quotations for cotton yam and the Stuttgart quotation for cotton
goods having been dropped from both actual and relative tables.
The table of actual average yearly prices as published in 1913 con
tains 320 series of quotations, some of which are themselves calcu
lations from more than one variety, as, for example, the first Mann
heim quotation for barley. This table also includes five commodi
ties not represented in the tables of relatives, namely, raw sugar,
refined sugar, molasses, cocoa, and rubber. The comparability of
all series of sugar quotations has been interrupted by changes in
tariff laws; there is no continuous series of yearly average prices
on molasses for the base period, and the last two articles have been
added to the list of commodities since the publication of index num
bers was begun— cocoa in 1907 and rubber in the following year.
GERMANY---- FORMER IM PERIAL STATISTICAL OFFICE.
219
The commodities included are not classified into raw materials and
manufactured products. The great majority are raw materials but
a number are so-called semimanufactures ( Halhfabrikaten) .
Some commodities originally omitted from the list were considered
desirable but were not included because satisfactory data could not
be secured. The original number of articles (30) has been consider
ably increased, but still certain important articles, as, for example,
lumber and flax, are not included even in the tables of actual prices.
DESCRIPTION AND GROUPING OF COMMODITIES.
Index numbers are shown only for single commodities and not for
groups. The description of commodities as published in 1917 follows:
Rye (1,000 kilograms).
Berlin, good, m inim um 712 grams per liter [51.3 pounds per bushel].
Breslau, m edium grade.
Danzig, goods for free exchange ( Ware z.freien Verlcehr).
Frankfort on the Main, m inim um 70 kilograms per hectoliter [54.4 pounds per bushel].
Ham burg, Russian, in bond.
Konigsberg, good, 714 grams per liter [51.5 pounds per bushel].
Leipzig, German, good.
Liibeck , Russian, 71.3 kilograms per hectoliter [55.4 pounds per bushel].
Mannheim , various origins, m edium .
M unich, Bavarian, best.
M unich, Bavarian, good m edium .
Wheat (1,000 kilograms).
Berlin, good, m inim um 755 grams per liter [54.4 pounds per bushel].
Breslau, m edium grade.
Dtanzig, goods for free exchange.
Frankfort on the M ain, m inim um 75 kilograms per hectoliter [58.3 pounds per bushel].
Ham burg, Holstein, M ecklenburg.
Konigsberg, good, 749 to 754 grams per liter [54.0 to 54.4 pounds per bushel}.
L eipzig, German, good.
L indau, 78 to 79 kilograms per hectoliter [60.6 to 61.4 pounds per bushel], various
origins.
Mannheim, various origins, m edium . .
M unich, Bavarian, best.
M unich, good m edium .
Oats (1,000 kilograms).
Berlin, good, m inim um 450 grams per liter [32.4 pounds per bushel].
Breslau, m edium grade.
D anzig, domestic.
Frankfort on the M ain, good, native.
Konigsberg, good, 447 grams per liter [32.2 pounds per bushel].
L eipzig, German, good.
Lindau, Bavarian, 44 to 45 kilograms per hectoliter [34.2 to 35.0 pounds per bushel].
Mannheim, from Baden, from W u rt tern berg, m edium .
M unich, Bavarian, best.
M unich, Bavarian, good m edium .
Corn ( L,000 kilograms),
Bremen, American, best, in bond.
Breslau, Russian, m edium grade.
Hamburg, American, in bond.
Leipzig, various origins.
Barley (1,000 kilograms].
Breslau, m edium grade.
D anzig, brewing, domestic.
Frankfort on the M ain, brewing.
Konigsberg, 647 to 652 grams per liter [46.6 to 47.0 pounds per bushel].
Leipzig, German, good.
220
INDEX NUM BERS OF WHOLESALE PRICES.
Lindau, Hungarian, 65 to 66 kilograms per hectoliter [50.5 to 51.3 pounds per bushel].
Madgeburg, Chevalier, good m edium . (N ot after 1912.)
M annheim , from Baden, from the Palatinate, m edium .
M unich, Hungarian, Moravian etc., best.
M unich, Bavarian, best.
M unich, Bavarian, good m edium .
Hops (100 kilograms, ivithout wrappings).
Nuremberg,
Nuremberg,
Nuremberg,
Nuremberg,
Nuremberg,
Nuremberg,
market.
mountain.
W urttem berg.
Hallertauer.
Hallertauer seal,
Spalt.
Potatoes {1,000 kilograms, without sack).
Berlin, early red, for distilling.
Berlin, early red, for food, assorted.
Breslau, good Silesian, food.
Madgeburg, Saxon, for food.
Madgeburg, distilling.
Stettin, sorted, red, for food.
Stettin, sorted, white, for food.
Live stock for slaughter.
Cattle (100 kilograms [220.5 pounds]) Berlin, slaughter weight.73
Hogs (100 kilograms [220.5 pounds]) Berlin, slaughter w eight.73
Calves (100 kilograms [220.5 pounds]) Berlin, slaughter w eight.73
Sheep (100 kilograms [220.5 pounds]) Berlin, slaughter weight.73
Rye flour (100 kilograms with sack).
Berlin, N o. 0 /1 , good average grade.
Danzig, N o. 0 /1, dom estic price ( Inlands preis).
Cologne, N o. 0 /1 .
M unich, N o. 0.
Posen, dom estic, No. 0/1.
Wheat flour (100 kilograms).
Berlin, No. 00, w ith sack.
Danzig, No. 00, with sack, domestic price.
Cologne, R henish, No. 00, w ith sack.
L u b e ck , German, No. 0, w ithout sack.
M unich, Bavarian, No. 2, with sack.
Posen, dom estic, No. 00, w ith sack.
Butter (100 kilograms).
Berlin, I quality.
Berlin, I I quality.
M unich, finest Swiss.
M unich, m ountain.
Raw sugar (100 kilograms, net weight).1*
Brunswick, 88 per cent centrifugal, w ithout sack, 3 m onths’ tim e.
H alle, 88 per cent centrifugal, w ithout sack, 3 m onths’ tim e.
Magdeburg, I product, 88 per cent centifugal, without sack, 3 m onths’ tim e.
Stettin, 88 per cent centrifugal, w ithout sack, 3 m onths’ tim e.
Refined sugar (100 kilograms).74
Brunswick, w ithout container in paper.
Magdeburg, I loaf (Brot), w ithout container, in paper.
Stettin, I loaf ( Brot), w ithout container, in paper.
73 Slaughter weight (Schlachtgewicht) is the presumptive weight of the four quarters on which the price
of the animal, without the deduction of the value of hide, head, feet, entrails, etc., has been apportioned.
Prior to July 1 , 1897, quotations were not on slaughter weight. A t the time the change was made it
was stated that according to information from authoritative sources the quotations on slaughter weight
are about 8 £ per cent higher than on dressed weight.
7 * Descriptions from table of actual prices. Article not included in table of relative prices.
GERMANY---- FORMER IMPERIAL STATISTICAL OFFICE.
Molasses (100 kilograms, net weight).74
Magdeburg, for distilling.
Potato alcohol, crude (100 liters), alcohol.
Ham burg, with container.
Rapeseed oil (100 kilograms).
Berlin, crude, w ithout container.
Danzig, crude, with container, export price.
Frankfort on the M ain, with container.
Ham burg, crude, with container.
Cologne, crude, good and clear, with container.
Konigsberg, crude, clear, w ithout container.
Leipzig, crude, light color and clear, without container.
M annheim , standard q u ality, with container.
Herrings (1 cask, 150 kilograms).
Danzig, with container, Crown and full.
Danzig, with container, Crown, Ihlen.
H am burg, with container, in bond, Norwegian.
H am burg, with container, in bond, Scotch W est Coast.
Stettin, clear, with container, Norwegian, commercial.
Stettin, clear, with container, Norwegian, large m edium .
Stettin, clear, with container, Norwegian, fair m edium .
Stettin, clear, with container, Norwegian, m edium .
Stettin, clear, with container, Scotch, Crown, full brand.
Stettin, clear, with container, Scotch, Crown, M atfulls.
Stettin, clear, with container, Scotch, Crown, Ihlen.
Coffee (100 kilograms).
Brem en, clear, with sack, in bond, Sabanilla, fair ordinary.
Brem en, clear, with sack, in bond, Santos, good average.
Ham burg, net weight, in bond, Santos.
Ham burg, net weight, in bond, R io.
Ham burg, net weight, in bond, Campinas.
Ham burg, net weight, in bond, L a Guaira, unwashed.
Cologne, net weight, with sack, Java, good m edium.
Cologne, net weight, with sack, Santos, good m edium.
Mannheim , Santos, average quality.
Frankfort on the M ain, with sack, Santos choice.
Frankfort on the M ain, with sack, blue Java or Central Am erican.
Cocoa {100 kilograms, in bond).14,
Ham burg,
H am burg,
H am burg,
H am burg,
H am burg,
Ham burg,
Akkra current.
St. Thom e, fine.
Bahia, fair, fermented.
Trinidad current.
Samina current.
Arriba, choice, summer.
Tea (1 kilogram, in bond).
Hamburg, Kongo, Foochow.
Ham burg, Kongo, Shanghai.
Ham burg, Souchong.
Konigsberg, common Moning.
Konigsberg, fine Moning.
Konigsberg, finest Moning.
Rice (100 kilograms, in bond).
Bremen, Rangoon, shelled, 4 m onths’ tim e.
Brem en, broken, No. 0, shelled, 4 m onths’ tim e.
Ham burg, Rangoon, shelled, highest price, 1 per cent discount.
Ham burg, Rangoon, shelled, lowest price, 1 per cent discount.
Hamburg, broken, shelled, lowest price, 1 per cent discount.
74 Descriptions from table of actual prices. Article not included in table of relative prices.
221
222
IN DEX NUM BERS
OF WHOLESALE PRICES.
Pepper (100 kilograms, in bond).
Bremen, Singapore, 4 m onths’ tim e.
Hamburg, Singapore, 1 per cent discount.
Lard (100 kilograms, in bond).
Bremen, refined Am erican, 4 m onths’ tim e.
Leaf tobacco {100 kilograms).
Brem en, with packings, in bond, K en tuck y, ordinary, 6 m onths’ tim e.
Bremen, with packings, in bond, Brazil, 6 m onths’ tim e.
Bremen, with packings, in bond, V irginia, stems, 6 m onths’ tim e.
Hamburg, in bond, Domingo, wrapper and filler leaves, 6 m onths’ tim e or 2J per cent
discount.
Hamburg, in bond, B razil, 6 m onths’ tim e or 2J per cent discount.
M annheim, in bond, wrapper leaves, Palatinate, 6 m onths’ tim e or 1 per cent discount.
Mannheim, in bond, wrapper leaves, and filler leaves, Palatinate, 6 m onths’ tim e or
1 per cent discount.
Mannheim, in bond, cut, Palatinate, 6 m onths’ tim e or 1 per cent discount.
Hides and shins.
Bremen, 100 kilograms, ox hides, best dry, Buenos Aires, 6 m onths’ tim e.
Bremen, 100 kilograms, ox hides, Buenos Aires, Saladero, 6 m onths’ tim e.
Bremen, 100 kilograms, k ip hides, Durbunga, arsenic slaughtered, 6 m onths’ tim e.
Bremen, 100 kilograms, kip hides, H u gli, slaughtered, 6 m onths’ tim e.
Bremen, 100 kilograms, kip hides, Bakka, best, 6 m onths’ tim e.
Ham burg, 100 kilograms, ox hides, R io Grande, salted.
Hamburg, 100 kilograms, ox hides, dry, W est Indian, Central Am erican, etc.
Cologne, 100 kilograms, ox hides, best, green, Uruguay, 6 m onths’ tim e.
Cologne, 100 kilograms, kip hides, dry, East Indian, best Bakka, 6 m onths’ tim e.
M unich, 100 kilograms, ox and cow hides, best, green.
Frankfort on the M ain, 100 kilograms, calfskins, 3 to 4 m onths’ tim e.
Frankfort on the M ain, 100 kilograms, goatskins, 3 to 4 months’ tim e.
Frankfort on the M ain, 100 kilograms, hare skins, 3 to 4 m onths’ tim e.
Leipzig, 500 skins, hare skins, German.
Leipzig, 500 skins, hare skins, Russian.
Wool ( 1QO kilograms).76
Berlin, North German sheep, m edium .
Bremen, washed, Buenos Aires, I.
M unich, South German, m edium .
Leipzig, top, Australian, A .
Leipzig, top, La Plata, choice.
Leipzig, top, La Plata, half blood.
Leipzig, top, La Plata, quarter blood.
L eipzig, noil, Australian, current.
Leipzig, noil, La Plata, m edium .
Cotton (100 kilograms).
Bremen, m iddling upland.
Bremen, good Oomrawuttee, I I .
Hamburg, New Orleans, m iddling.
Cotton yarn (1 ldlogram).
Augsburg, 36 warp, 42 woof.
Augsburg, 20 warp, 20 woof.
Krefeld, English N os., 40-12 0 , twofold, singed, cash, 6 per cent discount.
Krefeld, English N os., 130-200, twofold, singed, cash, 6 per cent discount.
M iilhausen, in Alsace, metric Nos., warp N o. 16, 30 days, 2 per cent discount.
Mtilhausen, in Alsace, metric N os., warp N o. 28, 30 days, 2 per cent discount.
Miilhausen, in Alsace, metric N os., card. Maco No. 4 0, 30 days, 2 per cent discount.
75Descriptions for quotations at Munich and Leipzig taken from actual prices. Wool quoted ki these
two markets is not included in the table of relative prices.
GERMANY— FORMER IMPERIAL* STATISTICAL OFFICE.
223
Miilhausen, in Alsace, metric N os., woof N o. 16, 30 days, 2 per cent discount.
Miilhausen, in Alsace, metric N os., woof No. 37, 30 days, 2 per cent discount.
Miilhausen, in Alsace, metric Nos., card. Maco No. 50, 30 days, 2 per cent discount.
M unich-G ladbach, M ule No. 8, f. o. b. factory, 3 m onths’ tim e.
M unich-G ladbach, W ater No. 12, f. o. b. factory, 3 m onths’ tim e.
M unich-G ladbach, W ater N o. 20, !. o. b . factory, 3 m onths’ tim e.
Calico (1 meter).
M iilhausen, in Alsace, 90 centimeters [35.4 inches] w ide, 20/20 thread.
M unich-Gladbach, 78 centimeters [30,7 inches] wide, 16/16 thread.
Linen yarn (1 kilogram).
B ielefeld, English N os., average price for I and I I , No. 30, flax y a m .
B ielefeld, English N os., average price for 1 and I I , No. 50, flax yarn.
Bielefeld, English N os., average price for I and I I , No. 10, tow y a m .
Bielefeld, English N os., average price for 1 and I I , N o. 20, tow y a m .
Landeshut, in Silesia, English Nos., average price for I , No. 30, flax y a m .
Landeshut, in Silesia, English Nos., average price for I , No. 50, flax yarn.
Landeshut, in Silesia, English Nos., average price for I , No. 10, tow ya m .
Landeshut, in Silesia, English N os., average price for 1, N o. 20, tow y a m .
Raw silk (1 kilogram).
K refeld,
Krefeld,
Krefeld,
Krefeld,
Krefeld,
Krefeld,
Italian organzine, 18-20, 9 m onths’ tim e, or cash 5 per cent discount.
Italian tram, 2 4-26 , 9 m onths’ tim e, or cash 5 pei* cent discount.
Italian raw (grege.), 12-14, 9 m onths’ tim e, or cash 5 per cent discount.
Japanese organzine, 22-24, 9 m onths’ tim e, or cash 5 per cent discount.
Japanese tram, 3 4-4 0 , 9 m onths’ tim e, or cash 5 per cent discount.
Chinese tram, 3 6 -4 0 , 9 m onths’ tim e, or cash 5 per cent discount.
jHemp (100 kilograms).
Liibeck, Petersburg dressed hemp.
Mexican fiber (100 kilograms).
Hamburg, in bales.
T) pi
Hamburg, Brand
Raw jute (100 kilograms).
^
Hamburg, Good I, native brands.
Hamburg, Good I I , native brands.
Rubber, crude (Jt kilogram).1*
Hamburg,
Hamburg,
Hamburg,
Ham burg,
Ham burg,
Hamburg,
Hamburg,
H am burg,
H am burg,
Ham burg,
H am burg,
South Kam erun.
Benguela I I .
Upper Kongo I.
Kassai I, red.
Massai.
M ozambique I .
fine Para, hard.
Manaos, Negro, heads.
Cam eta.
P e n m a n balls.
M exican gum .
Iron (1,000 kilograms),
Germ an, pig:
Breslau, at the foundry, puddle,
Breslau, at the foundry, foundry pig.
Dortmund, at the foundry, Bessemer.
Dortm und, at the foundry, puddle-1.
D ortm und, at the foundry, Thomas.
Diisseldorf, at the foundry, puddle.
Diisseldorf, at the foundry, foundry pig.
Diisseldorf, at the foundry, Luxem burg No. 3.
^ Descriptions from table of actual prices.
Articles not included in table of relative prices.
224
INDEX NUM BERS OF WHOLESALE PRICES,
E nglish, pig:
H am burg, Scotch No. 1.
H am burg, Middlesboro N o. 1.
Swedish, bar:
L iibeck, I Stockholm.
Lead (100 kilograms).
Berlin, various German brands.
Frankfort on the M ain, R henish, double refined.
H alberstadt, refined, H arz, soft.
H alberstadt, refined, Silesian, soft.
H am burg, H arz, soft, double refined.
Cologne, R henish, soft, double refined.
Copper (100 kilograms).
Berlin, Mansfeld.
Berlin, foreign I , B ed e brand.
Frankfort on the M ain, Germ an double refined, in sheets.
H am burg, English, best selected.
Zinc (100 kilograms).
Breslau, good, Silesian.
Frankfort on the M ain, refined, zinc blend.
Halberstadt, Rhenish W estphalian, crude.
Ham burg, Silesian, in sheets.
Cologne, R henish, crude, “ W H und S S .”
Tin (100 kilograms}.
Frankfort on the M ain, Banca.
H am burg, Banca, in blocks.
Hard coal ( 1,000 kilograms).
Germ an:
Breslau, p it price, Lower Silesian, gas.
Breslau, pit price, U pper Silesian, gas.
Dortm und, at the m ine, fallen, lum p (run of m ine).
Dortm und, at th e m ine, puddle.
Dusseldorf, at the m ine, open-burning.
Duseldorf, at the m ine, anthracite.
Dusseldorf, at the m ine, uninflam m able.
Dusseldorf, at th e m ine, gas.
Essen, at th e m ine, open-burning.
Essen, at th e m ine, anthracite.
Essen, at the m ine, uninflammable.
Essen, at the m ine, gas.
Saarbrucken, at the m ine, open-burning.
Saarbrucken, at the m ine, anthracite.
E nglish:
D anzig, f. o. b ., E nglish, pea.
Danzig, f. o. b ., Scotch, m achine.
H am burg, f. o. b ., W e s t H artley.
H am burg, f. o. b ., Sunderland.
Petroleum (100 kilograms), with container, tare 20 per cen*,
Am erican:
Standard w hite, Berlin.
Standard white, Danzig.
Standard white, H am burg in bond.,
Standard white, Magdeburg.
Standard white, M annheim .
Standard w hite, Posen.
Standard w hite, S tettin.
R u ssian :
Breslau.
L ubeck, “ N o b e l.”
SUBSTITUTIONS, ADDITIONS, AND OMISSIONS.
Within the period of years covered by the index numbers three
series of relatives have been changed, namely, those for hides and
skins, cotton yarn, and cotton cloth. The change in the series for
GERMANY— FORMER IMPERIAL STATISTICAL OFFICE.
225
hides and skins was occasioned by the substitution in 1909 of a new
set of quotations for Frankfort hare skins (German and Russian)
with no alteration in the description of the article and with some of
the earlier actual average prices identical with the old figures. The
data for six years of the base period are incomplete for the new
series. No reason was assigned for the substitution of the new
series and no explanation was given in regard to its source. The
later series, being published in the 10-year table 1900-1909, does
not include the first year for which index numbers have been regu
larly shown. Minor substitutions of varieties or brands which
apparently do not affect prices are occasionally indicated in footnotes
to the tables (e. g., the Mannheim quotations on oats as given in the
tables published in 1912). Actual price series printed in “ old-style”
type on account of a break in the comparability of the figures are in
a number of cases represented by relative series (e. g., the Cologne
coffee quotations).
The changes in the series of index numbers for cotton yarn and
for cotton cloth were occasioned by the discontinuance of quotations
from the Stuttgart market. In this case new series of relatives
with the Stuttgart quotations eliminated were constructed for the
whole period covered by the index numbers. Series of actual prices
with data lacking for one or more years are in several cases repre
sented by series of relative prices.
Only two new commodities, cocoa and rubber, have been added
in the period covered by the index numbers. In these cases only
actual prices are shown.
TABLE OF RESULTS.
As has been stated on a preceding page, index numbers made by
averaging the different series of relative prices for a particular com
modity were published in former years. Beginning with 1913,
however, this practice was discontinued by the German Statistical
Office. Therefore, in order to bring the information up to the
latest possible date it has been necessary, in the preparation of the
present 'bulletin, to compute such index numbers for the years 1913
to 1917 by adding the several series of relative prices shown for a
commodity and dividing the sum so obtained by the number of
price series.
It should be noted, moreover, that during the war most of the
commodities for which prices formerly were collected by the German
Statistical Office were under Government control and all free trading
in them was prohibited. In a number of instances commodities
imported from abroad or manufactured from imported raw materials
disappeared entirely from the markets. For these reasons prices for
a considerable number of commodities could not be obtained after
1914 and many of the series of index numbers were in consequence
interrupted.
Table 48 is taken from more than one report, because no table of
relatives as published covers more than 10 years. In cases where
the earlier and later series do not agree, both sets of figures have
been copied.
3 3 2 2 6 °— 21— B u ll. 284--------15
226
INDEX NUM BERS OF W HOLESALE PRICES.
T a b l e 4 8.—R E L A T I V E P R IC E S OF
[Data from Vierteljahreshefte zur Statistik des
(Average prices in 1889-1898=100.)
Commodity.
W heat.............. .......................................
Oats..........................................................
Com (maize)...........................................
Barley......................................................
Hops.........................................................
Potatoes...................................................
Cattle........................................................
Hogs..........................................................
Calves.......................................................
Sheep........................................................
Rye flour.................................................
Wheat flour.............................................
Butter......................................................
Potato alcohol........................................
Rapeseed oil...........................................
Herrings........................... ......................
Coffee........................................................
Rice............. ............................................
Pepger.....................................................
1899
100
91
98
91
99
118
93
101
91
(i)
106
97
91
100
102
90
129
53
93
106
147
82
103
Leaf tobacco............................................
Hides and skies..................................... 1 3 1 0 7
127
W ool.........................................................
79
C o tto n .................... ............. ................
f < 92
Cotton ya m ............................................. \ &91
J 0 86
\ 5 86
96
Linen ya m ..............................................
Raw silk..................................................
109
112
Hemp.......................................................
106
Mexican fiber.........................................
98
Raw jute..................................................
122
Iron, German, pig.................................
Iron, English, pig...... ...........................
128
120
Iron, Swedish, bar..................... ..........
130
Lead.........................................................
Copper.......... ...........................................
149
131
Zinc...........................................................
153
106
Coal, anthracite, German.....................
Coal, anthracite, English.....................
113
106
Petroleum, American...........................
102
Petroleum, Russian..............................
1900
1901
1902
1903
1904
97
88
96
103
96
90
103
104
92
0)
108
95
86
100
97
109
144
63
92
104
168
105
108
2 113
3 HO
117
120
4 117
5 116
6 107
5 108
118
102
124
112
114
153
145
148
149
148
108
166
120
159
110
106
96
94
101
106
98
81
87
102
107
0)
109
94
92
105
81
107
116
56
84
104
164
127
108
2 111
3 107
97
94
108
116
95
88
81
106
114
0)
117
95
92
101
71
100
128
55
83
95
161
152
101
a 122
« 118
104
105
* 105
6 104
«94
s 94
102
99
128
98
103
106
115
115
97
108
98
150
116
118
103
91
91
90
93
106
92
138
102
113
96
0)
128
89
89
105
86
90
106
51
87
104
167
126
93
2 117
3 115
117
128
4 121
&121
6110
5 111
110
107
121
117
111
105
108
113
100
120
110
158
112
114
110
99
90
98
92
108
94
166
133
115
94
0)
123
87
94
107
127
85
99
60
90
101
159
105
91
2 121
3 118
118
144
4 131
5 131
6 118
5 118
120
93
114
128
113
104
102
112
102
119
118
158
111
108
104
98
Q4
102
4 105
5 104
6 89
5 89
120
91
135
97
107
115
113
118
112
141
91
148
123
125
106
97
1905
1906
98
98
102
115
101
107
126
120
123
0)
134
91
94
112
107
87
136
62
81
100
155
109
92
107
100
114
119
103
72
83
129
128
0)
146
101
97
115
91
105
146
62
87
103
143
130
103
2 132
2 146
s 129
3 142
134
123
114
130
4 108
5 117 V 134
6 10 8
5 109 | 127
130
116
99
107
114
125
122
121
194
149
104
119
108
117
114
118
118
149
141
176
133
141
180
225
113
118
108
113
106
99
96
102
1 No index number published.
New series occasioned by substitution in Frankfort quotation for German and Russian hare skins,
a Old series.
2
GERMANY— FORMER IMPERIAL STATISTICAL OFFICE.
227
AR TICLES OF W H O L E S A L E T R A D E .
Deutschen Reichs: 1917 and preceding years.]
(Average prices in 1880-1898=100.)
1907
127
116
125
131
114
178
122
128
106
0)
144
122
112
114
119
135
115
59
94
115
123
335
121
2 137
3 134
138
135
159
145
157
133
132
117
185
136
126
118
165
179
126
216
127
137
108
106
1908
123
118
114
140
118
54
121
121
112
0)
136
120
115
119
143
129
92
60
88
113
87
113
118
2 125
3 122 \
121
122
137
119
132
96
128
113
152
119
112
115
116
121
107
168
133
129
113
112
1909
1910
1911
315
130
119
139
112
97
115
115
128
0)
137
no
126
120
108
107
107
62
87
107
90
171
112
143
132
137
131
101
116
107
127
103
142
98
127
127
O)
143
97
113
123
107
107
121
74
94
105
103
183
129
156
135
172
149
m
114
120
127
121
195
146
134
110
167
146
106
111
129
104
118
124
97
102
116
122
136
141
144
129
159
142
125
121
134
147
129
184
165
145
141
181
160
115
115
134
141
123
145
107
102
142
143
156
123
180
131
142
136
113
110
99
130
100
128
101
110
115
112
119
119
169
129
118
110
106
121
124
96
138
104
131
107
112
124
111
116
124
193
127
120
105
103
119
134
95
145
114
180
108
110
123
120
114
135
238
125
118
105
104
123
134
93
167
113
191
123
129
123
154
148
141
259
130
139
125
120
4 New series including Stuttgart quotations.
5 Old series including Stuttgart quotation.
e New series excluding Stuttgart quotations.
1912
1913
.
110
112
"13
126
107
147
111
151
140
195
176
107
111
127
151
122
157
93
105
121
134
166
127
197
144
150
134
124
145
98 i
169
115
233
130
132
12»
160
139
124
255
139
146
126
125
1914
1915
1916
;
1917
119
122
127
119
137
129
145
119
174
177
121
125
128
152
134
190
89
138
144
157
127
181
185
141
54
228
193
59
286
94
372
143
153
183
14a
171
207
308
29a
132
210
645
751
123
2,390
7,283
209
148
455
229
208
236
255
151
167
173
273
139
234
351
320
121
150
103
180
190
211
116
239
194
337
316
155
180
256
158
268
194
175
159
172
268
194
196
293
227
152
161
155
259
123
122
124
172
141
120
206
137
134
126
126
133
247
228
INDEX NUMBERS OF WHOLESALE PRICES.
INDEX NUMBERS OF THE JAHRBUCHER Ft)R NATIONALOKONOMIE
UND STATISTIK.
PUBLICATION.
Three general “ indexes” and one table of index prices for which no
general relative is computed, all based on German wholesale price
statistics, are found in former issues of the Jahrbiicher fur Nationalokonomie und Statistik, published monthly at Jena, Germany. They
are the work of Dr. Johannes Conrad, a professor of the University
of Halle, although in the years through which the index numbers
have been carried, different persons have assisted in their com
pilation.
No distinctive name, so far as known, has become identified with
any of the three, though they are variously spoken of as “ Conrad’s
indexes” and the “ Jahrbiicher indexes,” and are now and then
referred to as the “ Hamburg indexes.”
These index numbers appeared in published form about two years
late, and with some irregularity. Thus the figures for 1911 appeared
in the August, 1913, issue, and those for 1910 in the issue of July, 1912.
From 1887 to 1914, however, presentation of the figures at some time
during the year was made. Each issue of the Jahrbiicher reproduced
the figures from the beginning, though those for the earlier years
were grouped by periods in the later issues. No index numbers ap
pear to have been published in the Jahrbiicher since 1914.
HISTORY.
The following is a translation of the history of these price studies as
given by Dr. Conrad in volume 17 of the Jahrbiicher, third series*
1899, page 642:
In these Jahrbiicher, in volum e 3, 1864, appeared for the first tim e the results of
an investigation into the course of prices, based upon the Ham burg quotations on
imported goods subject to taxation. Prof. Laspeyres was the author. This study
was based upon the one already made b y Soetbeer which brought the data up to 1856,
Laspeyres carrying the figures to 1862 for 48 articles. H e compared the years 18511862 with both of the preceding decades, taldng the arithmetical mean of the prices,
in order to demonstrate the alteration of prices in consequence of the gold importation
resulting from the discovery of gold in California.
In the year 1874, in volume 23, Prof. H . Paasche, then a student, at m y suggestion
carried the investigation further upon the basis of the same materials but b y a different
method. Instead of computing the arithmetical mean he reckoned the quantities
of 22 articles consumed b y the population in the current year and m ultiplied the
quantity of each article b y the average price for the basic period (1847-1867) and b y
that for the current year, in order to get a more accurate relative.
Again in the year 1882 Richard van der Borght, now Prof. van der Borght, took up
this computation according to the same principles and for the same articles and,
printed the results of his investigations in volum e 5 (new series), 1882. However
he used other figures for the consumption quantities, since statistics on this subject
in the meantime had improved.
In the year 1887 we ourselves in volume 15 (new series) utilized the Ham burg
quotations for an investigation of the price reduction in the eighties, and carried the
data forward in the same manner. Since then we have annually compared the prices
of the current year with those for preceding years, in order to determine how the
m ovem ent of prices has further developed.
SOURCE OF QUOTATIONS.
Two of Dr. Conrad’s indexes and one table of 47 articles for which
a general index is not printed are based on the actual yearly average
prices appearing in the publication entitled “ Hamburg’s commerce
GERMANY— JAHRBUCHER FUR NATIONALOKONOM IE.
229
and shipping” (Hamburgs Handel und Schiffahrt), issued by the
Hamburg Bureau of Trade Statistics (Handelsstatistisches Amt) .
This publication contains actual prices, per 100 kilograms, net,
stated in marks. The report for the year 1911 contains prices for 174
articles and subdesignations of articles.
The price is that of sea
borne commerce declared at entry at the port of Hamburg.
In the first study published in the Jahrbucher, that appearing in
volume 3, 1864, by Dr. E. Laspeyres, professor in the University of
Basel, and entitled “ Hamburg prices from 1851 to 1863, and the Cali
fornia and Australian gold discoveries since 1848 ” (Hamburger Warenpreise 1851-1863 und die californisch-australischen Goldentdeckungen
seit 1848), Dr. Laspe}7Tes stated that the prices for the 48 articles used
by him in preparing his relative had appeared every Friday since the
year 1736 m the official “ General price-current” (.Allgemeiner Preis
Courant). However, so far as known, no use of them for purposes of
a relative had been made prior to Prof. Soetbeer’s compilation
beginning with 1831. In volume 23, 1874, Prof. Paasche stated that
the current report on Hamburg’s commerce and shipping for that year
contained prices for more than 300 articles for the years 1847-1872.
Prof. Soetbeer, in a study entitled “ The movement of prices in the
years 1886-1890” (Das Niveau der Warenpr.eis in den Jahren 18861890), published in the Jahrbucher, volume 58 (3d series, vol. 3), 1892,
made the same statement for the years 1886-1888. In other words,
no material change had occurred in the method of quoting Hamburg
prices from 1847 to 1888.
During all these years Hamburg had been a free port, collecting
duties on all goods entering the city, even if they came from other
States of Germany, and therefore the prices declared at entry on all
goods, whether received into the city by sea or river, by rail or
wagon, appeared in the official price statistics. On October 15, 1888,
however, Hamburg entered the German customs union (Zollverein) .
The following statement as to the effect of this change on the Ham
burg quotations is abstracted from Prof. A. Soetbeer’s statement in
the Jahrbiicher fur Nationalokonomie und Statistik, volume 3 , third
series, 1892, pages 590, 591:
W ith the entry of Hamburg into the German customs union the Hamburg trade
statistics underwent a fundamental alteration which no longer permitted of a direct
comparison with former years. The quotations upon articles of domestic commerce—
i. e ., articles received from neighboring German States— ceased, and quotations upon
imports entering b y sea only were available from that date. Therefore, for only those
articles which had previously been exclusively or in overwhelming proportion brought
in b y sea were the figures after 1888 comparable. If the long series of preceding cal
culations of average yearly prices was not to be finally terminated and an entirely
new series started, it would be necessary to make a complete revision and recomputa
tion of the preceding tables and a new computation of average actual prices to include
through the entire period from 1847 to 1888 only those articles w hich had been im
ported b y sea.
The great importance of the Hamburg quotations as a basis for price statistics being
generally recognized, and the great desirability of a continuous series from 1847 being
evident, the director of the Ham burg bureau of trade statistics decided to make this
recomputation. B y means of subsidiary material, b y use of the price declarations on
exports b y sea for the period 1847-1888, b y the utilization of trustworthy market
reports for the period, and in some degree b y the use of expert estimates, the average
actual prices were brought to the more restricted basis for the entire period.
Prof. Soetbeer concludes his discussion in these words: “ These com
bined means— i. e., of recomputation—have cost much work and
time, but one may* with all good confidence put trust in them that in
230
INDEX NUM BERS OF WHOLESALE PRICES.
spite of all difficulties success has been attained in securing the con
tinuity of this most important work on the movement of prices,
which is not only desirable but necessary,”
The effect of this change is discussed by Dr. Conrad in the Jahr
bucher for 1893 (3d series, vol. 6 , p. 695):
Unfortunately the first source— i. e ., Hamburgs H andel und Schiffahrt— has under
gone an alteration which has rendered necessary a reconstruction of the tables based
upon it. U p to the year 1888 th e prices were reckoned upon the customs declarations
of goods entering b y all means of transportation. A s a result of the entry of Ham burg
into the German customs union the customs declarations now embrace on ly goods
imported b y sea. A s a result of this a comparison w ith preceding years could not be
m ade with a n y accuracy. Now, however, the yearly average actual prices for 1 8471888 have been recomputed on the basis of 1888. This has been done b y restricting
the prices for 1847-1888 to those articles which were imported b y sea. This reduced
the num ber of articles upon which quotations were given from 320 to 180.
No changes of importance were made after Dr. Conrad’s article was
written. An average actual quotation was carried for each of 171
articles and subdesignations of articles for all years up to 1912, when
publication of the index numbers was discontinued.
The fourth series of relative prices published in the Jahrbucher and
prepared by Dr. Conrad is not based upon the famous Hamburg
wholesale price quotations but upon official statistics of the German
customs union. The history and source of this series are separately
treated on page 238 of the present bulletin.
From this point it will be necessary to consider separately the three
series of index numbers based on the Hamburg trade statistics which
Dr. Conrad presented annually in the Jahrbucher up to 1914.
Index n um bers co m p u ted from prices o f 29 selected articles.
The first series contains relatives for each of 39 articles from 1871 to
1912 and appeared in current issues of the Jahrbucher. No general
relative was computed for this group. This is the table used in th©
Report from the Committee on Finance of the United States Senate,
1893.76
BASE PERIOD.
This table appears in the first study of this series, volume 3 of the
Jahrbucher, 1864, b y Dr. E. Laspeyres in practically its latest form.
For 42 of the 48 articles then included in it the period 1831-1840 was
used as a base. Relatives for 3 articles had 1841-1850 as a base,
while for 3 articles 1851-1853 was the base. A general relative, not
weighted, was computed with 1831-1840 as the base.
When the study of relative prices was resumed in the Jahrbucher by
Prof. Paasche, volume 23, 1874, the period 1847-1867 was used as a
base and the relative for each of 47 articles was computed, but no
general relative for this table was made.
Of the three articles for which Dr. Laspeyres had been forced to use
1851-1853 as a base, Dr. Paasche dropped two (soda and Java coffee),
while rapeseed appeared under a slightly different designation. He
makes no comment as to how he adjusted it to the basis of 1847-1867.
Dr. Richard van der Borght, by whom the work was continued in
1882, did not print this table at all. Its publication was resumed,
™Report from the Committee on Finance of the United States Senate on Wholesale Prices, Wages, and
Transportation. Mar. 3,1893. 52d Cong., 2d sess., R eport N o. 1394. P t. I, pp. 297-301.
GERMANY---- JAHRBITCHER FUR NATIGSTALOKONOMIE.
231
however, by Dr. Conrad in volume 1 , third series, of the Jahrbucher,
1891, with 1847-1870 as the base. This base remained unaltered
from that date,
PRICES: H O W SH O W N AND COMPUTED.
In the latest available presentation of the table, that published in
June, 1914, relatives are snown by articles and decades from 1871 up
to and including 1900. Relatives appear for 1901-1905, 1906-1910,
and for each o f the years separately from 1906 to 1912, inclusive.
By the use of the tables in preceding issues, beginning with volume 1 ,
third series, 1891, it is possible to get the decennial relative for each
article since 1851-1860 and the yearly relative from 1886 to 1912, on
the latest base. No yearly relatives are printed for the years 18641885, inclusive, and the yearly relatives for 1851 to 1863, appearing
in the Jahrbucher (vol. 3, 1864) are, as already noted, on a different
base. Their reduction to the new base has not been made except by
decades.
Actual average prices are also shown in the June, 1914, issue of the
Jahrblicher by zentners (50 kilograms) in marks, for 1847-1870, for
the decades 1871-1880, 1881-1890, and 1891-1900, for the five-year
periods 1901-1905 and 1906-1910, and by single years from 1906 to
1912. By the use of preceding issues, actual average prices b y years
are available, beginning with 1886.
A comparison with the official figures shows that from 1891 to
1912 the actual prices of the table under consideration are those of
Hamburgs Handel und Schiffahrt as published in its current issue,
reduced from the dopplezentner (100 kilograms) to the zentner
(50 kilograms). The data for years prior to 1891, however, are not
those shown in the current official publication but those published
previous to the complete recomputation of the official figures, which
was made after Hamburg entered the German customs union; in
other words, Dr. Conrad’s table of actual prices for years prior to
1888 has not been readjusted to the basis of 1888, as are the official
quotations published for Hamburg.
Two relatives are given for each article for the current year,
namely, the per cent which the average price for that year is of the
average price for 1847-1850 and for 1871-1880, respectively.
No general relative has ever been printed for this table except
as it first appeared in 1864, although the Committee on Finance of the
United States Senate in 1893 computed a relative for its own useJG
NUMBER AND CLASS OF COMMODITIES.
The table shown in the Jahrblicher for June, 1914, the latest avail
able, embraces 39 articles. As originally computed by Dr. Laspeyres
the table contained 48 articles, but Prof. Paasche in 1874 made a
number of changes in his selection and reduced the number to 47.
He brought the table from 1847 to 1872 to the basis of his selection,
however, and since he computed no general relative, the changes
were of no especial significance. As already stated, this table was
not printed from 1874 until Dr. Conrad resumed it in 1891. {Jahr
bucher, 3d series, vol. 1 , pp. 916, 917.) In it he used the same 47
articles that Dr. Paasche had used.
76Report from the Committee on Finance o f the U nited States Senate on W holesale Prices, Wages, and
Transportation. Mar. 3, 1893. 52d Congress, 2d session, R eport No. 1394. Pt. I, p p. 297-301.
232
INDEX NUMBERS OF WHOLESALE PRICES.
In later issues of the Jahrbucher actual prices as well as relatives
appear for all articles up to the year 1891, inclusive. In 1892, how
ever, neither actual prices nor relatives were shown for raw sugar,
silk, flax, hops, rapeseed oil, horsehair, or butter, and no prices or rela
tives have appeared for these articles since that year, although they
were not finally eliminated from the table until the issue of the
Jahrbucher for July, 1911. No explanation of the change has been
made.
All the articles embraced in this table are either raw materials or
semimanufactured materials.
DESCRIPTION AND GROUPING OF COMMODITIES.
No description of the articles in this table has accompanied it after
its first appearance in 1864. At that time Dr. Laspeyres gave a
detailed description, but his list of articles is not the same as is used
in later tables, and the time that has elapsed since its appearance
makes his description of no great present significance. It is there
fore not reproduced. It may be found in the Jahrbucher, vol. 3,
pp. 89-92, inclusive.
As the table now stands, the description of articles is an exact copy
of the description found in Hamburgs Handel und Schiffahrt with
the following exceptions:
Cotton.— The table as published in the Jahrbucher merely says
“ cotton” CBaumwolle), but in the original table the boxhead reads
“ Cotton and cotton waste” (Baumwolle und Baumwoll-abfall). The
actual prices are those of “ Commerce and shipping” reduced to a
basis o f 50 kilograms.
Coal.— Dr. Conrad’s designation is “ Coal and coke” (Steinkohlen
und Koks), but the official designation is “ Coal” {Steinkohlen). Dr.
Conrad’s actual prices are the same as the official figures, reduced to
zentners.
Iron, English har.— This designation does not appear at all in the
official Hamburg figures. “ Bar iron from all sources ” (Stangen-Eisen
im Ganzen) and “ English strap iron” (Stangen-Eisen, englisch) are
quoted. Comparison of the actual prices, as quoted by Dr. Conrad
in the Jahrbucher, shows that from 1881 to 1890 the quotation appear
ing under the heading “ English bar” is “ English strap iron.” From
1891 to 1905 it is the quotation of the official figures for the “ bar iron
from all sources.” From 1906 to 1910 it is again the quotation for
“ English strap iron.” So far as can be ascertained, no explanation
of this irregularity is offered.
No grouping of commodities is made in the table. No general
relative has been published since 1864. The relative for 1864 was
not weighted.
I/idex n u m bers co m p u ted from prices o f 19 articles in 6 groups.
A second table of weighted relatives for six groups containing 19
articles, together with a combined weighted relative for all articles,
appears in issues of the Jahrbucher up to 1914. The weighted
relative for all articles is in each issue compared with a simple relative
based on all the articles quoted in the official Hamburg prices for
which the continuity of the quotations is such as to make them usable.
This latter relative is discussed on pages 236-238 of this bulletin.
GERMANY— JAHRBUCHER FUR NATIONALOKONOM IE.
233
BASE PERIOD.
The weighted relative as published is calculated on two bases
for purposes of comparison: 1847-1880 and 1871-1880. The table
also contains a relative for the years 1871-1880 with the period
1847-1867 as a base.
This table first appeared in Prof. Faasche’s study published in vol.
23, 1874. At that time 1847-1867 was used as the base. This same
period was used as the base period by Dr. van der Borght in 1882, but
he also made a relative for the years 1876-1880, with 1847-1875 as a
base. When Dr. Conrad calculated this table in 1887 he reduced it
to its present form.
PRICES: H OW SHOW N AND COMPUTED.
As the table was published up to 1912, relatives are shown for the
10-year periods 1881-1890 and 1891-1900, for the 5-year periods
1901-1905, and 1906-1910, and for each year after 1905. By the use
of preceding issues a yearly relative for the 19 articles is made avail
able beginning with 1888. and also a relative by 5-year periods begin
ning with 1881-1885.
NUMBER AND CLASS OF COMMODITIES.
The first issue of the table embraced 22 commodities, all raw
materials. As it appeared for 1912 it contained 19 raw materials.
The reduction from 22 to 19 articles really occurred in 1892, since
in the table of actual prices from which the relatives were made no
quotation appeared for sugar, silk, or unforged zinc after 1891. The
table, however, was preserved in its original form until volume 31 of
the Jahrbucher, 1906 (containing the relative for 1904), when the
table was recast to include only the 19 articles. No explanation of
the reason for the omission of the three articles noted above was
made, but since they were later not quoted in Hamburgs Handel und
Schiffahrt, their omission from the official statistics undoubtedly
forced their omission also from Dr. Conrad’s weighted relative.
DESCRIPTION AND GROUPING OF COMMODITIES.
The articles used by Dr. Conrad are described in his table exactly
as they appear in their original source (Hamburgs Handel und
Schiffahrt), with four exceptions:
1. Rice.— The quotation on rice is that for all rice, without dis
tinction of kind.
2 . Cotton.— In the Jahrbiicker the quotation under the heading
“ Cotton” (Baumwolle) is the price appearing in Hamburgs Handel
und Schiffahrt under the heading “ Cotton and cotton waste” {Baumwolle und BaumwoTl-abfall).
3. Fish oil.—This article appears in the Hamburg tables merely as
“ Tran,” in the Jahrbucher as “ Fischtran.”
4. Coal.— The quotation in Hamburgs Handel und Schiffahrt is
for “ coal and coke” (Steinkohlen und Koks), while in the Jahrbucher
the heading used is merely “ Coal” 0Steinkohlen). The price, how
ever, is that for coal and coke.
Actual average prices are also shown in issues of the Jahrbucher up
to June, 1914, by zentners (50 kilograms) in marks from 1847 to date.
234
INDEX NUM BERS OF WHOLESALE PBICES.
The grouping of articles is as follows:
I. Coffee, cocoa, tea, pepper, and rice.
II. Cotton.
III. Indigo, saltpeter, fish oil, and palm oil.
IV. Iron (pig and cast), tin, copper, and lead.
V. Coal.
VI. Wheat, rye, barley, and oats.
WEIGHTING.
The index number under consideration was first weighted by Prof.
Paasche according to consumption in the German Empire in 1874.
He states that he determined the consumption quantities which he
used by various means. For those articles which were not pro
duced in Germany it was possible to get fairly accurate data from the
import and export statistics of the German customs union. The arti
cles which came to Germany exclusively by importation were coffee,
cocoa, tea, pepper,/ice, cotton, indigo, saltpeter, fish oil, and palm oil.
Domestic production of sugar being at that time subject to a tax, it was
possible to get accurate statistics of the consumption of sugar from the
official imperial statistics (Reidisstatistik) . The same was true of the
mineral products, iron, zinc, tin, copper, lead, and coal. Prof. Paasche
stated that the least satisfactory were his statistics for the consump
tion of grain. He found that it was absolutely impossible to make
accurate tables of grain production, and he therefore used the figures
of Hausner in his “ Comparative European Statistics7’ ( Vergleichends
Statistik Europas), 1864.
The weighted relative was computed according to the method
of Drobisch. To illustrate, the relative for each group for the year
1868 was computed as follows:
German consumption for that year of each article in the group was
ascertained as indicated above. It was then multiplied by the aver
age price for the base period, 1847-1867, and bv the average price
for 1868. The sum of the products for 1847-1867 was then ascer
tained and likewise the sum of the products for 1868. The relative
for 1868 was the ratio of the sum of the products for 1868 to the sum
of the products for 1847-1867.
The relative for the entire table of 22 articles (now 19 articles) in
six groups was made in the same way. That is, the sum of all the
products for 1847-1867 was found and the sum of the products for
1868 was compared with it to get the general relative. Thus, the
textile group consisted of cotton and silk. The average price of
cotton for 1847-1867 was 26.83 thalers per zentner; for 1868, 25.92
thalers per zentner. The consumption of cotton in 1868 was 1,509,961 zentners. The average price of silk for 1847-1867 was 616.31
thalers per zentner; for 1868, 858.12 thalers per zentner. The con
sumption of silk for 1868 was 22,088 zentners. The relative for the
group was therefore secured as below.
GERMANY---- JAHRBUCHER FUR NATIONALOKONOM IE.
2 6.8 3 X 1,5 0 9,9 6 1 equals...........................................
616..31X 22,088 equals...............................................
235
40, 512, 254
13, 613, 055
5 4,1 2 5, 309
3 9 ,1 3 8 ,1 8 9
18, 954,155
2 5.92X 1,509,961 equals
858.1 2X 2 2 ,08 8 e q u a ls ...
58, 092, 344
R elative for 1868 equals......................................... equals 107
o 4, iz o , oU y
The sum of the 22 products for 1847-1867 was 954,341,370; for
1868 it was 1,130,430,232. The relative for the 22 articles for 1868
,
113,043,023,200 .
, 11QK
therefore was — 9 5 4 3 4 f 379 — equals 118.5.
The table was reweighted for each year from 1868 to 1872, inclu
sive, according to the consumption for that year.
When Dr. Richard van der Borght computed this table in 1882,
he followed the same method as Prof. Paasche, weighting anew for
each year the figures for the years 1873 to 1880, inclusive. As to
the source of his statistics of consumption he states that for grains,
instead of using Hausner’s results for the years 1873-1877, he sub
stituted the figures of X . von Neumann-Spallart which had appeared
in his “ Review of International Traffic” in 1878. For the years
1878-1880 he used the imperial statistics of domestic grain produc
tion which had become available.
The estimates of consumption of other articles were all based on
official imperial statistics (ReichsstatistiTc), including statistics of
domestic production, and of export and import.
When i)r. Conrad computed the table for the years 1880 to 1886,
inclusive (Jahrbucher, vol. 15, pp. 322-331), he used the consumption
statistics of 1880 for the entire table, having abandoned Dr. van
der Borght’s method of weighting each year separately. Dr. Conrad
continued to weight this table according to the consumption statis
tics of 1880 up to 1912, the latest year for which figures were shown.
So far as can be ascertained, the general relative for the entire 19
articles was weighted for all years prior to 1906, inclusive, as indi
cated in the preceding paragraphs. From 1907 to 1912, however,
the general relative was obtained by computing the simple arith
metical average of the relatives for the six groups.
TESTING.
It can not be said that any method of testing was applied to Dr.
Conrad’s weighted relative prior to 1889. From that year, however,
the weighted relative has been compared to the unweighted relative
for all articles with continuous quotations included in Hamburgs
Handel und Schiffahrt. For 1889 this simple relative embraced 318
articles. In 1912 it included only 157 articles. Dr. Conrad also
included in his periodical price studies after 1889 a relative based
on the official imperial quotations (ReichsstatistiTc), and in addition
reprinted Sauerbeck’s index as found in the Journal of the Royal
Statistical Society, London. These various relatives he utilized for
comparison with his own work.
236
INDEX NUM BERS OF WHOLESALE PEICES.
Index num bers com p u ted from prices o f 157 articles.
Beginning with volume 1 , third series, of the Jahrbucher for 1891,
Dr. Conrad’s price studies have also included, as stated in the preced
ing paragraph, a simple relative for all the' articles contained in
Hamburgs Handel und Schiffahrt for which the quotations have
been continuous. The relative is printed in the same table as the
weighted relative described above and is used for purpose of compari
son with it.
BASE PERIOD.
The base periods are the same as for the weighted relative already
discussed. A relative for the years from 1871 to 1880, with 1847-1867
as the base, and two relatives for the years from 1881 to 1912 are
shown. Of the latter the first has as its base the period 1847-1880,
while the second is computed upon 1871-1880 as a base.
PRICES: H OW SHOW N AND COMPUTED.
In the issue of the Jahrbucher for June, 1914, relatives are shown
for each of the decades 1881-1890 and 1891-1900, for the five-year
periods 1901-1905 and 1906-1910, and for the years 1906 to i912
separately. By the use of preceding issues it is possible to get a
yearly relative beginning with 1886 and a relative by five-year periods
beginning with 1881-1885.
NUMBER AND CLASS OF COMMODITIES.
As printed in its latest form the index number includes 157 articles,
largely raw materials, but including also some manufactured and
semimanufactured articles, as varied in character as possible. A
yearly relative on the later basis of 157 articles is available only from
1902 to 1912, This is occasioned by the fact that when the relative
w^as recomputed on its later basis the years prior to 1902 were grouped.
When first computed for 1889 and preceding years 318 articles were
included. The relative for 1891 and preceding years, which appeared
in volume 5, third series, embraces 320 articles. However, by the
time that Dr. Conrad prepared his relative for 1892 the recomputation of the Hamburg prices necessitated by the entry of Hamburg
into the German customs union, to which repeated reference has been
made, had taken place, and Dr. Conrad found it necessary to recom
pute his relative on the reduced basis of 163 articles.
Dr. Conrad’s statement concerning this is as follows:
I t became necessary for us to reduce tlie number of articles included in our com
putation of the arithmetical mean from 320 to 163 articles, and on this basis we have
available figures from 1847 to date. This fact is bound to have an appreciable effect
on our results. The decrease in prices as a result of this in recent tim es is rendered
less important, while the rise in prices in the seventies compared to 1847-1867 becomes
greater. Thus with the old list of articles the relative for 1871-1880, with 1847-1867
as a base, is 104. W ith the restricted list it is 111. W ith 1847-1880 as a base the
average of 320 articles for the decennium 1871-1880 has a relative 79, bu t the relative
for the new list of 163 articles is 95.
The difference is yet greater for the years 1890-91, for which the relative according
to the old method on the base 1847-1880 was 74, bu t now is 95, and on the base 18711880 was 72 and is now 85 plus. The year 1892, so far as comparison is possible, with
the base 1871-1880 would have shown a decrease of from 70 for 1891 to 68 for 1892.
W ith the present list of articles, however, the relative for 1892 is 82.7 as against 87.4
for 1891.
This is a difference of about 15 per cent, arising from the fact that the prices of
domestic products had decreased m uch more than the prices of those imported b y sea.
GERMANY---- JAHRBUCHER EUR NATIONALOKONOM IE.
237
This difference will serve to warn against the putting of too great faith in these tables,
for th ey show how results differ according to the number of articles considered and
show how necessary it is to use as large a number of articles as possible.
With the publication of the relative for 1906 and preceding years,
in volume 34, third series, of the Jahrbucher, 1907, the number of
articles was further reduced to 158 and the relative was recomputed
from the beginning on the new basis, Dr. Conrad does not state
what articles were dropped. When the index for 1907 was printed
the number was still further reduced to 157, without recomputation.
DESCRIPTION AND GROUPING OF COMMODITIES.
No list of included articles is printed in the Jahrbucher. The 157
articles, however, are from the following list of 174 articles for which
average annual prices from 1850 to 1912 appear in “ Hamburgs Handel
und Schiffahrt.” The quotations for the articles marked with an
asterisk (*) are in some degree imperfect, and it seems safe to con
clude that the excluded articles are among those so designated:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Aloes.
A ntim ony.
Oranges.
Arrack.
Asphalt.
Balsam of copaiba.
Peruvian balsam.
Cotton and cotton waste.
B a y leaves.
T in plate.
Sheet iron.
Lead.
W h ite lead.
Borax.
Bristles.
Cinchona bark.
Cochineal.
D iv i-d iv i.
Iron wire.
Pig and cast iron.
Bar iron from all sources
(im
Ganzen).
22.
23.
24.
*25.
26.
27.
*28.
*29.
*30.
31.
32.
33.
34.
35.
36.
*37.
*38.
39.
*40.
* 4 i:
42.
43.
Strap iron, English.
Sheet billets, Swedish.
Iron rails and fishplates.
E lep h ant’s tusks and ivory.
Peas.
E xtract of logwood.
E xtract of redwood.
E xtract of yellow wood.
E xtract of quercitin (dyer’s oak).
Figs.
Deerskins, doeskins, and reindeer
skins.
Calfskins.
Sheep and goat skins.
Dried fish.
Meats, fresh and cured.
M eat extracts.
Nutgalls.
Cotton yarn.
Coconut fiber yarn.
Jute and m anila hemp yarn.
L inen varn.
W oolen and half-woolen yarn.
*44.
45.
46.
47.
48.
49.
50.
*51.
*52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
*69.
70.
71.
72.
73.
74.
75.
*76.
77.
78.
79.
80.
81.
82.
83.
84.
*85.
86.
87.
Casings (i. e ., for sausages).
Y ellow metal and brass.
G in.
Barley.
Plate glass.
Natural guano.
G um arabic.
G um benzoin.
Dammar.
Raw rubber.
Copal.
Gutta-percha and chicle.
India-rubber shoes.
Dried and salted hides.
Oats.
H em p .
Resin and galipot.
Herring.
Logwood.
Y ellow wood.
Redwood.
E bony wood.
Rosewood ( Jacciranda).
Corkwood.
Mahogany.
W alnut wood.
Cedar wood.
H oney.
O x and cow horns.
Indigo, natural and manufactured.
Raw ginger (zinziber).
Iodine and iodine preparations.
Jute.
Cheese.
Coffee, raw, without designation of
kind (im Ganzen).
Coffee, raw, Brazil.
Coffee, raw, San Domingo.
Coffee, raw, Java.
Coffee, raw, L a Guaira.
Coffee, raw, Porto Rico.
Cocoa.
Potassium monochromate.
Camphor.
Cinnamon.
INDEX NUM BERS O F WHOLESALE PRICES.
238*
.88,
89.
90.
*91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
*103.
104.
105.
106.
107.
*108.
109.
110.
*111.
112.
113.
114.
115.
116.
*117.
*118.
119.
*120.
121.
122.
123.
124.
*125.
126.
Cassia lignea and cassia vera.
Catechu, brown and yellow .
Bones.
Bone ash.
Bone charcoal and bone m eal.
Cognac.
Currants.
Corks.
M adder.
Capper.
Licorice*
Leather.
Candles.77
M aize.
A lm onds.
M anila hem p, sisal, etc.
Nutm eg flowers.
Nutm egs.
N ails, iron.
Cloves.
Corozo nuts and coconuts (for use
in making buttons, e tc.).
W alnuts and hazelnuts.
Caster oil.
Cottonseed oil.
Coconut oil.
Linseed oil.
O live oil.
Palm oil.
Turpentine.
Oil cake.
Paraffin and vaseline.
Mother-of-pearl shells.
Refined petroleum.
Pepper.
Phosphorus.
Piassaba.
A llspice.
Quicksilver.
R ice, kind not specified (i. e ., im
*127.
128.
129.
130.
131.
R ice, Japanese.
R ice, Java.
R ye.
Raisins.
Rum .
Ganzeri).
132.
133.
134.
135.
*136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
* 152.
153.
*154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
*171.
172.
173.
174.
Grass seed.
Clover seed.
Flaxseed.
Rapeseed ( Raps Undrubsaat).
Sesame seeds.
Sago and tapioca.
Saltpeter.
A nehovies.
Sardines.
Shellac, and gum lac.
Slate.
Grease.
Sulphur.
Sail twine^(canvas yarn).
Soda, calcined and crystallized.
Steel.
Stearine.
Coal.
R attan.
Sumac.
Tobacco, without designation as to
origin (im Ganzen).
Tobacco, San Domingan.
Tobacco, Cuban.
Tobacco, Porto Rican.
Tallow .
Cordage, new.
T ea.
Fish oil.
W ax.
Baleen.
Spermaceti and margarine.
W ines without designation of origin
or kind (im Ganzen).
W ines, exclusive of champagne.
Champagne.
French wines.
Portuguese wines.
Spanish wines.
Tartar.
W heat.
W oolen waste and shoddy.
Sheep's wool.
T in .
Lemons.
TESTING.
The only test of this relative so far as is shown in the Jahrbucher is
its comparison with Dr. Conrad's weighted index, with Sauerbeck's
index for England, and with that which Dr. Conrad bases on the
German imperial statistics {Ueiehsstatistik) .
Index nu m ber based on prices o f the G erm an C u stom s Union.
In addition to the three series of index numbers described in the
foregoing paragraphs Dr. Conrad regularly included in his annual
study of prices a table of actual and relative prices derived from
official quotations of the German customs union. This table first
appeared in the Jahrbucher when Dr. Conrad took up the work of
n This term is not further defined in current issues. The word “ Lichte/’ which in technical usage Is
equivalent to “ Kerzen,” appears without qualification. In the issue for 1889 it was defined as “ StearrnLichte,” i. e., composite candles; but it can. not be safely inferred that at present composite candles only are
included, since the technical term “ Lichte” covers all sorts of candles, as tallow, paraffin, wax, etc.
GERMANY— JAHRBUCHER FUR IfATIONALOKONOMIE.
239
studying German prices, as published in the fifteenth volume of the
new series, 1887 (forty-ninth volume of entire series). This index
has usually appeared with those based on Hamburg prices, but in a
few cases has been issued separately. In general it has appeared
more regularly than the Hamburg index.
SOURCE OF QUOTATION^.
In the volume of the Jahrbucher in which this table first appeared
Dr. Conrad merely states that “ Table III, which follows, presents the
movement of prices from 1871 to 1886, as they are so satisfactorily
published in the official imperial statistics, and we compare the
periods 1879-1882 and 1883-1886.”
In the next issue, however, the table is credited to the “ Monthly
Statistical Journal of the German Empire ** ( Monatshefte der Statistik
des Deutschen Reichs), and it continued to be so credited until the
publication of the table for the year 1892 (Jahrbucher, 3d series, vol.
61, entire series, 1893). From that date to the latest year for which
information is shown, 1912, it was from the “ Monthly Statistical
News of Foreign Commerce in the German Customs U nion” (Statische Nachweise uber den miswartigen Handel des deutsehen Zollgebiets.)
BASE PERIOD.
As issued for 1912 the index is computed upon two bases— 18791883 and 1879-1889. The period 1879-1882 was the base as the table
was originally printed. It was changed to 1879-1883 with the pre
sentation of the table for the year 1888 (Jahrbucher, vol. 18, whole
series vol. 52, 1889). The second index, with the base 1879-1889,
was added when the table was printed for 1890 (Jahrbucher, 3d
series, vol. 1, whole series vol. 56, 1891).
PRICES: H O W SHO W N AND COMPUTED.
On the base 1879-1883, the table as last published 78 shows relatives
by five-year periods from 1884 to 1913, inclusive, and annual rela
tives for years since 1908. B y the use of preceding issues of the
Jahrbucher an annual relative is made available from 1888 to 1912,
excepting for 1889 and 1902. For these years relatives were not
computed, although the actual prices were available.
On the base 1879-1889 the table shows relatives by five-year
periods from 1889 to 1913 and annual relatives from 1908 to date.
B y the use of preceding issues an annual relative is made available
from 1890 to date, except for 1889 and 1902, as noted above.
Actual prices are presented in the same way, and are available for
every year of the period covered.
NUMBER AND CLASS OF COMMODITIES.
The table as constructed embraces quotations on 33 articles. In
this number are included two price series each for sugar, coffee,
tobacco, and cotton yarn. However, for the years after 1909 the
quotation on herring is lacking, while from 1907 to 1912 no quotation
on copper is shown. Therefore the relative as printed is actually
based on 33 articles from 1884 to 1906, 32 articles from 1907 to 1909,
and 31 articles from 1909 to 1912.
7®Jahrbiicher fur Nationalokonomie uud Statistik, 102, Band (III. Folge, 47. Band), Heft
p. 800.
6
. Juni, 1914,
240
INDEX NUMBERS OF WHOLESALE PRICES.
The relative, however, is not computed on 31 quotations but on 103,
as for many articles several quotations are utilized in making up the
average annual price on which the relative is based. There has been
a slight variation in the number of quotations from time to time, but
no change of any significance. As first issued the relative w&s based
upon 96 quotations.
Of the 33 articles in the list some are raw materials, others are fin
ished manufactures, while a few are semimanufactured products.
DESCRIPTION AND GROUPING OF COMMODITIES.
The articles are not divided into groups. They are described as
follows in the issue of the Jahrbucher for June, 1914:79
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
W heat, 14 quotations (1892 to date).
R y e , 14 quotations (1892 to date).
Barley, 15 quotations (1892 to date).
Maize, 5 quotations (1892 to date).
Oats, 14 quotations (1892 to date).
W h eat flour, 6 quotations (1892 to date).
R ye flour, Berlin.
Rapeseed oil, Berlin.
Potato alcohol, Berlin (1892-1903), Ham burg (1904 to date).
Raw sugar, Madgeburg.
Refined sugar, Madgeburg.
Coffee, R io, good ordinary, Bremen (Sabanilla, 1896 to date).
Coffee, plantation, Ceylon, m edium , Frankfort on the M ain.
R ice, Rangoon table, Bremen.
Pepper, Bremen.
Herring, Norwegian, Hamburg (Scotch, 1904 to date)
Leaf tobacco, ordinary K en tu ck y, Bremen.
Leaf tobacco, second-grade Brazil, Bremen.
Cotton, m iddling upland, Bremen.
W ool, Berlin.
H e m p , Lubeck.
Raw silk, M ilan organzine Krefeld.
Cotton yarn, Nos. 4 0-120, Krefeld.
Cotton yarn, warp 16, M iilhausen, in Alsace.
Cotton cloth, Miilhausen, in Alsace.
Linen yarn, No. 30, flax yarn, Bielefeld.
Lead, 6 quotations.
Copper, Mansfeld, Berlin.
Zin c, 5 quotations.
T in, 3 quotations.
Pig iron, Scotch N o. 1, Ham burg (up to 1900, inclusive, Berlin).
Petroleum, Hamburg (up to 1900, Brem en), in bond.
Coal, Westphalian, Berlin.
SUBSTITUTIONS AND ADDITIONS.
Such substitutions as have been made are with respect to grade or
place of quotation and are indicated in the description above. No
additions to the list of articles have been made.
TESTING.
This relative is presented for purposes of comparison with the two
other general relatives published currently in the Jahrbucher, namely:
The weighted relative of 19 articles (based on Hamburg quotations)
and the simple relative of 157 articles (also based on Hamburg
quotations).
79
Jahrbiicher fur Nationalokonomie und Statistik,
1914, pp. 799-801.
102,
Band (III. Folge, 47. Band), Heft
6
. Juni,
GERMANY---- JAHRBUCHER FUR NATIONALOKONOM IE.
241
Dr. Conrad also reprints Sauerbeck's index for comparison with
his own, and in addition compares them with other studies which
appear from time to time.
TABLES OF RESULTS.
Tables 49, 50, and 51, compiled mainly from the issue of the
Jahrbucher for June, 1914, show the principal index numbers com•puted by Dr. Conrad:
T able 4 9 .—R ELATIV E PRICES OF 39 SELECTED ARTICLES (BASED ON HAM BURG
T R A DE STATISTICS), B Y SPECIFIED PERIODS, 1871 TO 1910, AND B Y Y E A R S, 1906 TO
1912.1
(Average prices in 1847-1870=100.)
Article.
1871-
Coffee (Brazil).!
Cocoa.................
Tea.....................
Currants...........
Raisins..............
Almonds...........
Pepper...............
Coconut oil.......
Palm oil............
Indigo...............
Mahogany.........
Cotton...............
Hemp.............. .
Rice........ ..........
Wheat...............
155.13
116. 53
86. 57
89. 50
99. 51
110.91
140. 35
80.13
100.46
111. 41
92.56
81.06
98.01
81.43
104. 38
Rye............. 106.26
Barley............... 127. 79
Oats................... 109. 97
Clover seed....... 115.02
Rape and rapeseed................
97. 88
Linseed oil.......
90. 54
Calfskins........... 103.00
Bristles............. 155. 22
W a x ..................
76. 98
Tallow...............
89.74
Fish oil.............
82.17
Lard..................
86.92
Herrings........... 121. 94
Iron, pig........... 117. 71
T in..................... 96. 32
Copper............... 88.02
Lead................... 112.19
Quicksilver___ 129.54
Coal and coke.. 109.88
Saltpeter...........
96. 71
Bar iron (Eng
lish)................ 113. 53
Cotton yarn___ 115.60
Woolen a n d
h a 1 f-woolen
yarn............... 101. 73
Linen yarn.......
80.55
18811890
18911900
19011905
19061910
1907
120.29 125. 62 75.90 83. 37 82.99 73. 75
130. 98 119.07 109. 48 111. 84 103. 48 154. 58
67.20 52.96 53. 98 58.83 60.86 59.90
80. 33 63.44 79.07 96. 39 90.88 97.40
88.10 79.43 94.19 102. 87 99. 51 118. 89
111. 22 101. 34 161. 56 122.96 89.96 125. 70
182.93 104. 65 162.61 127.35 144.41 135. 27
68.54 61.26 70.88 76.93 71.92 77.30
71.04 57. 61 62. 71 74.67 69.47 78.41
93. 36 77. 48 54.23 40.85 47.15 51. 89
83.14 68. 59 56.97 42.86 43. 53 44.29
62.53 45. 95 48.24 49.70 48.57 48.87
85. 55 80.37 90.13 91.69 83.95 81.88
68.15 59. 36 57. 78 65.39 62.24 68.46
76.26 61.73 60.73 73. 24 63.65 67. 31
82. 54 71. 34 66.46 80.10 70.84 82.60
89.95 57.76 51. 34 67.84 63.35 74.15
89.34 78.07 82.24 88.93 68. 47 101. 64
98.00 82.46 85.84 93. 58 78. 51
1908
1909
1911
81.84 84. 55 102. 28 130. 44
118.15 97.66 95.80 97.05
51.04 61.75 60.48 63. 50
95.50 94.32 104.34 109.77
102.76 89.10 108.44 119. 52
114.17 126.09 128. 26 135. 48
126.64 143. 51
115. 40 114.
71. 48 72.40 85.11 78.08
66.63 69.15 85. 31 83.16
38.24 33. 75 34. 58 43.95
44.13 37.11 43.28 57.06
50.52 45. 23 55.55 29.43
97. 46 96.73 101.17 97.15
69.84 64.47 62.62 74.60
73.97 84.11 73.97 71.
90.61 89.96 73.34 76.10
71. 36 69.90 62.62 72.69
86. 48 91.53 86. 89 88.66
94.09 90.02 108.48 114. 58
84.99 72.13 68. 52 74.68 86.16 78.00 73.82 70.56
69. 48 66.22 78.36 77.26 68.06 75.24 76.86 76.57
74.42 59. 58 70.38 87.89 94.10 82.66 86. 48 87.31
145. 76 88.75 74.42 81.07 82.12 78.52 77.91 8a 76
54.12 60.99 74.41 75.45 81.98 83.06 74.42 67.14
74.04 57. 52 66.11 74. 78 68. 82 74.76 78.47 73. 51
66.65 49.24 52.95 55.64 49.77 57. 55 51.97 54.30
78.25 62.50 75.88 87.92 76.23 83.36 84.71 98.89
109.15 103. 77 110. 82 117. 54 123. 41 126.40 111. 75 110. 35
76.57 78.20 83. 38 97.55 94.55 97. 28 97. 55 96.46
84.72 84. 43 95.06 108. 85 105. 33 111. 65 100. 27 107.18
63. 91 57. 20 67.35 68.90 70.80 81.95 65. 36 65. 77
83. 77 86.44 95.06 102. 94 112.18 116. 76 102. 59 96.23
91. 37 87.74 84. 08 92.43 94.67
83.40 87.16 102.
77.78 85.85 87.65 87.65 83.95 93. 83 92.59 83. 95
73.71 56. 30 65. 62 69.05 75.14 75.70 68.98 65.90
72.37
105.39
1910
1912
150. 41
102.30
57.56
110. 71
122. 88
138. 64
164.27
72.83
81.78
35.94
2 7.78
51.85
53.27
69.84
76.53
87.48
86.53
103.69
111. 24
76.61 73. 75 85.78
96.50 115.13 102. 08
86.77 88.76 85.20
85.73 93. 92 98.92
72.72 80.79 82.48
78.69 78.67 84.69
62.10 65.61 56.68
108. 59 89. 45 99.65
115.11 117.
176.49
104. 09 96.46 103. 81
118.15 143. 34 153.85
63.89 63.46 72.31
91.14 86.49 98.24
95.65 98.04 92.25
86.42 82.72 95.06
62.11 67. 37 74. 58
81.37 76.69 83.39 88.35 90. 74 74. 92 92. 51 90.74
113. 48 114.12 109.64 110.96 133. 41! 121.13 127. 38 132. 44
13. 31
87.32
69. 98 61.72 54.88 65.61 65. 92 68.72 63.18 62.90, 66.38 67.65 63.32
111. 49 117.35 121.11 125.60
98.17 109. 91 116. 56 117.10 116. 30 121. 75 117.
1 Jahrbucher fiir Nation alokonomie und Statistik, 102. Band (III. Folge, 47. Band), Heft 6. Juni, 1914,
p. 796.
2 These figures appear thus in the original.
83226°— 21— Bull. 284-
-16
INDEX NUMBERS OF WHOLESALE PRICES.
242
T able 5 0 .—R EL ATIV E PRICES OF CER TAIN GROUPS OF ARTICLES W E IG H T E D IN PRO
PORTION TO T H E IR CONSUMPTION (BASED ON H AM B U R G T R A D E STATISTICS), B Y
SPECIFIED PERIODS, 1871 TO 1910, AND BY Y E A R S , 1906 TO 1912.*
(Average prices in 1847-1880=100.)
Article.
Prices
18711880
(based 1881on
1890
18171867“
1 0 0 ).
18911900
19011905
19061910
1906
1907
1908
1009
1910
1911
1912
75. 47
77.31
80.41
83. 98
92.32
53.54
47.93
58. SS
62. 98
54.-85
46.26-
39.06
|
i
i ..
Coffee, Brazil...
Cocoa.................
Tea..................... 141.66 100. 46 101. 50
Pepper...............
Rice...................
66.82
79.69
72.31
87.35
i
!
II.
Cotton...............
81.84
66.28
48.70
51.07
52.68
51.78
51. 79
|
in.
Indigo.. - ........... )
Saltpeter...........
Fish oil............. 101.65
Palm oil............ J
78. 87
61.65
61.15
42.73
63. 36
42. 57
39.91
36.26i 38.09
72.97
73.26
79. 51'
94.48
89.50 102.72
88.69
91.68
96.96 107.71 117. 90
109.88
75.90
83. 74
S6.03
85.54
81.93
91. 57
90.36
81. 93
84. 34
8a
72
92.77
W heat...............
Rye.....................
Baxley............... ■112. 51
O ats...................
78.01
63.90
63.40
69. 8 8
67.06
72.52
72.60
76.02
67.14
69. 59
79.29
77. 43
6 8 .44
67.93
71.31
7a 55
76.42
70.10
68.52
70.79
76.04
79. 38
91.70
84.10
76.37
83.03
80.52
89.47
82.07
80.22
82.87
84.80
IV .
Pig iron............. 1
Tin.....................
Copper............... •111.80
Lead................... |
V.
Coal....................
V I.
Average.. 105. 54
Arith m etical
mean, com
puted from
157 Hamburg
average prices 111. 31
2 86.
51
1 Jahrbucher fur Nationalokonomie und Statistik, 102.
Band (III. Folge, 47. Band) Heft 6 . Juni, 1914,
p. 708.
2 Gomputed from prices of only 134 articles, since the separate listing of several qualities of an article
such as tea, coffee, rice, wool, oil, etc., has been discontinued.
GERMANY— JAH RBUCH ER FUR NATIONALOKONOM IE.
243
TARLESa—R ELATIV E PRICES OF CER TAIN GROUPS OF ARTICLES W E IG H T E D IN PR O
PORTION TO T H E IR CONSUMPTION (BASED ON HAM BUR G TRA DE STATISTICS), B Y
SPECIFIED PERIODS, 1871 TO 1910, AND B Y Y E A R S , 1906 TO 1912—Concluded.
(Average prices in 1871=100.)
Article.
Prices
187118S0
(based 1881on
1890
1817-1867—1 0 0 ).
18911900
19011905
19061910
1906
1907
1908
1909 : 1910
1911
1912
i.
Coffee, Brazil...
Cocoa................ )
I
Tea............. .
Pepper...............
Rice................... ]
79.43
88.23
52.62
74.09
57.16
81.'21
70.16
71. 87
74.76
82.72
85.82
79.63
56.69
59.44
61.39
59.92
60.29
62.32
55.79
68.61
73.40
64.04
78.97
61.69
61.22
40. 28
93.43
49.57
37.64
34.19
35.91
43.63
38.83
67.19
67.48
73.21
96. 76
82.49 105.18
90.83
93.90
99.30 110. 31 120. 74
70. 79
78.09
80. 23
78. 76
76.40
85.39
84. 27
76.40
78.65
75. 28
8 6 .52
76.37
62.65
62.07
6 8 .8 8
65.65
75. 48
71.56
74.94
66.18
6 8 . 60
78.16
74.86
66.19
65.56
70. 30
6 8 .1 2
75.52
69. 46
67. 85
70. 57
75.66
78.69
8-5.18
78. 47
71.07
77.30; 74.87
83.19
76. 51
74. 74
77. 21
78. 8 8 * 80.47
ii.
Cotton...............
in.
Indigo................ )
Saltpeter........... 1>.........
Fish oil.............
Palm oil............ J
TV.
Pig iron.............
Tin..................... {
Copper............... >.........
Lead................. )
V.
Coal....................
V I.
W heat...............
Rye....................
Barley............... 1>.........
Oats................... J
Average
A rithm etical
mean, com
puted from
157Hamburg
average prices
8 Computed from prices of only 134 articles, since the separate listing of several qualities of an article
Such as tea, coffee, rice, wool, oil, etc., has been discontinued.
244
T
able
INDEX NUMBERS OF WHOLESALE PRICES.
5 1 .—R EL ATIV E PRICES
OF
SELECTED
ARTICLES
1890 TO
1913 (BASED ON
CUSTOMS
[Jahrbucher fur Nationalokonomie und Statistik,
(Average prices in
Mar
ginal
num
ber.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Article.
Wheat—141 quotations.................
Rye—141 quotations......................
Barley—15 * quotations.................
Maize—5 1 quotations....................
Oats—131 quotations....................
Wheat flour— 6 1 quotations........
Rye flour, Berlin3 ..........................
Rapeseed oil, Berlin......................
Potato alcohol, Hamburg4..........
Sugar, raw, Magdeburg...............
Sugar, refined, Magdeburg..........
Coffee, Rio, 6 good ordinary, Bre
men................................................
Coffee, plantation, Ceylon, 6 me
dium, Frankfort on Main.........
Rice, Rangoon, table, Bremen. .
iPepper, Bremen.............................
Herring, Norwegian, Hamburg..
Leaf tobacco, Kentucky, ordi
nary, Bremen..............................
Leaf tobacco, Brazil, second
grade, Bremen............................
Cotton, middling upland, Bre
men................................................
Wool, Berlin...................................
Hemp. Liibeck...............................
Raw silk, Krefeld..........................
Cotton yarn, Nos.40-120, Krefeld.
Cotton yarn, warp, 16, Miilhausen, in Alsace..............................
Calico, Miilhausen, in Alsace___
Linen yarn, No. 30, flax yarn,
Bielefeld...................................
Lead— 6 1 quotations.....................
Copper, Berlin................................
Zinc—5 1 quotations.......................
Tin—3 1 quotations........................
Pig iron, Sctoch, No. 1, Ham
burg. .............................................
Petroleum, Hamburg, in bond. .
Coal, Westphalian, Berlin...........
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
102.59118.95 99.73 82.70
110.271136.65 117.35 89.88
113.03112.12 104.26 101.41
91.53 116.60 96.35! 88.40
116.57115.49 105.43 124.49
100.60,115.14 94.33! 78.98
112 95 140.41 1 2 R 5 6 185.50
1 90 IK 11ft 99 95. 631 89.51
112.79 140.46 115.36 106.74
63.99 67.51 69.01! 58.48
83.14 83.98 85.22 , 85.15
72.61 75.78 83.37 92.24 104.15 93.88
77.95 79.09 80.01 85.351 97.77 106.26
93.95 89.15 95.43 99.96 107.92 107.54
82.50 84.84 69.69 68.09; 76.98 83.24
98.84 88.61 94.01 101.85,110.00 109.27
70.54 74.62 78.98 8 8 . 2 0 ; 96.88 89.71
74.77 79.75 1 7 8 .7 8
>
78 32 7 9 . 9Q1 8 9 9 8
104.59
92.68
100.90 105.63 107.75
45. 6 6 39.85! 44.27 36.48 39.10 47.68
72.98 6 6 .49 72.69 69.74 70.36 84.84
1
I
153.66140.79 124.89 141.96 140.79139.07 124.39 96.73 70.19 62.01
I
I
1
1
120.70 120.37115.75 120.03 112. 78108.38 105.20 98.87:! 85.33 81.04
103.13 102.01 95.58, 86.51 83.56 76.85i 79.76 86.73 104.47 105.29
88.57 ! 68.61 52.92 49.07 39.90! 39.90! 40.20 54.04 73.26 71.79
84.82111.79 8 6 . 46 61.16 77.64107.30 76.28 96.85 92.94 128.70
!
1
68.78 75.69 85.62 101.31 89.93 76.98; 67.64 6 6 .34 81.16 76.43:
i
1
136.45 120.16 95.56 83.24 71.93 67.63 88.72 107.52 120.67 141.27j
1
I
99.64 76.99 107.38 76.76 63.15 63.64 ! 71.44! 65.55 54.78 63.27
94.37 89.50 81.30 76.14 70.81 73.16; 77.04 72.34 76.88 108. 79
91.67 85.61 83. 43 95.08 106.58 107.66 105.72 101.53 107.05 103.19
95.03 77.73 83.56 1 0 2 . 8 8 68.44 77.30 73.81 70.48 71.63 94.22
100.94 91.61 81.55 88.47 79.87 81.13| 90.57 80.29 76. 73 98.88
1
90.50 81.56 78.21 84.92 78.77 77.90, 81.56
110.13 86.96 89.56 106.95 92.61 95.00' 94.78
I
88.63 89.57 90.52 102.84 100.57 85.31! 84.55
100.89 95.14 81. 73 75.97 73.64 80.13 85.41
101.48 93.45 83.95 79.44 70.63 77.27, 82.78
143.22 143.32 130.50,107.19 96.05 91.04 101.25
98.40 94.76 97. 6 8 92.34 73.86 66.69; 62.66
75.20 64.80 60.33
79.56 76.52 78.#33
82.84 82.84 87.38
93.80 99.33 124.53
83.86 89.18 139.02
108.20 127.04 167.46
63.47 73.46 118.98
120.35 106.08 95.04! 96.82 98.09 94.32 94.04! 92.64
89.76; 84.86 79.02 62.93 64.12 88.92 82.00 70.32 ’ 79. '8 8 ’ 97.'95
132.60^29.40 117.34114.13 114.68 114.58 113.64 114.41 117.56 127.59
Arithmetical average........... 105. 71j 98.14 95.32 91.52 82.54 | 81.75 81.82 82.65 84.04 99.60
1
1
2
Present number has varied from time to time.
Relative not computed for this year.
i
GERMANY---- JAHRBUCHER FUR NATION ALOKONOMIE.
245
ACTUAL A V E R A G E PRICES AS SHOW N B Y OFFICIAL STATISTICS OF THE GERMAN
UNION).
3. Folge,
1,
5,
11,
17, 24, 31, 34, 44, 46, and 47.]
1879-1889=100.)
1900
1901
1902
85. 51 90.04 (2)
93.13 96.12 (2)
104. 48 106.36 (2)
89. 67 91.84
102.30 107.34 (<*l
76.57 81.32
93.33 91.16 8
111.15 104.80 (2)
96.71 80.99 (2)
41.71 35.94 (2)
78.55 85. 57 (2)
1903
1904
85.55 92.55 89.92
89.70 90.31 93.85
96.69 100.96 102.97
91.99 94.35 97.26
97. 71 97. 6 8 105.57
79.30 83.66 82.02
86.85 84.82 98.88
87.63 82. 45 85. 45
1906
1907
33.90 37. 40 40.26
78.74 58.72 63.01
1 1 0 . 62
124.08
111.52
113.08
132. 77
99.15
122. 52
133.95
55.89
31.33 31.63
54.42 56.51
70.56
(2)
51.62
6 6 .55
83. 89 76.30
94.59 96.38
104.88 102.41
137.80 125.04
75.93
93.07
99.64
97.71
78.54 58.70
1905
95.04
102.16
105.99
103.14
119.44
85.68
101.50
103.85
1908
112.99
121.16
113. 57
133.98
120.34
1 0 2 .1 0
114.90
126. 60
66.90
38.70
60.30
1909
1910
1911
124.32 1 1 1 . 6 6 108.85
113. 67 99. 61 110.41
109.11 99.63 116.25
125.37 1 1 1 . 1 2 110.82
126. 51 113. 67 137.02
111.52 100.32 98.90
107. 54 92. 80 103.04
103. 54 104.98 117.13
50.50 50.12 48.85
39.79 46. 27 47.34
61.56 69.97 68.37
1912
1913
115. 6 6 108.33
1 1 2 . 2 0 108.47
126. 23 106.13
127. 72 110.40
142. 78 120.96
100. 04 98.58
1 1 0 . 1 0 100. 77
120.94 121. 67
65.92 70.94
46.21 34.83
71.41 57.73
Mar
ginal
num
ber.
1
2
3
4
5
6
7
8
9
10
11
66.60
78.95 110.59 128.14 109.97
12
(2)
(2)
(2)
(2)
72.90
99. 46
104.49
109.75
75.81 76.85 76.34 77. 41 73. 33
94.62 . 95.98 104.96 104. 46 97.01
96. 57 89.15 76. 56 54. 20 56.83
116. 43 148.26 112.83 1 0 1 . 1 1 128.38
82.15 96.35 104. 64 99.83
97.94 110.63 129.71 114.66
64.11 76.35 89. 37 83. 29
13
14
15
16
17
69.55
63.83
62.61
99.34 104.59
(2)
112. 54 93.82 92.86 109.51 136.56 156.13 124.47 145. 62 155. 41 153.98 141.31
125. 75 100.09
(2)
84. 28 89.81 98.86 109.99 137.73 133. 58 108.65 108.49 141.08 146.98 151.24
89.89 77.19
92.10 75.84
121.33 132.05
83. 40 73. 24
166.77 89.73
(2)
(2)
(2)
(2)
(2)
108. 47 85. 61 99.15 106.08 93.57 104.51 132. 41
98.11 97.56 102. 23 114. 23 114. 78 103.71 114. 23 110. 41
118.53 111.54 113.29 122.07 129.57 124.83 126.85 135. 31
87. 62 74. 56 78. 56 8 6 . 45 112. 71 79. 48 82.69 81.22
108.39 112.37 107.34 129.77 163.10 126. 62 128.09 151.36
75. 42 65.92
101,30 81.74
(2)
(2)
78.77 98.88 82. 77 1 0 1 . 1 2 118.44 97. 76 94. 41 109. 50 108.38 102.23 107.26
103.91 107.39 113.86 118.70 132. 61 109.56 104.35 1 0 6 . 5 2 107. 40 113.91 110.87
24
25
105.92 109. 43
131.94 98.55
125.26 1 2 1 . 1 2
126.86 106.55
135.95 121.76
(2)
(2)
(2)
(2)
(2)
96.87 106.35 104.73
89. 64 90.42 105.08
102.07 99.63 118.29
129.52 138.60 148.72
129.78 129.21 145.10
147. 48 124.93 139.13 145. 33 158.40 165.16 144.91
177.16 137. 42 138.28 157. 32 194.08 212.03 207. 38
26
27
28
29
30
97.76
127.72 134.51
(2)
(2)
(2)
105.30 98.11 100.57 110.17 1 2 0 . 0 2 105. 42 103.18 104. 64 103.44 119.23 125. 60
97.89 89.58 84. 44 92. 8 8 94.33 102.64 93.21 81.20
116. 89 116. 62
1 2 2 . 8 6 1 2 2 . 8 6 129.06 125.95 131.14 132.52 128.38 127.00 127.00 111.14 132.52
31
32
33
99.21 93.13
(2)
1 0 0 .2 0
115.16
131.45
147.42
165.89
184. 41
139.57 115.92
146.12 102.51
From 1892 to 1894 No. 0 0 with sack; from 1895 to date, No. 0/1.
* Berlin, 1879-1903, Hamburg, raw, 1904 to date.
104.94 113.00
112.87 1 2 1 . 2 0
163. 20 164.99
74.68 81.52
131. 45 138.57
99.38 111.23 122.56 119.29 121.04
82.79 98. 30 98.55 159.56 166.13
93.82 94.05 96. 46 108.01 113. 56 104.94 102.13 104.17
8
117. 6 6
108. 2 2
141.93
79.34
141. 51
1 1 2 .2 0
18
19
20
21
22
23
120.31 115.83
Coffee, Sabaniila from 1896 to date.
e Java, 1879-1890.
5
246
UtfDEX NUMBERS OF WHOLESALE PRICES.
INDEX NUMBERS OF OTTO SCHMITZ.
PUBLICATION.
The data embodying the results of Mr. Schmitz's original study
of the course of wholesale prices were published in book form in Berlin
in 19G3.80 “ Those from 1903 to 1906 are extracted from manuscript
figures supplied by the author himself and in the possession of the
Royal Statistical Society. As regards later years the index numbers
have been taken from the supplement to the Zeitschrift fur Sozialwissenschaft (Monatliche Ubersichten uber die aligemeine Wirtschaftslage); only the general index numbers for all articles and for one
group (the metal group) are there given, however.” (Great Britain.
Report of an Enquiry by the Board of Trade into Working-class
Rents and Retail Prices, 1912. p. 354.)
HISTORY.
This study was undertaken in order to furnish a convenient means
of measuring the fluctuations in German prices. It was considered
highly desirable to construct an index covering a long period of years
and reliable data from a single source were not available. By using
two sources the author was able to cover the second half of the nine
teenth century and to include six price waves, whose duration he
fixes as follows:
From 1849 to 1858.
From 1879 to 1886.
From 1858 to 1870.
From 1886 to 1895.
From 1895 to 1912.
From 1870 to 1879.
SOURCE OF QUOTATIONS.
The earlier series of index numbers (1851-1885) is based on quota
tions of the bureau of trade statistics at Hamburg .81 Mr. Soetbeer,
who used figures from the same source, says in regard to the way in
which the quotations were obtained:
“ Throughout this period (1847-1885) Hamburg was an important
market for almost all raw materials. Moreover, it has been a free
port, without duties or differential taxes. Commodities imported are
declared in writing, with a statement of their weight and of their
ordinary trade designations. Their value is stated separately for
each commodity, either according to its price on 'Change that day,
or if there were no quotations, according to the probable price plus
the cost of importation. For consigned goods a careful estimate of
the prices sufficed, sometimes supplemented with a statement of their
insured value. These declarations, which were carefully supervised,
were then collected by the bureau of trade statistics, and tables were
made out of the quantity and value of goods exported and imported " 82
The later index number (1879-1902) is based on data of the Im
perial Statistical Office, which published for each of the more impor
tant commodities several series of quotations representing a number
of important markets and varieties. From these the author selected
as a basis for his index number a single series from a representative
80 Die Bewegung der Warenpreise in Deutchsland von 1851 bis 1902; nebst zwei Erganzungen: Bankdiskont, Goldproduktion und Warenpreisstand, der Weizenpreis, von 400 v. Chr. bis 1900. Von Otto
Schmitz. Berlin, 1903. 443 pp.
81 In regard to Hamburg Bureau of Trade Statistics see also pp. 229 and 230.
83 Bimetalism in Europe, by Edward Atkinson, p. 226.
GERMANY— -OTTO SC H M ITZ.
market, and in connection with the detailed description of the article
stated the manner in which prices were obtained for the Imperial
Statistical Office at that market. These statements are retained with
the descriptions as given herewith. Eleven markets are represented:
Berlin and Bremen by 8 commodities each; Breslau and Hamburg
by 3 each; and Danzig., Magdeburg, Krefeld, Bielefeld, Dortmund,
Cologne, and Essen by 1 each.
In the few cases in which a hiatus occurs in the official series the
source of the substituted data is stated in connection with the descrip
tion of the commodity.
BASE PERIOD.
The average of the 10 annual prices as given by the Imperial Sta
tistical Office for the first 10 years covered by its reports (1879-1888)
was taken as the base for both series of index numbers. On the ground
that the period was a time of quiet and normal business development
and contained a period of falling prices, beginning in 1880, and a pe
riod o f rising prices, beginning in 1885, the average price was consid
ered comparatively normal and reasonably satisfactory as a basis.
The base period for lard is evidently the 9-year period, 1880 to 1888.
The Imperial Statistical Office did not publish quotations on lard for
1879 and no other quotations were substituted.
PRICES: H O W SHOW N AND COMPUTED.
The various methods of determining monthly averages are given
in the section on description and grouping of commodities. The
simple average method is used in computing the various averages for
the tables. All index numbers are printed with two decimals.
Data of the Imperial Statistical Office are in almost all cases
shown separately from Hamburg data and in much greater detail.
The principal table of the study shows for each article and group
of articles the index numbers for each month, quarter, half year,
and year from 1879 to 1900, while the corresponding table for Ham
burg gives the index number for articles and groups by years only.
Actual prices of the later series are shown by months and of the
earlier by years.
NUMBER AND CLASS OF COMMODITIES.
The index number based on the Imperial Statistical Office prices
(1879-1902) represents 29 commodities, while the earlier number
based on Hamburg prices (1851-1885) includes only 24 commodi
ties— corn, linen yarn, petroleum, and one quotation each for iron
and coal being lacking. The list of 29 commodities is the original
list of the Imperial Statistical Office,, with the following modifica
tions : ( 1 ) Wheat flour, rye flour, refined sugar, cotton yarn, and
cotton cloth are omitted as not being strictly raw materials. Linen
yam, however, is not omitted because the list does not include flax.
(2 ) Two kinds of iron and two of coal are carried as separate com
modities instead of one description for each group. (3) Lard and
butcher’s meat are added as being important commodities introduced
early in the period covered by the figures of the statistical office.
For the years 1909 to 1912, inclusive, however, copper and one
description of pig iron are excluded.
248
INDEX NUMBERS OF WHOLESALE PRICES.
DESCRIPTION AND GROUPING OF COMMODITIES.
The descriptions of commodities are based on the descriptions of
the selected varieties as given by the Imperial Statistical Office and
cover the years 1879-1902.83 Concerning the descriptions for the
period 1851-1885 the general statement is made that Hamburg quota
tions are for standards of quality similar to those of the Imperial
Statistical Office.
The grouping of commodities appears in the list which follows.
The description of the first commodity is given in full, but the descrip
tions of the others are somewhat abbreviated.
Group I.— Grains.
1. JVheat.
Official Berlin quotations. The average m onthly price is ascertained b y the com
mittee of senior merchants of Berlin on the basis of daily quotations based on hear
ings of the brokers. T h e early quotations are for 1,000 kilograms good, sound, yellow
wheat of all origins, 71.3 kilograms per hectoliter. On October 1, 1887, the standard
was fixed at a m inim um of 71.5 kilograms per hectoliter, and since January 1 ,1 8 9 3 , the
quotation is for good, sound, dry wheat, free from m usty ordor (hard wheat excluded),
of #11 origins, m inim um 72.5 kilograms per .hectoliter. In the author’s judgment
this repeated raising of the standard has exercised no demonstrable influence on
prices
A further rise in the standard grading weight to 75.5 kilograms per hectoliter on
January 1, 1893, was caused b y the introduction of the new grain tester and was of
merely formal nature.
On account of the closing of the Berlin grain exchange there are no official quota
tions for wheat, rye, and oats from January 1, 1897, to A pril 1, 1899. Therefore the
official Breslau quotations, suitably adjusted (unter entsprechender Anpassung), are
used. For January, February, and March, 1899, information founded upon the daily
publications of the central quotation office of the Prussian agricultural chambers
was used since the standard of quality was the same. From A pril, 1899, the official
Berlin quotations are again given. The in d ex numbers calculated on the Breslau
figures are printed in italics.
1879, d uty free.
1880-1885, duty per 1,000 kilograms gross, 10 marks.
1885-1887, duty per 1,000 kilograms gross, 30 marks.
1887-1891, duty per 1,000 kilograms gross, 50 marks.
1891 and following, duty per 1,000 kilograms gross, 35 marks.
2. Rye.
Official Berlin quotations. The m onthly average price is ascertained as in the
case of wheat up to October, 1887. The quotations are for 1,000 kilograms good,
sound rye of all origins. T he standard grading weight rose during the period from
65.9 to 67.8 kilograms per hectoliter without demonstrable influence on the price.
In the absence of Berlin quotations for the period January, 1897, to A pril, 1899, the
procedure was the same as in the case of wheat.
1879, duty free.
1880-1885, duty per 1,000 kilograms gross, 10 marks.
1885-1887, duty per 1,000 kilograms gross, 30 marks.
1887-1891, duty per 1,000 kilograms gross, 50 marks.
1891 and following, d u ty per 1,000 kilograms gross, 35 marks.
3. Barley.
Breslau quotations. The price is ascertained every week day b y the m unicipal
market commissioner. The m onthly average price is derived from the Breslau
chamber of commerce. The quotation is for 1,000 kilograms m edium heavy barley.
1879, d u ty free.
1880-1885, d u ty per 1,000 kilograms gross, 15 marks.
1885-4887, d u ty per 1,000 kilograms gross, 23 marks.
1887-1891, d uty per 1,000 kilograms gross, 23 marks.
1891 and following, d u ty per 1,000 kilograms gross, 22.50 marks.
_____________
83 N o detailed description is available after 1902.
•
GERMANY---- OTTO SCH M ITZ.
249
4. Oats.
Official Berlin quotations. The m onthly average price is ascertained as in the case
of wheat. The quotation is for 1,000 kilograms good, sound oats of all origins. The
standard grading weight has increased from 38.6 to 41.5 kilograms per hectoliter
without demonstrable influence on the price. In the absence o£ Berlin quotations
from January, 1897, to A p ril, 1899, the procedure was the same as in the case of wheat.
1879, d u ty free.
1880-1885, duty per 1,000 kilograms gross, 10 mayks.
1885-1887, d uty per 1,000 kilograms gross, 15 marks.
1887-1891, duty per 1,000 kilograms gross, 40 marks.
1891 an.d following, d u ty per 1,000 kilograms gross, 28 marks.
5. Corn (maize).
Bremen quotations. T he price is determined every Saturday b y the chamber of
commerce through licensed -brokers, and the m onthly average price is derived from
the m edium prices actually paid. The quotations are for 1,000 kilograms m ix ed
American corn of prime quality, in bond. On account of a shortage in corn, occa
sioned b y poor crops in America, there are no quotations from A p ril, 1882, to March,
1883. There are likewise no quotations from September to Decem ber, 1884, because
only L a Plata and Danube varieties were in the market. Although corn is quoted in
bond, duty rates are given for the sake of completeness.
1879, d u ty free.
1880-1885, duty per 1,000 kilograms gross, 5 marks.
1885-1887, duty per 1,000 kilograms gross, 10 marks.
1887-1891, duty per 1,000 kilograms gross, 20 marks.
1891 and following, duty per 1,000 kilograms gross, 16 marks.
Group II.— Other products of agriculture and products of fishing.
0. Herrings.
Danzig quotations. The m onthly average price is ascertained on the basis of the
prices actually paid -every Saturday or on the last exchange day of every week as
determined b y the bourse commission. The quotations are for 1 cask of 150 kilo
grams gross, “ Crown and f u ll ,” in bond. The custom rate, which is given although
the article is quoted in bond, is 3 marks per cask.
7. Rapeseed oil.
Berlin quotations. The m onthly average price is ascertained as in the case of wheat.
T h e quotations are for 100 kilograms good, raw rapeseed oil. Owing to the closing
of the Berlin produce exchange, January, 1897, to Decem ber, 1898, Konigsberg quo
tations were used for these two years and continued through 1899 and 1900. Since
1901 Berlin quotations have been used again. The relatives for 1897 and 1898 are
printed in italics.
1879, d u ty per 100 kilograms gross, including container, 3 marks.
1880-1885, duty per 100 kilograms gross, including container, 4 marks.
1885 and following, d u ty per 100 kilograms gross, including container, 9 marks.
8. Alcohol.
Hamburg quotations. The average price is ascertained through the brokers b y the
chamber of commerce on the basis of prices actually p aid and the average of the
m edium prices of every m onth. T he quotation is for raw potato alcohol (per 100 liters
pure alcohol) in bond. For the sake of completeness the customs rates so far as could
be ascertained are appended.
January, 1879, to July, 1879, d u ty per 100 kilograms net, 36 marks.
July 5, 1879 to 1885, d u ty per 100 kilograms net, 48 marks.
1885-1891, d uty per 100 kilograms net, 80 marks.
July 1, 1891, to July 14, 1900, d u ty per 100 kilograms net, 125 marks.
July 1 4 ,1 9 0 0 , and following, d uty per 100 kilograms net, 160 marks.
9. Raiv sugar.
Magdeburg quotations. The average price is determined on the basis of the highest
and lowest prices fixed on Friday of each week b y commissions and commissioners
of senior merchants. The quotations are for 100 kilograms first quality. The descrip
2S0
INDEX NUM BERS OF WHOLESALE PRICES.
tion was 96 per cent polarization until October, 1887, 92 per cent yield until F eb
ruary, 1897, and 88 per cent yield since that date. T he first two descriptions are
considered as corresponding fairly w ell. T h e difference in v a lu e between 92 per
cent yield an d 88 per cent y ield is estim ated at \ mark, bu t this difference is not taken
into account in calculating the in d ex num ber. From Septem ber, 1888, the quota
tion includes customs d u ty and excludes excise ta x , and the in d ex num ber has been,
revised accordingly.
10. Butcher's meat.
Berlin quotations. Prices have been quoted regularly b y the m onth since M ay,
1887. For the years 1881 to 1886 the yearly average prices 'were subsequently ascer
tained and communicated b y the Imperial Statistical Office in Decem ber, 1889.
The price for 1881 is the average price for the months March to December. For 1879
and 1880 Ham burg prices form the basis of the in dex numbers.
T he in d ex number is basfed on the average of the quotations for beef, veal, pork,
and m utton. The earlier quotations Were for 100 kilograms.
Beef, dressed weight, average of the priees for second grade.
Pork, liv e w eight, w ith 20 per cent tare, highest quotation for second grade.
Y e a l, dressed weight, lowest quotation for first grade.
M utton, dressed weight, m e d iu m of the prices for first grade.
Since July, 1897, the prices for beef, veal, and mutton are for 100 kilograms,
daughter weight. Both dressed weight and slaughter w eight are estimated on the
presum ptive weight of the four quarters on w hich the price of the anim al has been
apportioned, dressed weight w ith and slaughter w eight w ithout the deduction of
the value of the hide, head, feet, entrails, etc. T h e price based on slaughter
Weight, according to information obtained b y the Im perial Statistical Office from
authoritative sources, is about 8J per cent higher than the price based on dressed
weight. In com puting the in d e x number, allowance has been made for this change
in the manner of quoting.
11. Lard.
Brem en quotations. Tho m on th ly average price is ascertained as in the case of
co m . T he quotations are for 100 kilograms refined A m erican lard, W ilc o x brand, in
bond. T h e Im perial Statistical Office did not publish lard quotations until 1880.
The prices for September, October, and Novem ber, 1888, and for January, 1892, were
m erely nom inal, owing to lack of the com m odity, b u t were included in the yearly
average. A lthough lard is quoted in bond, th e rate of d u ty , 10 marks throughout the
period, is given.
Group III.— Colonial goods, etc.
12. Leaf tobacco.
Brem en quotations. T h e m o n th ly average price is ascertained as in the case of
c o m . T he quotations are for 100 kilograms K en tu ck y ordinary, container included,
in bond. A lthough th e article is quoted in bond, the rates of d u ty are given.
In 1879 the d u ty on 100 kilograms net was 24 marks.
A t present the d u ty on tobacco leaves, stems, and ribs, as w ell as tobacco sauce, is
85 marks.
A t present the d u ty on stripped Leaves an d smoking tobacco is 180 marks.
IS. Coffee.
Brem en quotations. T he m onthly average price is ascertained as in the case of
co m . T he quotations are for 100 kilograms net, including sack, in bond. Prior to
1896 th e quotations were for “ B io , good ord in ary,” and since January, 1896, for
“ Sabanilla, fair ordinary. ” T he change was made because R io, good ordinary, had
become relatively of sm all importance in the Bremen market. Although quoted in
bond, the d u ty rates are given.
Prior to July, 1879, d u ty per 100 kilograms net, 35 marks.
Since July, 1879, d u ty per 100 kilograms net, 40 marks.
l i . Rice.
Bremen quotations. The m onthly average price is ascertained as in the case of
co m . The quotations are for 100 kilograms Rangoon table rice,, shelled, in bond.
Although quoted in bond, d u ty rates are given .
Prior t o Ju ly, 1879, th e d u ty per 100 kilograms gross, 3 marks.
Since July, 1879, the d u ty per 100 kilograms gross, 4 marks.
G E S M A N Y — OTTO S C H M IT Z.
251
15. Pepper.
Bremen quotations. TJie m onthly average price is ascertained as in the ease of
corn. The quotations are for 100 kilograms, black Singapore pepper, in bond,
Although quoted in bond, d u ty rates are given.
Prior to July, 1879, duty per 100 kilograms net, 39 marks.
Since July, 1879, duty per 100 kilograms net, 50 marks.
Group IV.
Raw materials of the textile industry.
16. Cotton.
Bremen quotations. T h e m onthly average price is ascertained as in the case of
com . T h e quotations are for 100 kilograms m iddling upland, in bond, duty free.
17. Wool.
Berlin quotations. The m onthly average price is ascertained on the basis of weekly
quotations b y the committee of senior merchants of Berlin after having heard the
brokers. T h e quotations are for 100 kilograms North German sheep wool, medium
grade, duty free.
18. Hemp.
Hamburg quotations. Quotations every Friday. T h e m onthly average price is
ascertained as in the case of alcohol. The quotations are for 100 kilograms M exican,
net, in bales of about 350 pounds, 7 pounds tare, in bond, duty free.
19. Raw sttk.
Krefeld quotations. Quotations are ascertained on the second M onday of every
subsequent month b y a committee of merchants and manufacturers on the basis of
local transactions and of the m edium price. T h e quotations are for 1 kilogram M ilan
ese organzine, classique, 18-20, duty free.
20. Linen yarn.
Bielefeld quotations. Quotations represent prices determined on the first and
m iddle of every month b y the secretary of the chamber of commerce on the basis of
the prices obtained at the sales occurring at the spinning mills. T h e quotations are
for 1 kilogram linen yarn, No. 30 (English number), m edium price, between grades
la and I la , but for the sake of accuracy the author lias given the yearly average price
per 100 kilograms.
1879, duty per 100 kilograms gross, 3 markk
G r o u p V .-— M e t a ls .
21. Foundry pig iron.
Breslau quotations. Prices are obtained through private persons at the end of the
month from the average of all the quotations for delivery on the last day of the month.
The quotations are for 1,000 kilograms Silesian foundry pig iron at the foundry. The
grade of Silesian foundry pig iron quoted at Breslau is lower than that quoted at Diieseld ori Since 1897 other lower prices have been quoted for pig iron sold to points in
Lower Silesia, but these have not been included in the ind ex number,
U n til June, 1879, duty free.
Since June, 1879, duty per 1,000 kilograms gross, 10 marks.
22. Bessemer pig iron (from the Ruhr districts, Rhenish Westphalia).
Dortmund quotations. Prices are determined at the end of every month b y a com
mittee of the chamber of commerce or the secretary of the chamber of commerce on
the basis of schedules filled out b y producers, consumers, and dealers and on the basis
of the lowest and highest prices. T he quotations are for 1,000 kilograms Bessemer
pig iron from the districts of the Ruhr at the foundry.
U ntil June, 1879, duty free.
Since June, 1879, duty per 1,000 kilograms gross, 10 marks.
252
INDEX NUM BEKS OF WHOLESALE PRICES.
23. Lead.
Berlin quotations. M onthly average price is ascertained as in the case of wool.
T he quotations until M ay, 1899, are for 100 kilograms Tarnowitz lead, Saxonia brand,
and since M ay, 1899, are for 100 kilograms Tarnowitz and Harz lead. For Novem ber
and December, 1887, on account of the great fluctuation in prices, averages could not
be given. D u ty free.
24. Copper.
Berlin quotations. The m onthly average price is ascertained as in the case of wooL
T he quotations are for 100 kilograms Mansfeld copper. For the year 1887 trie Imperial
Statistical Office, owing to the great fluctuation in prices, gives a nominal average
covering only the months January to M ay. The author, convinced that this nominal
figure was too low, computed an average for the greater part of the second half of the
year on the basis of information which he himself obtained. Taking the average of
the Imperial Statistical Office for the first five months and his own average as repre
senting the last seven months, he obtained the average which he has used for the year.
H e states that this is the only case in which he has deemed it expedient to depart from
the official figures, and in this case the low nominal figure of the statistical office would
have affected the average for the base period sufficiently to have raised the total index
number for the period 1889-1900 about 10 points (i. e., one-tenth of a unit. Index
numbers are printed with two decimals). In March, 1899, when prices fluctuated
greatly the m onthly average price given in the tables is merely approximately correct.
D u ty free.
25. Zinc.
Cologne quotations. Quotations are obtained every W ednesday b y Cologne whole
sale establishments or the chamber of commerce. The information is based on the
booking of sales and represents the prices paid b y buyers— for the indicated grade of
the com m odity— settling their accounts regularly. The quotations are for 100 kilo
grams Rhenish crude zinc, brand “ W II und S S .” T he price for March, 1889, is
nominal, since there were no sales. D u ty free.
26. Tin.
Hamburg quotations. Quotations are obtained every Friday. T he average price
is ascertained as in the case of alcohol. The quotations are for 100 kilograms Banca
tin in blocks. D u ty free.
Group VI.— Coal and petroleum.
21. Ruhr anthracite coal.
Essen quotations. Quotations are obtained once a month at the industrial exchange
in Essen b y an exchange committee of the chamber of commerce on the basis of tne
sales at all the mines in the district and on the basis of the highest and lowest prices.
T he quotations are for 1,000 kilograms anthracite coal at the mine. From M ay to
August, 1889, there are no quotations on account of the strike. D u ty free.
28. Upper Silesia gas lump coal (Gas-Stiickkohle).
Breslau quotations. The price is obtained through private persons at the close of
the month from the average of all quotations for delivery on the last day of the month.
The quotations are for 1,000 kilograms. U pper Silesian lum p coal for gas, f . o. b. m ine.
D u ty free.
29. Petroleum.
Bremen quotations. T he m onthly average price is ascertained as in the case of corn.
T he quotations are for 100 kilograms Am erican white refined, including container, in
bond. From January 1 ,1 8 9 2 , to July 1 ,1 8 9 3 , hogshead d u ty am ounting to about 0.95
marks per 100 kilograms net was charged. A fter the lifting of the hogshead d u ty, July
1, 1893, the quotation was again for the com m odity in bond. A lthough quoted in
bond, the duty rates are given.
From January, 1879, d uty free.
From 1891, d uty, 6 marks.
GERMANY— OTTO SC H M ITZ.
253
SUBSTITUTIONS AND ADDITIONS.
The report includes parallel data from Hamburg and from the
Imperial Statistical Office for the 7-year period 1879 to 1885. Ham
burg index numbers were obtained by finding the average actual
Hamburg price for the 7-year period and the average relative Imperial
Statistical Office price for the same period. It was then calculated
what Hamburg price corresponded to the index number 100 , and on
the basis of the resulting figure index numbers for the Hamburg
quotations from 1851 to 1885 were computed.
In the case of a few commodities, breaks occur in the series of
quotations. In the absence of Berlin prices Breslau prices have been
substituted from January, 1897, to April, 1899, in the series for wheat,
rye, and oats. These are said to have been “ suitably adjusted”
(unter entsprecJiender Anpassung) but the process is not described.
Neither is the method oi changing from Hamburg to statistical office
prices of butcher’s meat at the end of 1880 described. No actual
prices are given for lard for the year 1879, and the index numbers
for all months of that year are given as 100 . This procedure is
equivalent to the substitution of the average of the actual prices
from 1880 to 1888 as the actual price for 1879. The base period for
this commodity is therefore the 9-year period 1880 to 1888.
WEIGHTING.
The total index is the simple arithmetic mean of the index num
bers of the 29 articles. No system of weighting is used. The author
holds that the simple average of wholesale prices of important raw
materials roughly indicates the course of prices and that this is its
only purpose. He does not think that manipulation on the basis
of estimated consumption makes it a satisfactory index of the stand
ard of life or increases its value as an index of price movements.
He may be considered, however, to have weighted his own index, in
a loose sense, for two of the 29 commodities are different varieties of
coal and two others are separate varieties of iron.
TESTING.
No test of the index number is made. A comparison is shown,
however, with an index number derived from Soetbeer’s figures for
the totai of the 24 articles in question, as follows:
Period.
1851-1855...................... '
1856-1860......................
1861-1865......................
1866-1870......................
1871-1875......................
Soetbeer.
Schmitz.
117.18
123.35
121.46
120.98
135.56
117.32
122.06
120.58
119.62
130. 75
TABLES OF RESULTS.
Table 52 shows the index numbers for the total of all articles.
The numbers from 1851 to 1878 represent Hamburg quotations for
24 commodities, while the numbers from 1879 to 1902 represent
Imperial Statistical Office quotations for 5 additional commodities,
or 29 in all. The statement is made that the addition of the 5
articles affects the index number only slightly.
254
T able
INDEX NUM BERS OF WHOLESALE PRICES.
5 3 . — IN D E X
Total
index
num
Year. ber:
18791888
equals
100.
1851
1852
1853
1854
1855
1856
1857
1858
1859
100.60
103.33
118.91
131.79
131. 96
132. 59
132. 76
112.08
114.16
N UM BERS OF W H O L E SA L E PRICES IN
ACCORDING TO OTTO SCHMITZ.
Total
index
num
Year. ber:
18791888
equals
100.
1860
1861
1862
1863
1864
1865
1866
1867
1868
118.73
117.24
120.31
122. 68
125. 28
117. 37
119. 88
120.69
120.09
Total
index
num
ber:
Year. 18791888
equals
100.
1869
1870
1871
1872
1873
1874
1875
1876
1877
120.14
117. 32
123. 02
136.12
141.56
130.60
122. 41
119. 52
119.86
Total
index
num
ber:
Year.
18791888
equals
100.
1878
1879
1880
1881
1882
1883
1884
1885
1886
110.62
100.87
111. 71
109.26
106. 52
104. 06
99.62
92. 88
88.00
G ER M AN Y, 1851 TO
Total
index
num
Year. ber:
18791888
equals
100.
Total
index
num
ber:
Year.
18791888
equals
100.
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
90.98
96. 07
100. 87
107. 54
104. 75
95.46
92.21
83. 79
83. 55
83.91
85. 79
90. 65
98. 43
106. 49
100. 11
99. 19
100. 64.
100.22
1913,
Total
index
num
Year. ber:
18791888
equals
100.
1905
1906
1907
1908
1909
1910
1911
1912
1913
103. 50
112.08
119.43
112. 87
111.65
113.65
118.95
130.41
127.07
N otes .—1. Index from 1851 to 1878, inclusive, based on actual wholesale prices of the Hamburg Bureau
of Trade Statistics; from 1879 to 1912 on actual wholesale prices of the Imperial Statistical Office of Germany.
2. The index numbers for the years 1909, 1910, 1911, and 1912 do not include two articles, copper and pig
iron (one kind).
3. Index for 1851 to 1902 inclusive from “ Die Bewegung der Warenpreise in Deutschland von 1851 bis
1902” ; index for 1903 to 1912, inclusive, from the British “ Report of an Enquiry by the Board of Trade
intG Working-class Rents and Retail Prices, 1912."
The two series of index numbers for the 7-year period for which
parallel data were available are:
Hamburg.
Year.
1879................................
1880................................
1881................................
1882................................
1883................................
1884................................
1885................................
104.47
109. 94
110.19
106. 21
104.52
98.31
90.69
Imperial
Statistical
Officc.
100.87
111.71
109.26
106.52
104.06
99.62
92.88
A table of index numbers for each of the 29 articles, and a total
index: number, has been prepared for a series of 5, 10, and 25 year
periods, as well as for the 22 -year period, 1879 to 1900. These are
calculated on the basis of 1879-1888 equals 100 . It is obvious that
any one of these period indexes could be made the basis ( 100 ) of a
new* series of calculations for the different commodities if so desired.
The index numbers are as follows:
1851-1855
1856-1860
1861-1865
1866-1870
1871-1875
1851-1875
.
.
117.32
122. 06
120.58
119. 62
130. 75
1876-1880, .
1881-1885 .
1886-1890 .
1891-1895
1896-1900
122.07
1876-1900
112. 52
102. 47
96.69
91. 95
93.05
99.34
119.69
1851-1860
120.10
1861-1870
1871-1880
121.63
1881-1890
99. 58
1891-1900........ 92. 50
1851-1900
110.70
1879-1900= 96. 97
In summary the author shows the price level of the latest decade
included in the study, by means of cumulative figures, as follows:
The index number for the period 1851 to 1900 is 110.70; for 1879 to
GERMANY-- ADOLF SOETBEER.
*255
INDEX NUMBERS OP ADOLF SOETBEER.
P U B LICA TIO N A N D H IS TO R Y.
Dr. Adolf Soetbeer, a German economist, published the results
of a study of wholesale prices in Hamburg as early as 1858,84 but his
main contribution to the study of prices and the one that contains the
index: number which he continued until his death, in 1892, appeared
in 1885, under the title Materials toward the Elucidation of the
Economic Conditions Affecting the Precious Metals and the Question
of Monetary Standards” ( Materialien zur Erlduterung und Beurteilung
der wirtschaftlichen EdelmetaUverhdltnisse und der Wahrungsfrage.
Berlin, 1885; 2. Aufgahe, Berlin, 1886). This publication has been
translated into English; in full, by Prof. F. W. Taussig and the trans
lation is included in the UniteH States Senate Document No. 34
(pp. 57-258), 1 st session, 50th Congress, 1887. Dr. Soetbeer added
the index: numbers for 1886 in the Hamburg Borsen-Halle, Nos. 181
and 182 (a translation of which also appears in Senate Document No.
34, pp. 271-276), and published the indexes for succeeding years up to
1890, inclusive, in his article on “ The course of prices from 1886 to
1890” (Das Niveau der Warenpreise in den Jahren 1886-1890), which
appeared in the Jahrbucher fur Nationalokonomie und Statistik,
1892, 3, Folge, 3, pp. 588-596.
The original tables compiled by
Soetbeer show average prices and indexes for each individual article,
for groups, and for five-year periods from 1851 to 1885.
The table
for 1886 omits some articles of minor importance but continues the
data for the rest and for the groups—with the exception of the group
of British exports. The article in the Jahrbucher continued the data
to include 1890, by groups only, but stated that the figures for the
individual articles were in the writer’s hands in manuscript but
would have to be reserved for future publication on account of lack
of room in the Jahrbucher. Soetbeer’s death in 1892 prevented the
realization of his plan. Although his index numbers were not con
tinued beyond 1891, two other important indexes have been based
on Hamburg prices, namely, those of Dr. Heinz, published in Ham
burgs Handel und Schiffahrt, by the bureau of trade statistics of
Hamburg and those of Prof. Conrad, of Halle, published in the Jahr
bucher fur Nationalokonomie und Statistik.85
S O U R C E O F Q U O T A T IO N S — B A S E P E R IO D .
Soetbeer used the average wholesale prices published by the
bureau of trade statistics of Hamburg.
These prices do not go back
farther than 1847, since prior to that time no such statistical bureau
existed. For that reason Dr. Soetbeer found it necessary to select
1847-1850 for the base period, although he himself states that he
would have much preferred to use 1841-1850.
si Beitrage zur Statistik der Preise: I. Uebersieht der Durchschnitts-Preise verschiedener Handelsartikei
nach den Angaben im Hamburger Borsen-Preiaeourante in den Jahren 1851-1857 unter Vergleichung rait
den Durchschnittspreisen der Jahrzehnte 1831-1840 und 1841-1850; II. Zusammenstellung der jahriichen
Durchschnitts-Preisefiir Weizen in Hamburg, Hanover, Braunschweig, Berlin, Frankreich und England
wahrend der Jahre 1851-1857. Hamburg, 1858.
85 For a description of these index numbers see pp. 228-238 of the present bulletin.
256
INDEX NUMBERS OF WHOLESALE PRICES.
PR IC E S: H O W S H O W N A N D C O M P U T E D .
TI19 annual average prices were computed by a simple arithmetical
process from the total quantity and total value of each article im
ported as recorded by the Hamburg bureau.
Since the bureau of
trade statistics entered the weight and kind o f each article imported
into Hamburg, and the price of each on the Hamburg exchange
on the day of importation, this was most easy.
When the price of
an article was not quoted on the exchange, then the invoice value,
plus freight, insurance, and other charges, was entered.
As long
as Hamburg was a free harbor and all goods entering the city in any
manner whatever were recorded, these average prices were of the
utmost importance. But when, in 1888, Hamburg joined the German
customs union, and only articles entering by sea were required to be
recorded, the figures for land importations could no longer be ob
tained with any accuracy, nor could those of articles entering by
both land and sea.
In consequence, the quotations of the bureau
of trade statistics, which up to that time had numbered over 300,
now dropped to 163. Dr. Heinz, director of the Hamburg Bureau
of Trade Statistics, undertook the task of going back over the records
of the bureau to separate sea importations from land importations
for Soetbeer’s list o f articles, so as to furnish reliable average prices
on them for years up to 1891, inclusive.
But after the death of
Soetbeer, in 1892, Dr. Heinz carried on this investigation for a
different list of articles, selecting only such as presented data pre
vious to 1888 that could also be quoted subsequent to 1888. This
list contains only 70. of Soetbeer’s 114 articles, but adds 110 new quo
tations.
The average prices of these 180 commodities were carried
back to 1850.
D E SC R IP TIO N A N D G R O U P IN G O F C O M M O D IT IE S .
The 114 articles on which Soetbeer computed his index numbers
were grouped as follows:
I. Products of agriculture, etc. {20 articles).
W heat.
W h eat flour.
R ye.
R ye flour.
Oats.
Barley.
Malt.
Buckw heat.
Peas.
W hite beans.
Potatoes.*
Hops.
Clover seed.
Rapeseed.
Rapeseed oil.
Linseed oil.
O il cake.
R aw sugar.
Refined sugar.
Spirits from corn or potatoes.
II. Animal and fish products {22 articles).
B e e f*
V e a l*
M utton.*
P ork*
M i lk *
B u tter.*
Cheese.
Tallow.
Lard.
H ides.
Calfskins.
Leather.
Horsehair.
Bristles.
Bed feathers.
Bones.
Buffalo horns.
Glue.
E ggs*
Herrings.
Dried fish.
Fish oil.
GERMANY---- ADOLF SOETBEER.
257
I II. Southern 'products (7 articles).
O live oil.
W ine in casks.
Champagne.
Raisins.
Currants.
A lm onds.
Dried prunes.
IV . Colonial products (19 articles).
Coffee.
Cocoa.
Tea.
Pepper.
Allspice.
Cassia bark.
Rice.
Tobacco.
Indigo.
Cochineal.
Logwood.
Redwood.
Mahogany.
Cane.
Palm oil.
Ivory.
Arrack.
Rum .
V. Minerals and metals (14 articles).
Copper.
Quicksilver.
Sulphur, raw.
Saltpeter, raw, Chile.
Salt.
Lim e.
Cement.
Coal.
Pig iron.
Bar iron.
Steel.
Lead.
Zinc.
T in.
VI. Textile matei'ials (7 articles).
Cotton.
W ool.
F lax.
H em p .
Silk.
Cordage.
Rags.
V II . Miscellaneous (11 articles).
Potash.
Soda.
Tallow candles.
Tar.
W ax.
Guano.
India rubber.
Gutta-percha.
Rosin.
Pearl ash.
P itch.
V III. British articles of export (14 articles).
Cotton y a m .
Piece goods, plain.
Cotton piece goods, printed.
Cotton stockings and socks.
Thread for sewing.
Common glass bottles.
Linen yarn.
Linen, plain.
Linen sail cloth and sails.
W oolen and worsted yarn.
Woolen cloths, etc.
Flannels, etc.
Worsteds.
Carpets, etc.
N o t e . —The prices of articles marked with an asterisk are the average of the prices paid by Hamburg
institutions (hospitals, etc.) for large purchases.
SUBSTITUTIONS AND ADDITIONS.
Although no substitutions or additions are mentioned specifically
as having been made, the procedure of the Hamburg Bureau of Trade
Statistics in securing its average annual wholesale prices, as described
by Dr. Soetbeer, would readily admit of such being done. Dr.
3 3 2 2 6 °— 21— B u ll. 284--------17
258
INDEX NUM BERS OF WHOLESALE PRICES.
Soetbeer states 86 that since the quantities and kinds of many impor
tant articles undergo changes in the course of decades, it had seemed
proper to the bureau to take no account of the different kinds of each
article but to treat all kinds as one in order to get a general indication
of the changes in prices. He also adds that to meet objections to
his selection of articles for his compilation, he has carefully revised
the list, so as to exclude those for which the Hamburg wholesale
prices are not to be considered fairly indicative of prices in the
general trade; that a number of very important articles were included
for which the official statements give no figures and for which whole
sale prices have been ascertained from the yearly accounts of large
institutions at Hamburg, as in tihe case of meat, butter, milk, and
eggs; that the prices of yams and cloths which appeared in the earlier
compilation were later excluded, since the indirect influence of the
German import duties on the importation of cheaper grades prevents
the prices from indicating the general range of prices of such articles,
and that in their place have been given corresponding average prices
of yarns and cloths exported from England, as well as the prices of
some other articles of manufacture, all derived from the British trade
statistics. Likewise, to incorporate the results of a thorough and
detailed revision made by the Hamburg Bureau of Trade Statistics
of its earlier price tables, it was necessary to make some slight changes
in the figures for some articles in the second edition of the Materialien
as compared with the figures for those same articles in the first edition.
Since Soetbeer’s tables were discontinued only six years after their
first publication there was no occasion to resort to interpolation, and
no trace of any is evident.
WEIGHTING.
Soetbeer’s index numbers were not weighted, although the problem
was recognized and met half way by his discriminating selection of
articles and by his including more than one variety of a commodity
in the list, as in the case of wheat, rye, sugar, etc.
TESTING.
To test his index Dr. Soetbeer constructed a comparative table
showing the successive annual average prices from 1871 to 1890 of
three different groups of Sauerbeck’s articles as compared with three
similar groups of his own, recomputing for this purpose his own index
numbers on Sauerbeck's base period of 1867-1877. The comparisons
are made, first, between the general index number for Sauerbeck'S
complete list of 45 articles 87 and his own index number for 100 articles;
second, between the index numbers for their respective groups of
agricultural products, which include 7 articles in Sauerbeck’s list and
20 in Soetbeer’s; and, third, between the index numbers for their
respective groups of minerals and metals, consisting of 8 articles for
Sauerbeck and 14 for Soetbeer. In the same article he makes a
further test by contrasting his index number for 100 articles for the
separate years from 1881 to 1890 with the index mumbers for exports,
for imports, and for both, which are published by the Imperial Statis
ts UVS.IBenate Ex. Doc. No. 34 (p. 226), 50th Cong., 1st sess., 1887.
87 See pp. 276-278 of this bulletin.
GERMANY---- ADOLF SOETBEER.
tical Office.
is used.
259
For this comparison the base period 1881 of the bureau
TABLE OF RESULTS.
In his article on “ The course of prices from 1886 to 1890" 88 Dr.
Soetbeer presents his last published table, which gives his computa
tions of index numbers for the 114 articles by groups, as follows:
T
able
5 3 — SUM M ARY OF R E L A T IV E
PRICES OP COMMODITIES FOR TH E Y E A R S
1847-1890.
(Average prices in 1847-1850=100.)
Year.
1847-1850..
1851-1855..
1856-1860..
1861-1865..
1866-1870..
1871...........
1872............
1873............
1874............
1875............
1871-1875..
1878............
1877............
1878............
1879............
1880............
1876-1880..
1881............
1882............
1883............
1884............
1885............
1881-1885..
1886............
1887............
1888............
1889............
1890............
1886-1890..
I.
V.
VII.
II.
IV .
VI.
III.
Minerals
Agricul Animal Southern
Miscel
Colonial
Textile laneous
and
tural
products, products.
materials.
products.
products.
metals.
articles.
etc.
1 0 0 .0 0
1 0 0 .0 0
129.99
131.84
124.46
137. 74
144.76
144.17
146.21
150.99
138.1&
144.90
141.06
145.34
132. 50
132.92
138.11
138.12
137.59
138.45
143.33
123.85
110. 75
130.77
101.31
96.2S
98.18
102.06
107.53
101.06
114.79
132.31
128. 24
136. 35
144.14
155. 82
156.72
157.76
158.59
154.57
155.79
152.51
141.53
137.60
147.30
146.76
151.21
155.17
156.40
150.26
140.45
150.65
133.53
1 2 a 93
128.97
130.95
129.85
130.41
m .m
1HX43
134.72
114.13
121.54
122.99
125.36
132.15
145.02
131.35
131.50
128.69
14a 55
134.34
139.10
154.65
138.91
146.57
139. 23
142.38
120.16
123.78
134.41
122.44
1 2 1 . 81
1 2 a 09
127.57
13a 61
126.08
1 0 0 .0 0
110.97
122.61
118.64
118.32
1 2 0 .2 2
130. 25
134.32
136. 74
132.11
130.72
129.74
130.29
125.61
123.34
122.92
126,38
122.60
122.47
1 2 a 17
117.90
116.39
119.91
115.45
116. 59
116.41
118.82
119. 35
117.32
loaoo
105.20
107.12
131.83
129.17
119. 23
122. 79
119. 58
112.80
111. 47
117.17
105.54
108.33
102.33
98.76
96. 72
102.33
99.29
95.10
95.93
97.02
95.89
96.65
89.76
81. 42
82.17
89.05
81.92
84.86
loaoo
107.03
113.59
1 0 2 .1 1
95.47
101.85
121.63
140.60
116.70
107. 49
116. 90
106. 27
98. 87
94.14
84.28
8 8 .33
94.35
84.87
86.99
82.93
78.69
74.23
8-1.55
70.52
72.50
75.57
78.55
as. 54
76.12
VIII.
British
exports.
1 0 0 .0 0
1 0 0 .0 0
106.65
108.21
144.33
105.90
117.48
128.54
119.14
1 1 2 . 21
98.74
114.98
1 0 1 .78
99.80
97.24
90.21
95.23
96.79
94.89
99.10
95.38
84.82
81.35
91.11
78.75
77.30
74.31
86.41
91.70
81.70
98.47
102.41
127.56
130. 55
122.64
130.07
128.52
126.06
124.96
126.44
119.23
114.04
111. 03
105.93
108.15
111.70
103.08
104.72
104.72
103.36
100.48
103.28
97.03
95.98
94.91
96.60
94.96
95.90
I-V III.
Total
articles,
114.
loaoo
112. 22
120.91
123. 59
123.57
127.03
135.62
138. 28
136.20
129.85
133.29
128.33
127.70
120.60
117.10
121. 89
123.07
121.07
122.14
122.24
114. 25
10&72
117.68
103.99
1 0 2 .0 2
102.04
106.13
108.12
104.41
GREAT BRITAIN,
INDEX NUMBERS OF THE BOARD OF TRADE.
PUBLICATION.
The first report of this series contains the results of an investiga
tion conducted by the Ministry of Labor for years prior to 1902.89
From that j^ear until 1920 an annual report on the subject was pub
lished in the January issue of The Labor Gazette, London. Since
1920 the index numbers have been published in one of the January
issues of the weekfy Board of Trade Journal and Commercial Gazette.
HISTORY.
The inquiry concerning the subject of
Ministry of Labor for several years, and,
amount of public attention devoted to all
of commodities, it was decided in 1903 to
ss Jahrbucher fur Nationalokonomie und Statistik, 1892, p. 593.
Report on Wholesale and Retail Prices, 1902. Great Britain.
89
prices had occupied the
on account of the great
questions affecting prices
publish the results up to
Board of Trade.
260
INDEX NUM BERS OF WHOLESALE PRICES.
1902 without delay. The first report in 1903 consisted of a series of
comparative tables of actual wholesale prices covering the years from
1871 to 1902.
The index of wholesale prices was computed upon 45 selected arti
cles, the number later being stated as 47. In respect to most of the
articles the actual prices were carried as many years back of 1871 as
was regarded safe, considering the nature of the data available. How
ever, the index was not computed back of 1871, even where the actual
prices of individual articles were secured for earlier years, as in the
instance of bread in the city of London, where the price was carried
back to 1758.
To show the average change of general prices, not only from 1871
to 1902, but extending over the whole of the nineteenth century, a
chart was published covering the period from 1801 to 1902.
The statement showing the course of prices from 1801 to 1846 is
based on Jevons* index number, from 1846 to 1871 on that of Sauer
beck, and from 1871 to 1902 on the Board of Trade index as shown
in these reports.
SOURCE OF QUOTATIONS.
The data used in the original report were import and export aver
age values, contract prices at hospitals and institutions, prices at
markets, ascertained values of coal and iron in different districts used
for the determination of wage rates, prices from private firms, asso
ciations, etc. The import values were based on the declarations of
the importer, those for exports also being declared values, but the
report states that various difficulties were encountered “ in tracing
back the average value of the same article throughout so long a period,
arising to a large extent from changes in classifications.” 90 The same
system of declared values was in force throughout the period. The
contract prices of certain articles for hospitals and asylums of the
London County Council represent the prices paid, throughout the
period, by a somewhat similar class o f consumers. Market prices
were compiled from official reports, newspapers, and market quota
tions. The ascertained values of coal and iron were from reports
made by accountants for use in the determination of the general rate
of wages by sliding scale or otherwise. In a few cases it was found
necessary to secure quotations from the original sources and from
private corporations. A memorandum states that it was proposed to
use either import or export values according as the article was chiefly
one of import or export, except for British corn, milk, potatoes, beef,
mutton, and brick .91
BASE PERIOD.
The year 1871 was originally adopted as the base period, and from
1871 to 1906 the index was computed upon this base for the average
price of all the articles. This was used as a standard until 1906, when
1900 was established as the base and the index for the 45 articles as
a whole was recomputed on the new basis from 1871 to 1906. From
1906 until 1920 the year 1900 has been used as the base period.
91
Report on Wholesale and Retail Prices, 1902, p. 427.
Idem, p. 439.
Great Britain.
Board of Trade.
GREAT BRITAIN---- BOARD OF TRADE.
261
PRICES: H O W SHOW N AND COMPUTED.
In the original report the prices were shown as yearly actual aver
ages for the separate articles. In succeeding reports up to 1905 the
actual average prices were not shown, but the index numbers for the
four groups and the general index number were published. After
1905 no price data other than the general index number were shown
until the publication of index numbers for the four groups was resumed
in the January issue of the Labor Gazette.
NUMBER AND CLASS OF COMMODITIES.
The list of articles covered by the reports includes principally raw
materials or materials at an early stage of manufacture. The Janu
ary, 1914, and subsequent issues of the Labor Gazette state that, in
compiling the general index number, the index numbers for 47 sepa
rate articles have been weighted in accordance with their estimated
consumption in 1881-1890. Counting milk, butter, and cheese as
separate articles (heretofore counted one) probably accounts for the
new number; however, no explanation is vouchsafed.
DESCRIPTION AND GROUPING OF COMMODITIES.
Below is shown the list of the 47 articles included in the group and
general index, the weight alloted to each article, and the source of
the quotations .92
Group I .— Coal and metals (6 articles).
Article.
Allotted
weight.
Pig iron.................................................................................
Copper (ore and regulus)..................................................
Crude zinc............................................................................
Block tin*.............................................................................
Lead......................................................................................
34
16
5
1J
li
li
Total..........................................................................
59*
Source of price quotations.
Export values of coal.
Export values of pig iron.
Import values of copper regulus.
Import values of crude zinc.
Import values of block tin.
Import values of pig and sheet lead.
Group I I .— Textiles (raw materials— 6 articles).
Cotton, raw ..
Wool, British.
Wool, foreign.
Jute, raw____
Flax, raw-----Silk, raw........
Import values of raw cotton.
Export values of sheep and lambs'
wool.
Import values of sheep and lambs*
wool.
Import values of jute.
Import values of flax.
Import values of silk.
T otal...
92
Report on Wholesale and Retail Prices, 1902, pp. xxxv-xxxvii.
Great Britain.
Board of Trade.
262
INDEX NUM BERS OF WHOLESALE PRICES.
Group I I I .— Food and drink (25 articles).
A . Com , etc.
Source of price quotations.
Article.
Wheat, British
Barley, British
Oats, British..
Wheat, foreign
Barley, foreign.
Oats, foreign...
Maize............... .
Hops................ .
Rice...................
.Potatoes.......... .
Gazette average of British wheat.
Gazette average of British barley.
Gazette average of British oats.
Import values of wheat.
Import values of barley.
Import values of oats.
Import values of maize.
Import values of hops.
Import values of rice.
Contract price, potetoesatSt. Thomas’s
Hospital.
Total___
B . Meat, fish, and dairy products.
Beef..................................
Mutton............................
Bacon...............................
Milk, butter, cheese, etc,
E£gs.....................
Herrings..........................
Beef (live), 1st class, Metropolitan Cat
tle Market.
Mutton (live), 1st class, Metropolitan
Cattle Market.
Import values of bacon.
Averageprice of milk at Bethlam Royal
Hospital and St. Thomas’s Hospital.
Import values of eggs.
Export values of herrings.
Total.....................
C. Tea, tobacco, ivine, and sugar.
Sugar..........
Tea.............
Coffee......
Cocoa..........
Rum ...........
W ine..........
Tobacco___
Import values of refined sugar.
Import values of tea.
Import values of coffee.
Import values of cocoa.
Import values of rum.
Import values of wine.
Import values of unmanufactured to
bacco.
Total
Group I V .— Miscellaneous (10 articles).
Cotton seed..........
Linseed.................
Olive oil................
Palm oil................
Paraffin.................
Petroleum............
Bricks...................
Hewn fir............. .
Caoutchouc..........
Hides.....................
Import values of cotton seed,
Import values of linseed,
Import values of olive oil.
Import values of palm oil.
Import values of paraffin,
Import values of petroleum,
Price of stocks at Glasgow,
Import values of hewn fir.
Import values of caoutchouc,
Import values of hides.
Total..........
Grand total
SUBSTITUTIONS AND ADDITIONS.
Various difficulties, as was previously stated, were met in tracing
average values throughout so long a period. Changes in classification
were the cause to a large extent. The methods adopted in making
substitutions are not fully explained. Apparently no new articles
have been added since the publication of the first report.
GREAT BRITAIN— BOARD OF TRADE.
§63
INTERPOLATION.
In the discussion of sources of information the statement is made
that where the data related only to some of the earlier years of the
period covered, or could not be continued to the present, they were
omitted, and that when large gaps existed in the records it was the
general practice as far as possible to start the table from dates sub
sequent to the gaps in order to preserve continuity . 03
WEIGHTING.
The method of weighting used in computing this index number was
that based on the amount of consumption o f the various articles in
the United Kingdom in the period 1881-90. The consumption of an
article is defined to mean any process by which the commodity is
substantially changed in character. The original report in its ex
planation of the u consumption standard” states that “ the theoreti
cal basis of the consumption standard is the proposition that the true
measure m the change of the value of money is the change in the
amount of gold that must be paid by consumers throughout the
country for all commodities in their finished state consumed by them
per unit of time. ” 94 The value of the national consumption of the
23 raw materials which were derived almost entirely from foreign
sources was taken to be the declared value of the imports less the
declared value of the exports. The value of the consumption of the
remaining articles was the value of the quantity produced plus the
value of the amount imported, if any, minus the value of the exports,
if any. The results thus obtained represent the estimated value in
millions sterling of the annual consumption of the articles in the base
period. The millions sterling constitute the weights allotted.
The weights assigned to the various articles were placed .against the
percentage variations in prices. The percentage variations were com
puted for each year by using 1871 as the base, or 100 . For example,
the percentage price of coal in 1872 was 161.1 per cent, the price in
1871 being 100. This percentage multiplied by 34— the weight
allotted to coal— produced 5,477.4, or what was termed the weighted
percentage. The sum of the weighted percentages of all the articles
in a group divided by the sum of the weighted percentages for the
base period produces the index number for the group in the specified
year. For example, the weights for the group of coal and metals were
5,950.0 for the base year, and for 1872 the total was 9,173.2., which
divided by the figure for the base year equals 154.1, the published
index number o f the group for 1872. A continuation of this process
produces the other group indexes, and the general index for the 47
commodities is computed in like manner.
TESTING.
Some comparison of results was made with the results of other
indexes. The principal test was made by using certain articles as
given by Sauerbeck. These articles covered only 28 of the 45 price
quotations in his report, but they formed nine-tenths of the total
weight of the Board of Trade index number. These articles were
used to form a special index number, making use of the weights
allotted as above. The results are shown below.
83
m
"Report on Wholesale and Retail Prices, 1902, p . 426.
Idem, p. 432.
Great Britain.
Board of Trade.
INDEX NUMBERS OF WHOLESALE PRICES.
264
Table
5 4 .—COMPARISON OF A SPECIALLY W E IG H T E D NU M BER FROM THE BOARD OF
T R A D E AN D SAU ER B EC K ’ S N U M B E R . 1
Specially
weighted
number.
Year.
1867................................
1868................................
1869...............................
1870...............................
1871...............................
1872...............................
1873................................
1874................................
1875................................
1876................................
1896...............................
Sauerbeck’s
number.
*00
100
100
99
98
96
96
94
96
107
113
104
97
96
62
i Report on Wholesale and Retail Prices, 1902, p. 449.
100
109
111
102
96
95
61
Great Britain.
Board of Trade.
TABLES OF RESULTS.
The principal table of the original report shows the index numbers
for the four general groups, three subgroups under food and drink,
and the index for all commodities .95 This table is reproduced be
low and carried forward to 1905, the last year that the index was
computed upon the basis of 1871 as 100.
T a b le 5 5 .—
BO AR D OF T R A D E IN D E X NUM BER S, B Y GROUPS, 1871 TO 1905.
fThe index numbers are the sum of the weighted percentages divided by the sum of the weights.]
(Average prices in 1871=100.)
Year.
1871.................
1872.................
1873.................
1874.................
1875.................
1876.................
1877.................
1878.................
1879.................
1880.................
1881.................
1 882..............
1883.................
1884.................
1885.................
1886.................
1887.................
1888.................
1889.................
1890.................
1891.................
1892.................
1893.................
1894.................
1895.................
1896.................
1897.................
1898.................
1899.................
1900.................
1901.................
1902........i . . . .
1903.................
1904.................
1905.................
*>
Index
number
for all
the 45
commodi
ties.
1 0 0 .0
1 1 0 .6
118.8
113.6
107.8
104.2
105.3
99.3
94 9
97.4
95.7
97.3
96.5
88.3
83.0
78 5
76.7
79.3
80.8
82.8
84.1
80.1
78.7
75.1
72.2
69.8
71.3
73.6
74.5
83.2
79.2
78.8
78.6
78.7
77.7
III.
1 0 0 .0
1 0 0 .0
1 0 0 .0
154.1
194.8
158.8
126.3
107.2
99.9
92.9
114.0
104.0
109.7
8 6 .8
94.9
91.1
91.8
90.3
86.4
82.1
78/8
80.1
83.0
94.1
113.6
106.6
98.8
89.4
91.7
85.8
83.3
84.4
92.7
107.5
151.9
124.7
114.9
1 1 1 .2
106.1
105.6
1 1 0 .2
1 0 2 .6
1 0 0 .2
93.6
91.7
88.5
84. J
8 8 .6
87.0
84.1
82.0
79.8
75.7
69.0
70.7
70.0
72.4
72.9
70.1
6 6 .1
6 6 .6
60.8
57.7
64.0
59.7
54.8
57.2
70.0
65.7
65.0
71.3
78.7
73.5
IV .
Miscella
neous.
1 0 0 .0
1 0 2 .0
1 0 0 .0
1 0 2 .0
1 0 0 .0
1 0 0 .0
109.3
98.1
94.9
93.9
90.7
96.9
88.4
85.2
102.7
108.2
108.6
107.3
106.8
110.3
104.2
100.5
8 6 .2
1 0 0 .8
84.4
83.6
80.9
70.7
66.7
63.7
61.4
65.0
99.4
102.3
105.1
109.5
108.4
99.3
97.9
97.3
88.3
81.5
89.2
88.5
89.1
87.5
82.2
80.2
73.8
69.3
71.0
74.3
72.9
70.1
1 1 0 .8
1 1 0 .1
99.9
98.7
107.9
98.9
97.4
98.0
94.9
95.8
93.9
81.1
76.7
71.8
71.2
71.7
70.6
72.0
83.2
73.4
68.3
63.1
62.2
57.6
62.7
73.1
63.7
62.4
64.0
63.7
63.8
66.7
64.8
116.6
117.2
114.9
R eport on Wholesale and Retail Prices ,1902, p. 34.
Food and drink.
II.
I.
Textiles
Coal and
III c .
i n B.
(raw
metals. materials).
Meat, fish, Sugar, tea,
I l l A.
Total,
Com, etc. and dairy wine, and group III.
tobacco.
produce.
1 1 2 .2
106.7
106.6
106.7
1 1 2 .0
113.7
104.7
96.3
92.7
8 8 .2
94.0
92.8
91.7
91.1
91.8
95.2
92.2
8 8 .2
81.9
84.6
81.8
85.8
90.3
89.8
94.4
92.1
89.0
88.4
Great Britain.
1 0 2 .1
6 8 .0
63.8
64.8
63.9
65.0
59.5
56.2
57.0
54.9
54.2
52.5
52.4
50.1
46.1
47.0
48.2
52.1
Board of Trade.
91.4
85.1
81.1
78.3
81.8
81.1
80.6
84.9
81.3
81.1
76.9
74.2
69.4
72.4
75.1
73.2
74.9
75.3
76.7
75.7
75.5
74.8
6 8 .1
6 6 .2
62.7
62.6
63.6
63.9
66.5
67.9
74.3
71.7
69.2
68.4
6 6 .0
68.7
GREAT BRITAIN---- BOARD OF TRADE.
265
A second table, reproduced below, shows the index as published up
to the end of 1920, the price in 1900 being used as the base, or 100.
The exact method of computation upon the new base period is not
stated, but a note to the report for 1906, issued in January, 1907,
states that “ the index number has, however, now been recalculated
with the year 1900 as its base year instead of 1871.” 96 It is pre
sumed that the calculation was made in the same way as in the
original report.
T able
5 6 .—BOARD OF TRA DE IN D E X NUM BERS, 1871 TO 1920.
[The Board of Trade Journal and Commercial Gazette, London, January 13,1921, p. 34.
(Average prices in 1900=100.)
Index
number.
Year.
1871
1872
1873
1874
1875........
1876
1877
1878
1879
1880
1881
Year.
1882
1883
1884
188 5
1886........
1887
1888
1S89
1890
1891
1892
135.6
145.2
151.9
146.9
140.4
137.1
140.4
131.1
125.0
129.0
126.6
Index
number.
127.7
125.9
114.1
107. 0
1 0 1 .0
98.8
101. 8
103.4
103.3
106.9
1 0 1 .1
Index
number.
Y ear.
1893........
1 8 9 4 .....
1895
1896...
1897...
1898........
1899. .
1900.
1901
1902........
1903...
99.4
93.5
90.7
8 8 .2
90.1
93.2
92.2
100.0
98.7
96.4
96.9
Year.
1904........
1905........
1906
1907...
1908...
1909
1910
1911
1912
1913. .
Index :
number.
98.2
97.6
1 0 0 .8
106.0
103.0
104.1
108.8
109.4
114.9
116.5
Index
number.
Year.
1914:
Jan.-July
Aug.-Dee
191 4
191 5
1916
1917
1918
191 9
j 19201___
113.6
1 2 2 .6
117.2
143.9
186.5
243.*)
267.4
296.3
371.4
i Preliminary.
In Table 57 are shown the index numbers of the 47 articles since
1900, classified into four groups.
T
able
5 7 . — IN D E X
NUM BERS OF 47 AR TICLES, CLASSIFIED B Y GROUPS, 1900 TO 1920.
[The Board of Trade Journal and Commercial Gazette, London, January 13, 1921, p. 34.]
(Average prices in 1900=100.)
Year.
1900..........................................
1 9 0 1 ... :
1 9 0 2 ... :
190 3 ..................................
190 4
190 5
190 6
190 7
1 9 0 8 ..:.....................................
190 9
191 0
191 1
191 2
191 3
1914 (January to July).........
1914 (August to December) a
1914 (year)...............................
191 5
191 6
.
191 7
.
191 8
..
191 9
.
1920 6 ..........................................
Coal and
metals.
Textiles
(raw ma
terials).
Food,
drink, and
tobacco.
100.0
100.0
100.0
82.2
76.1
74.1
70.9
71.3
78.3
86.9
78.5
73.6
76.6
74.7
84.9
92.5
86.2
88.8
86.7
116.7
165.8
182.0
204.9
280.2
419.3
a First five months of the war.
b Preliminary.
w The Board of Trade Labour Gazette, vol. 15, 1907, p. 4.
93.3
92.3
101.7
112.9
106.7
121.1
127.4
109.8
112.4
136.2
128.9
119.6
135.0
135.1
116.8
128.8
119.8
180.1
270.4
354.4
373.3
508.1
100.1
101.4
100.6
101.2
101.2
101.0
105.5
107.0
108.7
109.2
111.6
119.9
117.7
114.8
130.4
120.9
154.1
189.4
246.2
259.3
279.7
335.1
Miscella
neous.
100.0
96.3
92.5
91.7
88.3
.91.1
95.6
99.7
94.8
98.5
104.3
105.5
110.1
109.4
106.2
119.1
111.3
143.8
204.0
256.3
268.6
317.8
352.2
All articles
combined.
100.0
96.7
96.4
96.9
98.2
97.6
100.8
106.0
103.0
104.1
108.8
109.4
114.9
116.5
113.6
122.6
117.2
143.9
186.5
243.0
267.4
296.5
371.4
266
HSTDEX LU M BERS OF WHOLESALE PRICES.
CHANGE IN M ETH OD OF COMPILING WHOLESALE PRICE INDEX NUMBERS OF THE
BOARD OF TRADE.
In the issue of the Board of Trade Journal and Commercial Gazette
of January 13, 1921, it was stated that the index numbers for 1920
were the last to be calculated according to the method adopted for
the series given in the foregoing table, and that an account of the
method to be followed in compiling the index numbers for 1921 and
succeeding years would appear in the next issue of the journal. In
the issue of January 20 , 1921, the following explanation was given:
For a number of reasons it has been decided to revise the basis of
calculation, beginning with the present year. Several changes of
considerable importance are being made in the nature of the material
used in the compilation and also in the manner in which the different
elements of the calculation are combined. The average import or
export values hitherto used are to be replaced by market values.
In a few special cases, as in the past, so also in the future, reliance
will be placed on values furnished to the board bv experts in the
absence of a satisfactory published quotation of the prices of the
articles concerned. A further point of importance is that, instead of
multiplying the price percentages by suitable factors, the number of
separate prices used will be increased, so that articles of special im
portance, such as wheat, coal, iron, and cotton, will be represented
by several quotations. The number of quotations to be used in each
case has been determined on the same general principles as the mul
tipliers in the old number, but the results of the census of production
have rendered possible a complete revision, based on the values of
goods made in many cases in which the value of raw materials worked
up were alone ascertainable prior to that census. The total number
of series of commodity prices which it is proposed to use at present
is 150, or three times the number hitherto employed. The quotations
will relate in numerous instances to standard manufactured commo
dities, in others to raw or semimanufactured materials.
To avoid various inconveniences resulting from the use of prices
based on those of a fixed year, the calculation to be made in the first
place will be the percentage movement of prices over a period of one
year, each month’s figure showing the increase or decrease compared
with the corresponding month a year earlier. This procedure will
avoid the difficulties which occur when, owing to changes in business,
commodities once serving as standards o f comparison are superseded
by other commodities or other grades. The extension of the list of
commodities, when necessary, will also be facilitated in the same way,
the calculations forward from any date not being hampered b y the
necessity of securing comparisons with prices at a past date from
which tne calculations have started. The combination of the series
o f yearly comparisons so as to yield continuous record can be easily
made. The results will not be dependent on the initial date of the
series, as is generally the case when the procedure described above
as used in the old index number is followed.
The 150 series of quotations are to be arranged in eight groups of
approximately equal importance, three for foodstuffs and five for
industrial products. The foodstuffs groups comprise cereals, meat
and fish, and other foods; the industrial products are grouped as
iron and steel, other metals and minerals, cotton, other textiles, and
miscellaneous industrial products. The separate index numbers for
GREAT B R IT A IN -----BOARD OF TRADE.
267
the eight groups will be prepared, and the aggregate index number
will be the average of these eight numbers. Finally, the geometric
mean of the individual items is to be used in place of the more com
monly employed arithmetic mean, this course being adopted for
various technical reasons. It is anticipated that the new series of
numbers will reflect more closely than has been the case with the
older series in the recent past the movements of the wholesale mar
kets, the special conditions of trade during and since the war having
had the effect of diminishing in this respect the value of the index
number as hitherto calculated.
The Board of Trade Journal and Commercial Gazette for March 17,
1921, contains the following table of new index numbers for the
period from January, 1920, to February, 1921:
Percentages o f 'prices to the average 'prices o f 1920.
1920
Groups.
Jan. Feb. Mar. IApr. May June July. Aug. Sept. Oct. Nov. Dec. Jan. Feb.
Cereals................................. .
Meat and fish..................... .
Other food...........................
Total food................
Iron and steel......................
Other metals and minerals
Cotton................................. .
Other textiles.................... .
Other articles......................
Total not food......... .
All articles...............
i.2 90.5 78.4
89.3 88.8 97.7 101.6 102.5 104.3 103.0 102.2 103.3 106.7 106.3
91.9 88.6
6 91.8 92.9 102.4 106.0 106.9 109.1 111.4 110.6 108.0 100.8
98.
1 94.2 97.2 94.5 92.3 88.0 81. G
94.9 101.4 106.5 108.2 106.7 108.5 101. 4
94.2 94. 2 97.7 101.3 100.4 102.0 102.2 101.2 101.0 104.0 103.4 99.1 94.
82.7 90.2 96.4 100. 105.4 109.1 108.2 107.0 107.0 104.2 99.2
98.7 101.8 100.3 97.0 101. 0101.0 101.6 102.8 103.6 102.3 100.1
112.4 127.0 .127. 4 127.9 123. 115.7 :108.0 105.3 96.1 78.9 65.8
113.6 120.6 121.7 123. 4 117.7 108.7 99.2 94.6 93.9 84.2 76.3
99.0 104.0 107.1 108.0 104.9 100.2 99.9 98.2 100.8 99. S 94.3
98.3 105.7
108.0108.9 109.0 106. ‘J
>3.5 101.9 100.9 94.9
101.5104.2, 196.1 105.9 104. S 103.0 101.6 101.9
98.0.
92.3
52.9
67.0
85.9
8
8.6
85.5
46.7
61.1
80.4
86, 2
79.0
80.7
40.6
55. ft
78.6
79.2 73.1 67.3
85,7
73.4
N o t e . —In this table the aggregate index number is the geometric average of the 150 separate index num
bers for individual commodities.
INDEX NUMBERS OF THE ECONOMIST.
PUBLICATION.
This index represents the course of wholesale prices of commodi
ties in the United Kingdom.
It is compiled and published each month in the Economist, London,
a summary being given also for the period since January 1 , 1914.
HISTORY.
The object of this compilation originally was to throw some light
on the relation between the gold supply and prices. In 1849 gold had
been discovered in California and in 1850 in Australia, and the pouring
of this gold into Europe seemed to be accompanied by a general
upward movement of prices. It was to ascertain whether there had
been such a movement and, if so, its extent that the Economist index
numbers were developed.
In 1859 William Newmarch, then editor of the Journal of the
Royal Statistical Society, published an article in that periodical on the
prices of the previous year, in which the prices of 19 commodities in
the London market were expressed as percentages of the average
268
INDEX NUMBERS OF WHOLESALE PRICES.
of the prices of 1845-1850. These commodities were as follows:
Coffee, sugar, tea, tobacco, wheat, butcher’s meat, cotton, silk,
flax and hemp (average), wool, indigo, oils (average of 3 varieties),
timber, tallow, leather, copper, iron, lead, and tin. In 1860 and 1861
similar articles appeared in the Journal, when in addition to these
19 commodities three others were added: Raw cotton, cotton varn,
and cotton cloth— all at Manchester prices. The prices of these 22
commodities were expressed in the form of percentages, but no
general index number was constructed from them.
The Economist stated in its issue of February 20 , 1864, in which it
published for the first time its commercial history and review of the
past year, that in the table of actual wholesale prices it was following
the arrangement and method which were adopted by Tooke and Newmarch in their history of prices, and continued by Newmarch in the
Journal for 1859, 1860, and 1861.
The first table in the Economist report of 1864 presented the actual
prices in pounds sterling of 45 articles for the base period of 18451850 and for succeeding years down to 1862. The prices were for
a given date, being either those for January 1 or July 1 for all years
previous to 1863, tor which year they were the prices for the 1st of
each month.
The 45 articles were coffee, sugar (3 kinds), rum, tea, tobacco,
butter, wheat, beef (2 kinds), mutton (2 kinds), pork, cotton, silk,
flax, hemp, wool (4 kinds), dyes (2 kinds), oils (3 kinds), timber (2
kinds), tallow, leather, saltpeter, ashes, copper, iron (2 kinds), lead,
steel, tin, raw cotton (3 kinds), cotton yarn, and cotton cloth (2 kinds).
For a few of the articles the prices were not continuous throughout
the period.
A second table was printed entitled “ Proportionate results/ ’ being
the percentage that the actual price of each article for the given date
was of the actual average price for 1845-1850. In this table, instead
of 45 series of percentages, the number was reduced to 2 2 . This
number was made up of 37 of the 45 series of quotations under the
heads of coffee, sugar, tea, tobacco, wheat, butcher’s meat, cotton,
raw silk, flax and hemp, sheep’s wool, indigo, oils, timber, tallow,
leather, copper, iron, lead, tin, raw cotton (cotton wool), cotton
yarn, and cotton cloth. Of these 22 series of percentages, the 9
composed of more than one description or grade of the article were
sugar (2 kinds), butcher’s meat (2 kinds of beef and 2 of mutton),
flax and hemp (2 articles), sheep’s wool (4 kinds), oils (3 kinds), iron
(2 kinds), raw cotton (3 kinds), cotton cloth (2 kinds), and timber
(2 kinds).
The articles were divided into five groups as follows:
I. Colonial and tropical produce (food).
II. Wheat (England and Wales) and butcher’s meat (New
gate market).
III. Raw materials of manufacture.
IV. Metals.
V. Manchester markets.
The articles under the fifth head were raw cotton, cotton yarn, and
cotton cloth.
During the y^ars 1864 to 1867 the composition of these tables
remained the same, except that in 1865 the percentage for raw cotton
was computed upon one grade instead of three, as formerly.
GREAT BRITAIN— THE ECONOMIST.
269
The commercial history, published by the Economist, for 1868
gave for the first time the total index number. However, this was
simply the total at each date of the 22 percentage columns, no
general index being computed and it was not until 1869 that the
numbers were added together and divided by 22 , the result thus
becoming the “ Economist” index number, which has been pub
lished year by year since that date.
It was announced in the Economist of February 4, 1911, that it
was deemed desirable to change the basis upon which the index
number had been calculated. This statement recited the intention
of the publication to make this review of prices more far-reaching by
embodying in the index quotations of some important articles which
play a large part in modern commerce, and at the same time to retain
its character as a wholesale market index number. On account of
the inclusion of new articles, it became necessar}^ to adopt a base
period sufficiently recent to include standardized quotations of
modern commodities.
It Was stated that, owing to the fact that many commodities are
now important in the business life that were not so regarded at the
time of beginning the index, the list of commodities had been re
vised and the number increased. The result of this recasting was
published in the issue of November 18, 1911, showing how the new
index number was made comparable with the old figures by dividing
the total index number for the 44 commodities by 2, thus reducing
it to that of 22 articles as used formerly.
The chief change made was in respect of the coal and iron trades,
which were formerly represented by one quotation only, but wTere now
given a weight of 5 quotations out of 44. Quotations were added, for
the first time, for barley, oats, potatoes, rice, pork, and butter among
foodstuffs; Egyptian cotton and jute among textiles; iron bars, steel
rails, and coal among minerals; and petroleum, oilseeds, rubber, and
soda crystals in the miscellaneous group.
In order to show the relation between the percentage index as com
puted by the old method and the index number obtained by the new
plan, the latter has been reduced to the same basis. The index num
ber, however, is based on the prices of 44 articles, while the old per
centage number was computed on but 22 . They have been made
comparable, as stated, by dividing the index number for the 44 arti
cles by 2 .
SOURCE OF QUOTATIONS.
The quotations used in compiling this index are market prices as
ublished weekly in the Economist, which represent those of the
iondon markets.
E
BASE PERIOD.
As has been explained, the base period formerly was 1845-1850, but
in November, 1911, it was announced that the base period had been
changed to 1901-1905.
PRICES: H O W SHOW N AND COMPUTED.
As was previously stated, the quotations used are those published
weekly in the Economist. A review of price changes among the se
lected articles is published each month in the discussion of the index
number.
INDEX NUMBERS OF WHOLESALE PRICES.
270
NUMBER AND CLASS OF COM M ODITIES.
The original number of commodities, as has been stated, was 22.
In 1911 the number was increased to 44. Raw, or what might be
termed primary, commodities only are included in these quotations.
Table 58, which appeared in the Economist of November 18, 1911,
shows the number of quotations for each commodity, comparing the
old with the new base period:
5 8 .—N UM BER OF COMMODITIES: SERIES OF Q U O TATION S U N D E R THE OLD
BASIS OF 1845-1850, COMPxARED W IT H T H A T UN D ER THE N E W BASIS OF 1901-1905.
T able
Old basis New basis
quotation quotation
number. number.
Commodities.
Wheat and flour.......................
Barley.........................................
Oats.............................................
Potatoes.....................................
Rice.............................................
Beef..................
....
M utton.......................................
Pork............................................
Sugar...........................................
Coffee...........................................
Tea..............................................
Tobacco......................................
Butter.........................................
Cotton (raw, yarn, cloth).......
Wool............................................
Flax.......................... .................
Hemp
Jute..............................................
Silk..............................................
3
1
1
1
1
1
1
2
1
1
1
2
1
1
1
1
1
1
4
4
1
1
2
j
!
Commodities.
Old basis New basis
quotation quotation
number. number.
Pig iron.........................................
Steel rails.....................................
Iron ba^s.......................................
Coal................ ..............................
Copper..........................................
T in...............................................
Lead..............................................
Timber.........................................
Leather.........................................
Oil..................................................
1
1
1
1
1
1
1
1
1
Petroleum.................................. 7
Rubber.........................................
Tallow..........................................
Indigo...........................................
Soda crystals...............................
1
1
1
1
1
2
i
1 1i
Total.......... ......................
22
44
1
DESCRIPTION AND GROUPING OF COMMODITIES.
The detailed table of prices week by week as published until recently
includes 38 articles. These are arranged in five groups as follows: 97
Cereals and meats.
W heat, Gazette averages (English grain).
B arley, Gazette averages (English grain).
Oats, Gazette averages (English grain).
F lou r, town made, households.
B eef, inferior.
B eef, prime.
M utton, prime.
Potatoes, good English.
R ice, Rangoon.
Other foodstuffs.
Sugar, granulated.
T ea, Indian and Ceylon, broken and fannings.
T ea, Indian and Ceylon, broken Pekoes, common.
Coffee, Santos, good average.
Butter, Danish.
Tobacco, mean price Virginia leaf, common to fine.
Textiles.
Cotton, m iddling, American.
Cotton, yarn, 32’s, twist.
W ool, Victoria, scoured, good.
W ool, Southdown ewes and wethers.
S ilk, Canton.
H e m p , manila.
Jute, native firsts.
” Economist, Jan. 6, 1917. p. 5.
GREAT BRITAIN— TH E ECONOMIST.
271
Minerals.
Iron and steel, Cleveland, N o . 3, G . M . B .
Iron an d steel, com m on, bars.
Iron and steel, Ffeeei rails.
Coal, pit-head prices, best steam, Newcastle.
Coal, p it-h ea d prices, best Yorkshire (siikstone), house.
Copper, standard.
Tin, standard.
Lead, English pig.
Miscellaneous.
Tim ber, pitch pine.
Tim ber, yellow pine, large.
Leather, m ixed taanage, outts or bends.
Petroleum.
R ubber? fine, hard, Para.
Seeds, linseed.
Tallow , town.
Indigo, Bengal, good red violet.
SUBSTITUTIONS AND ADDITIONS.
It is stated that over so long a period of time some variations have
inevitably arisen in the mode of quoting prices, but in all such cases
the nearest approach possible has been made to a uniform quotation.
The articles cited are raw cotton, tea, sugar, flax, and wool. It is
further stated that in some cases,, where it has been considered
desirable to introduce a commodity to replace one no longer actively
dealt in, the current price of the substituted article has been taken as
equivalent to the same percentage of the basis price as was repre
sented by its predecessor.
INTERPOLATION.
The supplying of missing data, if such has been found necessary,
nas not been noted.
WEIGHTING.
The index is computed by means of simple arithmetical average.
Indirect weighting is attained by th§ selection of articles.
This method, as applied to the present index, has been frequently
criticized because of the small number of articles included. An
attempt to correct the fault of giving each article an equal weight
was made by Mr. R. H. Inglis Palgrave, in 1886, in a memorandum
to the Royal Commission on the Depression of Trade and Industry.
The method used b y him was to give each relative price an importance
proportional to the consumption of the article, which was ascer
tained by adding to the production the imports and deducting the
exports. He thus obtained a series of figures representing the im
portance, in each year, of the consumption of each commodity, and
used these in connection with the Economist figures for the years 1865
to 1885 upon the basis, 1865-1869 equals 100 . The data prepared by
Mr. Palgrave in 1886 have not been continued for subsequent years
(see Report of United States Senate Finance Committee on Whole
sale Prices, Wages, and Transportation, 1893, Pt. I, pp. 228, 229).
TESTING.
The testing of the accuracy of the results secured in this index is
made by comparison with the results in other index compilations.
Table 59 below shows a comparison by decades of the weighted and
unweighted indexes of the Economist with those of Sauerbeck (an
272
INDEX NUMBERS OF WHOLESALE PRICES.
unweighted index) and the Board of Trade (a weighted one). The
weights for the Economist index have been calculated on the basis
of consumption in the country as estimated by the Board of Trade .08
The decade 1891 to 1900 is considered as the base or 100.
T able
5 9 .-I N D E X NUMBERS OF T H E L E V E L OF PRICES IN TH E U N ITED KINGDOM,
B Y DECADES, 1861 TO 1910.
(Average prices in 1891-1900=* 100.)
Economist.
Sauerbeck.
Decade.
Weighted.
1861-1870..........
1871-1880..........
1881-1890..........
1891-1900.........
1901-1910..........
Board of
Trade.
Unweighted.
152
131
108
151
144
113
100
100
no
108
100
110
146
131
107
138
138
112
100
111
The Economist of August 26, 1911, presented as a test of the accu
racy of its index a table to show that retail prices have pursued n^uch
the same course as wholesale prices. The following comparison is
made of the Economist index with that of the Board of Trade for
retail prices in London from 1895 to 1910, in each case the year 1900
being taken as the base.
T a b l e 6 0 . — COMPARISON
OF W H O L E S A L E AN D R E T A IL PRICES, 1895 TO 1910.
(Average prices in 1900=100.)
Year.
1895............................
1896................................
1897................................
1898................................
1899................................
1900................................
1901...............................
1902................................
1903...............................
1904...............................
1905................................
1906................................
1907................................
1908................................
1909................................
1910................................
Board of
Economist Trade
in
index
number on dex num
ber
of
retail
Jan. 1 of
prices in
each year.
London.
89
93
91
88
93.2
92.0
96.2 1 0 0 .8
89
96.4
100
1 0 0 .0
99
91
93
103
99
109
117
108
103
101.9
112
1 0 1 .6
103.2
104.3
103.7
103.2
105.8
108.4
108.2
109.9
TABLES OF RESULTS.
Table 61, covering the period from 1851 to 1910, shows the total
index number for the 22 commodities in the form in which it was
published in earlier years.
“ The Econom ist, A ug. 26,1911, p p . 422, 423.
GREAT BRITAIN— TH E ECONOMIST.
273
T able 6 1 .—T H E ECONOMIST IN D E X (ORIGINAL).*
(Average prices in 1845-1850= 2200.)
1845-50: Aver
age..................
1851: January..
1853: July.........
1857: July.........
1858: January..
1859: January.,
1860: January..
1861: January..
1862: January.,
1863: January.,
1864: January.
1865: January.
1866: January.
1867: January.
1868:
January...
July..........
1869:
January...
July..........
1870:
January...
July..........
1871:
January...
July..........
1872:
January...
July..........
1873:
January...
July..........
1874:
January...
July..........
1
2200
2310
2463
3059
2667
2556
2713
2751
2878
3492
3787
3575
3564
3024
2582
2826
2666
(2)
2689
2711
2590
2640
2835
3054
2947
2914
2891
2779
1875:
January...
July..........
1876:
January...
July..........
1877:
January...
July..........
1878:
January...
July..........
1879:
January...
July..........
1880:
January...
July..........
1881:
January...
July..........
1882:
January...
July..........
1883:
January...
July..........
1884:
January...
July..........
1885:
January...
July..........
1886:
January...
July..........
2778
2692
1887:
January..
July.........
2059
2116
January..
July.........
1889:
January..
July.........
1890:
January..
July........
2239
January..
July........
1892:
January..
July........
2224
2190
2120
2098
2048
January..
July........
1894:
January..
July........
1895:
January..
July........
1896:
January..
July........
1897:
January..
July........
2023
2023
January..
July-------
27112531
2715
2625
2554
2457
2225
2299
2577
2479
2376
2302
2435
2442
2343
2220
2221
2169
The Economist, Sept. 2, 1911, pp. 490 and 491.
2
2121
2187
2161
2236
2259
2133
2081
2105
1974
1923
1931
1999
1947
1950
1885
1915
1899:
January.
July..........
1900:
January.
J u ly ....
1901:
January.
July----1902:
January.
July----1903:
JanuaryJ u ly ....
1904:
January.
J u ly ....
1905:
January.
July----1906:
January.
July----1907:
January.
July----1908:
January.
July----1909:
January.
July----1910:
January.
J u ly ....
1918
2145
2211
2126
2007
1948
1995
2003
2111
2197
2130
2136
2163
2342
2362
2499
2594
2310
2190
2197
2240
2390
2362
Figures not calculated for July 1,1869.
A comparison of the Economist index number as computed on the
old and new bases is afforded in Table 62:
T able
62»—COMPARISON OF ECONOMIST IN D E X NUM BERS COMPUTED ON OLD
AND N E W BASES .1
Date.
1896:
January 1
July 1 . . . .
1897:
January 1
July 1 . . . .
1898:
January 1
July 1___
1899:
January 1
July 1 . . . .
1900:
January 1
July 1 ----1901:
January 1
July 1 . . . .
1902:
January 1
July 1 ----1903:
January 1
July 1 . . . .
1
Old basis
(1845-1850=
100).
New basis
(1901-1905=
100).
91
88*
90
88
89
90
87
89
95
87
92
100*
110
111
97
91
106
103
97*
101
91
96
99*
104*
Date.
1904:
January 1
July 1 ----1905:
January..
July 1 ----1906:
January 1
July 1 . . . .
1907:
January 1
July 1 ----1908:
January 1
July 1----1909:
January 1
July 1 . . . .
1910:
January 1
July 1 . . . .
1911:
January 1
The^e data are taken from the Economist of Nov. 18,1911, p. 1035.
3 3 2 2 6 °— 21— B ull. 284--------18
Old basis
(1845-1850=
New basis
(1901-1905=
100
102
100).
100).
97
99
97
98
104
102*
106
107
109
114
118
115
105
111*
100
106*
100
102
104
109
107
113*
113
114
114
110
121
110
274
INDEX NUM BERS OF WHOLESALE PRICES.
Table 63, reproduced from the Economist of January 8 , 1921,
shows the manner in which the information relative to the index
number is now presented.
T able
6 3 ,—IN D E X NUM BERS OF W H O L E SA L E PRICES IN G R E A T B R IT A IN , IN SPECI
FIED MONTHS, 1914 TO 1920.
[The Economist, Jan. 8 , 1921, p. 44.]
Date.
Other
food
products
(tea
sugar,
etc.).
Cereals
and
meat.
Min
erals.
Tex
tiles.
Miscella
neous
(rubber,
timber,
oils,
etc.).
Total.
Percent
age
change.
Basis (average, 1901-1905)........
500
300
500
400
500
2200
100.0
January 1, 1914..............................
April 1 , 1914...................................
July 1, 1914.....................................
Ena July, 1914..............................
End December, 1£14....................
End December, 1915.....................
End December, 1916.....................
End December, 1917.....................
End October, 1918........................
End November, 1918...................
End December, 1918.....................
End January, 1919........................
End February, 1919.....................
End March, 1919...........................
End April, 1919.............................
End May, 1919...............................
End June, 1919..............................
563
560
565^
579
714
897
1294
12861
1271
1289
1303
1287
12881
1285
13061
1310£
1338
1339*
1380“
1-399
1412
14271
14411
1461
1454
1508
1498J
1484
1511
1499
1438
1504
15601
1478
1344
355
350%
345
352
4141
446
553
491
493
4711
4641
476
7111
8241
839*
878*
903“
572
567
551
553
6861
8481
780
7821
782|
782i
7821
782*
752“
776*
800“
805%
8221
817*
838“
642
626J
616
616*
509
731
11241
16841
1889
1848
18051
1618*
1596J
1502
15121
1643
17411
1854*
18771
19791
2123
866
22021
8811
857J
886*
914
908
9811
9291
934
927
928
900|
869£
805
24421
27021
29511
2974£
29381
2819
2562
2594
2521
23621
19511
1651
1284
2623
2597
2549
2565
2800
3634
4908
5845 ,
6210
6212
6094
5851
5796
5708
5774
5988
6188 ;
6450
6503
6587
6795
6985
7364
7768
8160
8352
8232
8199
7847
7876
7743
7645
7175
6594
5924
119.2
118.0
115.9
116.6
127.3
165,1
223.0
263.2
282.6
282.6
277.0
265.9
263.8
259.4
262.4
272.2
281.3
293,2
295.9
299.4
308.9
317.5
334.7
353.1
370.9
379.6
374.2
372.7
356.7
358.0
352.0
347.5
326.1
299.7
269.3
End August, 1919.........................
End September, 1919...................
End October, 1919........................
End November, 1919...................
End December, 1919.....................
End January, 1920........................
End February, 1920.....................
End March, 1920...........................
End April, 1920.............................
End May, 1920...............................
End June, 1920..............................
End July, 1920...............................
End August, 1920.........................
End September, 1920...................
End October, 1920........................
End November, 1920...................
End Deeember, 1920........... .........
686
i
866
828 :
818
8441
9121
931
937
1033*
1040“
1047
1064
1093
1145
12111
12531
1246
12321
12951
1289
1308
1302*
1311
13161
12591
1216
1112
13481
13911
13891
1337
1335
13101
1294
1290|
1327
13711
1417
1383
1344
1358
1396
14531
15351
16141
17091
16541
1619
15551
1541
15541
15391
1446
1336
1275
How each group has moved from the beginning of the war to the
end of 1919 may be seen from Table 64, in which the changes in
each half year are compared with July, 1914, as 10 0 :
T able
6 4 .— ECONOMIST IN D E X NU M BER S FOR DECEM BER A N D JUNE EACH Y E A R
1914 TO 1919.1
(Average prices in July, 1914=100.)
End of—
July, 1914....................................
December, 1914.........................
June, 1915.....................................
December, 1915...........................
June, 1916.....................................
December, 1916...........................
June, 1917.....................................
December, 1917...........................
June, 1918.....................................
December, 1918...........................
June, 1919.....................................
December, 1919...........................
Cereals
and
meat.
Other
food
products.
100
124
141
155
172
224
248
222
220
226
231
249
1 F rom the E conom ist, Jan. 10, 1920, p. 54.
100
117
122
126
143
157
185
194
220
222
227
250
Textiles.
100
82
90
119
129
183
235
274
294
293
284
396
Minerals.
*
100
103
135
154
193
178
182
181
186
186
202
247
Miscella
neous.
100
124
141
153
183
Total.
201
100
109
127
142
164
192
230
244
250
241
247
263
228
238
236
241
287
220
GREAT BRITAIN---- AUGUSTUS SAUERBECK
(S T A T IS T ).
275
INDEX NUMBERS OF AUGUSTUS SAUERBECK (STATIST).
PUBLICATION.
This index number represents the course of wholesale prices in the
United Kingdom. Prior to 1910 statements were published only once
a year. Prom January, 1910, to January, 1913, the general result
was published each month for the preceding month, and the yearly
resume in March, in the Journal of the Royal Statistical Society,
London, but since January, 1913, the information has appeared in
The Statist, London. A general discussion of the results for each
year is now also published in the Journal of the Royal Statistical
Society.
HISTORY-
During 1885, or in the early part of 1886, Mr. Augustus Sauerbeck,
a London wool merchant, prepared a paper upon the gold supply
and its relation to prices, which was published in the September,
1886, Journal of the Royal Statistical Society." In much of the
discussion relating to the causes of an “ extraordinary and almost
unprecedented fall of prices ’1 that had continued for 12 years, Mr.
Sauerbeck recognized the lack of statistical information and prepared
this paper in order to supply data upon the subject.
The work thus begun was continued by Mr. Sauerbeck until the
end of 1912, when he relinquished the task and it was taken up by
Sir George Paish, editor of The Statist.
SOURCE OF QUOTATIONS.
The statement is made in the appendix to the first article that
such of the prices from 1846 to 1885 as are not official returns were
received from private firms or collected from the Economist and
other publications. Further than this no information concerning the
source o f price quotations is given.
BASE PERIOIX
The 1 1 years 1867-1877 are taken as the standard period. A t the
time the period was choosen the study covered the 40 years 1846-1885
and the base period included the years of the highest prices as well
as a number o f years of low prices. The index number for the 1 1 -year
period was found to correspond exactly with the index number of
the 25 years 1853-1877, so that “ a comparison of the aggregate prices
of all commodities in a certain year with the 11 years 1867-1877, i&
equivalent to a comparison with the whole 25 years 1853-1877.” 1
PRICES: H O W SH O W N AND COMPUTED^
The prices upon which the index number is based are average
prices for each year. The prices quoted in the report covering the
years 1846-1885 are, with but few exceptions, “ the average prices
in each year, either those officially returned or the averages of the
12 quotations at the end of each month .” 2 Where a range of prices
is given the mean is taken between the highest and the lowest quo
tations. The prices as given in later reports are the averages of
99 Journal
1 Idem, p.
2
or thje Royal SlatisfcieaJ Soeiotyv September, 1886, \ol. 49, p. 581.
592.
idem, p. 632.
276
INDEX NUM BERS OF WHOLESALE PRICES.
12 monthly or 52 weekly quotations; in the case of potatoes, of 8
monthly quotations, January to April and September to December.
These annual averages are shown in the tables by articles, as are also
the corresponding relatives. The actual prices from which the
yearly averages are computed are nowhere shown, but relatives based
on the quarterly averages are shown by groups of commodities, cov
ering the period fron 1884 to the present time.
The statement is made in the report of 1893 that all articles have
been calculated at their actual prices and no corrections have been
made for extraordinary fluctuations. The treatment of cotton prices
at the time of the American Civil War is cited as an example.
The prices of all imported articles are quoted “ in bond.”
In the first report the general statement is made that in construct
ing his table of prices the author has on the whole been guided by the
system adopted in the Economist reports on the course of prices.
NUMBER AND CLASS OF COMMODITIES.
The number of articles used directly in computing the index num
ber was 43 from 1846 to 1866, 44 from 1867 to 1872, and 45 from
1873 to the present time. All are considered raw materials. In the
original report the statement is made that only commodities were
included which in the United Kingdom at that time represented a
value of about a million pounds or more, counting both domestic pro
duction and imports. A few important articles, like wine, spirits, and
tobacco, had to be left out, as no reliable data were obtainable.
Certain important commodities are represented by more than one
of the 45 articles; for example, two varieties of wheat are quoted,
and each variety is considered a separate article. The relative prices
of certain others of the 45 articles, as for example coffee, were obtained
b y averaging two relatives representing different varieties or grades
of the article. Thus in 1911, when the relative price of Ceylon coffee
was 95 and of good Rio was 91, the relative used for coffee was 93,
the average of these two. This method was followed in cases where
the price of a single variety was not considered sufficiently repre
sentative of the article. The number of quotations, including these
additional quotations used only indirectly in the computation of the
index number in the report for 1911, was 57. The table of average
actual prices, however, comprised 60 quotations, one each for tea,
copper, and coal being shown in the actual price form without being
represented in the index number. At the time the original report
was published the series of quotations in the table of average prices
comprised a total of 55.
An index number based on the prices of “ the 31 principal com
modities” from 1818 to 1845 was prepared by Mr. Sauerbeck and
published in his original report. These commodities are not enu
merated.
DESCRIPTION AND GROUPING OF COMMODITIES.
The 45 articles are divided into six groups. The grouping is shown
in Table 65, which also shows the number of series of price quotations
secured for each commodity, and the number of relative prices for
each commodity used directly in the computation of the index for the
year 1911. The table has been compiled from data appearing in
GREAT BRITAIN---- AUGUSTUS SAUERBECK
(S T A T IS T ).
277
the Journal of the Royal Statistical Society, March, 1911, pages
415 to 420.
T a b l e 6 5 . — NU M BER
OF SERIES OF PRICE QUOTATIONS SECURED AN D NU M BER OF
R E L A T IV E PRICES USED IN IN D E X , B Y COMMODITIES.
Commoditv,
Number Number
of series of relative
of quota
prices
used in
tions
secured.
index.
Commodity.
Number Number
of series of relative
of quota
prices
used in
tions
secured.
index.
Iron..............................................
Copper.........................................
T in..............................................
Lead............................................
Coal..............................................
W heat. . .
Flour___
Barley. . .
Oats........
Maize----Potatoes.
Rice........
Total minerals...............
Cotton.........................................
Flax............................................
Hemp..........................................
Jute..............................................
Wool............................................
Silk..............................................
Total vegetable food.
Beef.......
Mutton.
Pork___
Bacon...
Butter. .
Total textiles..................
Total animal food.
Sugar.
Coffee.
T e a ...
Total sugar, coffee, and
Total food..
19
Hides..........................................
Leather.......................................
Tallow.........................................
Oil................................................
Linseed oil and linseed (flax
seed).........................................
Petroleum..................................
Soda crystals.............................
Nitrate of soda..........................
Indigo.........................................
Timber.......................................
Total sundry materials.
16
11
Grand total....................
60
45
A description of the various articles included in the six groups of
commodities follows :3
Vegetable food (8 price series).
1.
2.
3.
4.
5.
6.
7.
8.
W h ea t, English Gazette.
W heat, Am erican.
Flour, town-m ade white.
Barley, English Gazette.
Oats, English Gazette.
M aize, Am erican, m ixed.
Potatoes, good English.
R ice, Rangoon, cargoes to arrive.
Animal food (7 price series).
9.
10.
11.
12.
Beef, prime.
B eef, m iddling.
M utton, prime.
Mutton, m iddling.
13. Pork, large and sm all, average.
14. Bacon, W aterford.
15. Butter, Friesland, fine to finest.
Sugar, coffee, and tea (8 price series).
16a.
16b.
17.
18a.
Sugar, British W est Indian, refining,
Sugar, beet, German, 88 p. c. f. o. b.
Sugar, Java, floating cargoes.
Coffee, Ceylon plantation, low m id
dling.
18b.
19a.
19b.
19c.
Coffee, R io, good.
T ea, Congou, common.
T ea, average import price.
T e a , In dian , good m edium .
3 Journal of the R oyal Statistical Society, March, 1918, p p . 344, 345.
278
INDEX NUM BERS OF WHOLESALE PRICES.
Minerals (10 price scrim).
2 0a .
20b .
21.
22.
—
Iron, Scotch pig.
I 23. T in , Straits.
Iron, Cleveland (Middlesbrough) pig.
24. L e ad , English pig.
Iron, barsTcommon.
25a. Coal, Wallsend* I le lto n , in London.
Copper, Chile bars.
25b. Coal, Newcastle steam.
Copper, English tough cake.
26. Coal, average export price.
Textiles (11 price series).
27.
2&.
29a.
29b.
30a.
30b.
31.
Cotton, m iddling Am erican.
Cotton, fair Dholera.
F lax, Petrqgrad.
F lax, Russian average import.
H em p , M anila fair ropiftg.
H e m p , Petrograd, clean.
Jtite, good m edium .
32a. W ool, merino, Port P hillip, average
fleece.
32b. W ool, merino, Adelaide, average
grease.
33. W ool, English, Lincoln half hogs.
34. Silk, Tsatlee.
Sundry materials (16 price series).
35a.
35b.
35c.
36a.
36b.
37.
38.
39.
40a.
H ides, R iver Plata, dry.
Hides, R iver Plata, salted.
H ides, average im port.
Leather, dressing hide#.
Leather, average import.
Tallow , town.
Oil, palm .
Oil, olive.
Oil, linseed.
40b .
41.
42.
43.
44.
45a.
45b.
Seeds, linseed.
Petroleum , refined.Soda, crystals.
Nitrate of soda.
Indigo, Bengal, good conswiBfting.
Tim ber, hewn, average im port.
T im ber, sawn or split, average
im port.
SUBSTITUTIONS AND ADDITIONS.
The method of calculating the index adopted when it was deemed
necessary to add or drop quotations for articles is not disclosed. No
mention is made of the necessity of quoting other grades of com
modities than those formerly quoted, but it is reasonable to believe
that in a period of this length it has been found necessary to do so.
INTERPOLATION.
It may have been impossible to secure complete statistical material
during the full period, but, if so, the author makes no mention of the
fact. Where prices were abnormal, as cotton during the Civil War in
the United States, no corrections were made, quotations being used as
found.
WEIGHTING.
The index number is unweighted. The author has, however, given
to certain important commodities a larger influence in computing the
index number by quoting as separate articles several different varieties
or grades of the same commodity. For example, English wheat and
American wheat constitute two separate articles, as do prime beef
and middling beef; also prime mutton and middling mutton. Simi
larly, sugar, iron, coal, cotton, wool, and oil are each given double
importance in the computation of the index.
TESTING.
Beginning with the index numbers of the year 1892 4 two tests
were applied, the one consisting of weighting the relative prices
according to the “ money-valnes” of the commodities in accordance
4 Journal of the R oyal Statistical Society, June, 1893, pp. 215-247, 254.
GREAT BRITAIN— AUGUSTUS SAUERBECK
(S T A T IS T ).
279
with their importance in the trade of the United Kingdom during the
3-year period 1889-1891; the second method consisting of weighting
them according to their “ mass-quantities” of other years. In the
latter method the quantities of imports and exports of any one year
are reduced to a nominal money value by multiplying the number
representing the quantity of the article by the number representing
the average prices of said articles during the years 1867-1877.
In his presentation for the year 1895 5 Mr. Sauerbeck used the
geometric average 6 of Jevons and calculated by that means a total
index for his 45 articles for the years 1880, 1894, and 1895, and com
pared it with his own arithmetic averages, both simple and weighted.
After 1907 the single test of weighting according to average
11money-values ’ 1 of the commodities for the 3-vear period 1904-1906
was employed.
The author makes no direct statement in regard to the process of
finding the nominal values of the several articles, beyond stating that
one factor is the average price of the article during the base period
1867-1877. The other factor, or that quantity which represents
the “ importance in the United Kingdom” of the article, appears
to be the average of the annual production plus imports for the
chosen 3-year period. Due warning is given that this figure, which
represents the total trade in the commodity including reexports,
must not be considered as representing the actual consumption of
the commodity in the United Kingdom.
A second test was in use up to and including the report for 1907.
It is stated that the estimated actual values of the 45 articles con
sumed in the United Kingdom were obtained by taking the produc
tion on the basis of Mr. Sauerbeck’s price and the imports at British
Board of Trade values. The ratio of these actual prices to the
nominal values on the basis of the average prices from 1867 to 1877
constituted a second weighted index. The explanation of this
second test is not fully stated.
TABLE OF RESULTS.
The principal table in this compilation shows the index number
for the four general groups, and the grand total index. In addition
there are three subindexes of food and an index of all materials
contained in the groups of minerals, textiles, and sundry materials.
The following data, showing the variations in the group index
numbers and in the general index, have been compiled from various
issues of the Journal of the Royal Statistical Society .7
5 Journal of the Royal Statistical Society, March, 1896, pp. 193, 194.
6 To find the geometrical moan, the logarithm is taken of the percentage figure of each article, the total
of all logarithms is divided by 45—the number of articles—and the antilogarithm, the number correspond
ing to the average logarithm, is the geometrical index number. (Journal of the Royal Statistical Society
March, 1896, p. 194.)
7 September, 1886, p. 648; March, 1891, p. 128; March, 1911, p. 408; April, 1914, p. 556; March, 1917, p. 289;
March, 1921, p. 255.
INDEX NUM BERS OF WHOLESALE PRICES.
6 6
.—SUM M AR Y OF IN D E X NUM BERS, 1846 TO 1920.
(Average prices in 1867-1877=100.)
Food.
Vege
table
food
(corn,
etc.).
106
129
92
79
74
73
80
100
120
120
109
105
87
85
99
102
98
87
79
84
95
115
113
91
88
94
Ani
mal
food
(meat,
etc.).
Sugar,
coffee,
and
tea.
81
98
87
69
77
87
84
75
87
85
89
97
119
97
88
83
71
67
68
69
82
87
87
88
89
83
85
91
91
86
85
89
97
96
89
88
96
98
101
100
101
106
105
93 .
92
109
103
108
108
100
101
101
95
87
89
84
84
82
71
68
65
64
67
65
65
75
65
59
55
54
53
60
67
60
62
62
63
62
63
63
62
69
70
71
65
70
78
69
75
108
133
177
168
Materials.
94
101
101
104
103
97
88
87
79
82
86“
82
81
84
85
80
78
73
79
77
79
85
85
87
84
83
87
89
88
89
89
96
90
96
99
100
126
152
192
207
Total
food.
95
105
84
76
75
74
75
91
101
101
99
Miner
als.
92
94
78
77
77
75
80
105
115
109
110
100
94
93
98
104
106
105
107
104
108
96
98
97
91
91
93
96
91
91
87
85
89
89
93
127
141
116
100
100
101
98
103
90
87
99
102
107
96
98
99
106
97
94
94
96
98
95
88
84
76
77
63
63
60
67
65
75
70
71
69
75
65
62
59
52
51
53
54
46
41
44
50
52
46
48
48
50
54
61
62
54
58
70
86
113
130
102
88
89
98
97
94
89
88
91
95
101
100
102
101
96
90
94
91
89
89
79
74
72
70
72
75
73
77
73
72
66
64
62
65
68
65
69
67
67
66
68
69
69
72
72
73
74
75
81
77
81
107
130
169
174
90
84
74
73
79
77
79
76
68
66
67
69
78
75
80
76
71
68
64
62
63
66
70
92
108
89
82
82
81
87
101
107
89
86
89
93
110
111
99
126
158
172
192
Tex
tiles.
77
78
64
67
78
75
78
87
88
84
89
92
81
Sundry
mate
rials.
86
86
77
75
80
79
84
101
109
109
109
119
Total
mate
rials.
85
86
73
73
78
76
81
97
104
101
102
88
107
90
92
123
149
162
134
130
107
94
98
111
100
109
106
99
107
115
119
108
107
110
100
102
100
100
99
105
108
106
96
92
95
94
99
106
109
106
103
114
103
92
88
85
85
78
74
81
77
73
70
68
65
63
65
64
70
66
59
57
59
53
52
54
51
51
58
66
60
61
66
102
101
98
97
99
88
85
89
86
85
84
81
76
69
67
67
68
69
69
67
68
64
65
63
62
63
65
71
71
71
69
67
71
72
80
77
62
64
73
76
76
84
81
92
129
192
74
78
73
76
81
81
82
83
87
109
136
174
222
202
68
99
100
101
115
114
100
93
91
89
81
78
84
80
80
77
73
70
67
67
69
70
71
68
65
65
60
60
60
59
61
70
80
72
71
72
72
75
83
86
74
75
81
83
88
91
88
108
140
179
206
>1.
89
95
78
74
77
75
78
95
102
101
101
105
91
94
99
98
101
103
105
101
102
100
99
98
96
100
111
102
109
96
95
94
87
83
88
85
84
82
76
72
69
68
70
72
72
72
68
68
63
62
61
62
64
68
75
70
69
69
70
72
77
80
73
74
78
80
85
85
85
108
136
175
192
GREAT BRITAIN— AUGUSTUS SAUERBECK
T a b l e 6 6 . —SU M M AR Y
OF IN D E X N U M BER S, 1846 TO 1920—Concludad.
Food.
Date.
1919............................
1920............................
Average:
1904-1913...
1890-1899...
1878-1887...
1818-1827...
Vege
table
food
(corn,
etc.).
179
227
68
61
79
109
Ani
mal
food
(meat,
etc.).
281
(S T A T IS T ).
Materials.
Sugar,
conee,
and
tea.
Total
food.
213
263
147
198
91
80
95
90
53
63
76
151
185
234
73
68
81
111
Miner
als.
Tex
tiles.
Sundry
mate
rials.
Total
mate
rials.
222
295
228
262
219
244
95
71
73
128
74
56
71
105
76
220
66
81
106
264
81
64
76
112
Grand
total.
206
251
77
66
79
111
INDIA.
INDEX NUMBERS OF THE DEPARTMENT OF STATISTICS.
PUBLICATION AND HISTORY.
Index numbers of rupee prices in India for the years 1861-1895
were first published by Fred. J. Atkinson, accountant general,
United Provinces, India, in the Journal of the Royal Statistical
Society for March, 1897. Those for the years 1896 and 1897 were
presented in the journal for June, 1898, and those for the years 1898
to 1901 in the number for March, 1903. In the number for Sep
tember, 1909, they were brought up to the end of 1908.
Atkinson’s work was continued by the Commercial Intelligence
Department of India, and, in addition, a series of index numbers of
prices in India for the years 1861 to 1904 was constructed by Mr.
J. A. Robertson, director general of statistics, and published in
August, 1905. The series and the charts were republished in 1910,
with figures brought up to 1909, and again in 1911, 1912, and 1913,
with the figures brought up to date. These publications appeared
under the title of u Variations in Indian Price Levels.” In 1919 the
Department of Statistics of India, which had superseded the Com
mercial Intelligence Department, published the. sixth volume of
the new series, with a few additional tables and with figures brought
up to 1918, under the title of “ Index Numbers of Indian Prices.77
SOURCE OF QUOTATIONS.
In preparing his index numbers Atkinson, while adopting Sauer
beck’s principles, did not base his figures on the prices of imported
articles, but on those of the native products of India. This was
necessary, because the bulk of the trade of India is concerned with
the products of the country, imports representing only some 8 or 9
per cent of the exported products. Moreover, a considerable portion
of the articles produced are mainly for the purpose of export and are
but little used by the natives of the country. It is evident, therefore,
that prices must necessarily be affected rather by the production in
common use than either by imports or exports.
Sauerbeck takes the majority of the prices he quotes from the
London market. In India, however, the production and prices vary
so greatly in different parts of the country that to take only one
market, as Calcutta or Bombay, and treat every article as of equal
or nearly equal importance would give a very inaccurate idea of the
282
INDEX NUM BERS OF WHOLESALE PRICES.
actual state of affairs. To avoid this, Atkinson prepared a statement
based on the agricultural returns and financial and commercial sta
tistics published by the Government of India, and the administration
reports of the various native States, giving for the Year 1893, with a
fair degree of accuracy, the agricultural and manufactured products
of India and the relative importance of each .8
Next, accepting the fact that India in its economic conditions
represents a cluster of different countries, it had to be ascertained in
what particular markets the prices of the various articles should be
taken. Atkinson accordingly prepared a table showing the area of
cultivation in 1893-94 of each product in each Province of India,
and the price of each product was, as far as possible, taken in the
province or provinces in which the area of its cultivation is greatest.
The prices taken were obtained from various sources; partly from
the prices current which the chambers of commerce of Calcutta,
Bombay, and Madras issue; partly from the publication “ Prices
and Wages in India,” issued by the Government of India; and
partly from private sources. A few only, when figures were not
elsewhere obtainable, were taken from the export accounts.
In later publications by the Department of Statistics the same
sources of price quotations were utilized.
BASE PERIOD.
In the first table of index numbers prepared in 1897 Atkinson took
the year 1871 to represent the number 100 . This year was selected
because in it Sauerbeck’s index number of gold prices was 100 , and
the price of silver was approximately the same (99.7). This selection
had the advantage of permitting a ready comparison of the course of
rupee prices with that of gold prices, which was useful in connection
with the currency question. It is obvious, however, that prices of a
single year can not be regarded as representing normal prices. As a
matter of fact, rupee prices in 1871 were exceptionally low and as
the currency question had been settled Atkinson, in 1903, thought
it best to use the average of the years 1868-1876, which may be re
garded as fairly normal years, as the base period for another table
of index numbers. Both tables, that with 1871 and that with 18681876 as the base period, were in 1908 brought up. to date. In 1910
the base year was changed to 1873 in reports of the Commercial Intel
ligence Department. This year was considered to have been of nor
mal character, and it was the year when silver began to fall conse
quent on the general demonetization of silver by Germany and other
countries.
PRICES: H O W SH OW N AND COMPUTED.
The yearly volume on “ Prices and Wages in India” which has been
published by the Government since 1878 contains numerous tables
on wholesale prices in Calcutta, the prices of certain staple articles
at Calcutta and Bombay, and prices of articles of export at Calcutta,
Bombay, Madras, and Rangoon. Each table of actual prices is fol
lowed by one showing relative prices, or index numbers, of individual
commodities from a given basic period, A separate table gives the
actual prices, through a number of years, of 25 articles among the
s See Journal of the R oyal Statistical Society, March, 1897, V ol. L X , p p. 124, 125.
INDIA— DEPARTMENT OP STATISTICS.
283
grains used for food, together with a comparison of prices in India
with those in the United Kingdom. Yearly relative prices are also
shown for articles of consumption by troops at 16 stations in India.
Other tables in the appendix give average yearly prices of rice,
wheat, cotton, and jute.
NUMBER AND CLASS OF COM M ODITIES.
In the volume entitled “ Index Numbers of Indian Prices, 18611918” (p. 22 ), published by the Department of Statistics in 1919,
the number of articles included in Atkinson’s index number is given
as 37. In this count, however, several related articles are in som©
instances counted as one article, due allowance being made in the
weighted index number. A straight count of the articles or kinds
of articles, arbitrarily made, gives the number as 60. Both raw and
manufactured commodities are included in the list, the raw commodi
ties predominating. The index number of the Department of Sta
tistics, as shown by page 19 of the report referred to above, includes
39 articles, of which 28 represent exported articles and 11 imported
articles. A majority of these consists of rawTmaterials.
DESCRIPTION AND GROUPING OF COMMODITIES.
The individual commodities included in the general index numbers
of both systems and the markets in which the prices were obtained
are given separately. The Atkinson commodities are the following.
The numerals at the beginning of each paragraph refer to the order
of arrangement of the articles in the total of 100 price series.
1-30. Rice (15 prices).®— Monghyr, Calcutta; common, Bengal, eastern d ivision; com
mon, Calcutta; common, Patna; common, Bengal, Deltaic division; common, Bengal,
Orissa division; common, Madras, southeast coast division; common, Madras, Salem ;
common, northwest provinces, eastern division; common, northwest provinces, central
division; common, Burma, Rangoon; common, Burma, Tenasserim division; common,
central provinces, Nagpur; common, central provinces, Jubbulpore; and common,
Hvderabad, Bolaram. The quotations used are those published in “ Prices and
W ages” and from prices supplied b y dealers.
S1-S5. Wheat (5 prices).— Common, northwest provinces, Cawnpore; comtaon, P un
jab, D elhi; common, central provinces, Nagpur; first quality, Central India, Nussirabad; and flour, B om bay, Poona.
36-38. Jawari10 (3 prices).— B om bay; Madras, Salem ; and H yderabad, Bolaram.
39-40 . Rag gee 11 (2 prices).— Madras, southeast coast division; and Mysore.
41-42. Gram 12 (2 prices
Punjab, central division; and northwest provinces, cen
tral division.
43-44. B ajra 13 (2 prices).— Bom bay, Deccan division; and Madras, Salem.
45. Maize (1 price). Chota, Magpur.
46. Barley (1 price). Delhi.
4^-51. Other grains (5 prices).—-Arhar, Allahabad; arhar, northwest provinces,
Sub-Montano division; mung, Dal, Lucknow ; masur, L uck now ; and clienna, Cawn
pore. This classification in the agricultural returns includes a large number of differ
ent varieties of minor grains and pulses grown in different parts of the country, though
the northwest provinces are preeminent in their cultivation. Reliable figures could
be obtained for only a few, and the figures for some of these are not complete. Arhar
(Cajanus indicrus), the most important of these minor pulses, for which two prices
have been given, and mung (Phaseolus mirago) are represented b y complete figures.
The prices for mastir (Erva lens) and ehenna (Cicer arietinum ), were supplied b y the
).—
—
—
9 As rice is given an importance of 30 and only 15 prices were taken, each price has been doubled in can*puting the general index number.
10 A cheap Indian grain used in making, a kind of unleavened bread.
11 A cereal grass (eleusina Caroeana) largely cultivated for food.
12 The chick-pea (Cicer arietinum) of East Indies, there extensively used as food for men, horses, and
cattle.
u The spiked or pearl millet (Penicillaria spicata) one of the commonest food cereals of southeastern Asia.
284
INDEX NUM BEBS OF WHOLESALE PRICES.
commissariat department and commence only from 1875 and are averages of the finan
cial year.
52. Vegetables ( 1 price).— Potatoes, Bom bay.
53-56. Sugar (4 prices).— Gurputty, Calcutta; Dhulloah, Calcutta; Jaggery, cane,
Madras; and Jaggery, Palmyra, Madras. Prices for the two refined sugars have been
taken from the Calcutta prices current, supplem ented b y prices supplied b y B issouath, Law & C o ., and those for raw sugar from the Madras prices current. Consid
erable difficulty was experienced in the case of refined sugar, as indeed in most of the
quotations taken from the prices current, b y the changes in nom enclature, w hich in
Bome cases m eant a change in actual quality. To continue the same quality through
out the entire period involved a method of calculation of comparisons. The result,
however, is said to be approxim ately accurate.14
57. Spices ( 1 price).— Ginger, export accounts. Prices taken from the export
accounts.
58. Ghee 15 (1 price).— Bom bay.
59-60. Meat (2 prices).— M utton, B o m bay ; beef, Bom bay.
Raw produce and materials.
61. Tea ( 1 price).— Taken from the export Accounts.
62. Coffee (1 price).— T aken from the export accounts.
63-65. Cotton (3 prices).— Dharwar; Broach; and Dholera.
A ll obtained from the
B o m bay prices current.
66-67 . Jute (2 prices).— P ick ed ; and double triangle M . From Calcutta prices
current.
68-69. Indigo (2 prices).— Bengal, good; and consuming. From Calcutta prices
current.
70-71. Opium (2 prices).— Behar; and M alwa. From the m o n th ly figures pub
lished b y the G overnm ent of India.
72-73. Tobacco (2 prices).— Central India, Nussirabad; and B o m bay .
74-77. Seeds (4 prices).— Linseed, bold ; til; rape-yellow, m ix ed ; and castor.
The prices for linseed and rape were taken from the Calcutta prices current, those of
til and castor seeds from the m onth ly figures published b y the Governm ent of India.
78-84• Miscellaneous (7 prices).— Saltpeter, 5 per cent refined; c u tch ,16 Rangoon;
myrobalans;17 manure, anim al bones; coal; raw silk , Surdales; and raw wool. The
prices of saltpeter and raw silk were taken from the Calcutta prices current, those
of cutch, myrobalans, and manure from the export accounts, and those of coal were
furnished b y the Bengal Coal Co.
85-87. Hides and skins (3 prices).— R aw hides, buffalo, Patna, slaughtered, arsenic;
raw hides, cow, Burdwan, slaughtered; and raw skins, goat, Calcutta. Prices taken
from the Calcutta prices current.
88-89. Timber (2 prices).— Bamboos, Calcutta; teak, Rangoon. T he prices for
bamboos are taken from the figures published in the Governm ent of India publication
“ Prices and W a g e s . T h e prices on 1st of January of each year beginning from 1871
only are given. M onthly prices for the series of years were not obtainable. The prices
for Rangoon teak were taken from the Calcutta prices current.
Manufactures.
90-91. Cotton goods (2 prices).— Y a m 1/203, and T cloth, 44 inches.
These are
taken from “ Prices and wages” and represent the prices as given to the Governm ent
of India b y the B om bay M ill O w ners’ Association on 1st of January and July of each
year.
92-93. Jute goods (2 prices).— Bags, N o. 2 tw ill; and bags for California. Prices
taken from the Calcutta prices current.
94-95. Oils (2 prices) .— Castor and coconut. The prices for castor oil were taken
from the Calcutta prices current. Those for coconut oil are from “ Prices and wages’ ’
the prices on January 1 of each year beginning from 1871 only are given, m onthly
prices not being obtainable.
96. Silk piece goods ( 1 price).— Corah No. 1, from the Calcutta prices current.
97-99. Hides and skins ( 3 prices). Tanned hides; cow ; tanned skins, goat; and
tanned skins, sheep. Prices are taken from the Madras prices current.
100. Shellac (1 price).— First quality, orange; from the Calcutta prices current.
14 Journal of the Royal Statistical Society, March, 1897, Vol. L X , p. 90.
Butter clarified by boiling or heating and skimming or straining until it becomes a liquid or semiliquid
oil, capable of being kept for many years. It enters into the composition of nearly everything eaten by
the Brahmans.
ifl An extract from the bark of the mangrove used in tanning and dyeing.
17 Prune-like fruits of several tropical plants of the genus terminalia, used for tanning and calico printing.
285
INDIA— DEPARTMENT OF STATISTICS.
A list of the commodities included in the index number of the
Department of Statistics of India follows :18
Exported articles.— R ice— average of M oonghy and B allam ; wheat; jawari; bajra;
gram; barley; ragee; sugar; tea— average of Pekoe Souchong and Congou; ghee;
linseed; rapeseed; sesamum (til or jin jili seed); poppy seed; castor oil; jute, raw— ■
average of picked and ordinary; jute— gunny bags; cotton, raw; cotton yarn; T
cloth; silk, raw; wool, raw; hides, raw; skins, dressed— average of goat and sheep;
coal; shellac— average; saltpeter; indigo.
Imported articles.— Sugar— M auritius; salt; gray shirtings; gray yarn; colored yarn;
silk, raw; iron; copper; spelter; coal; kerosene oil.
Retail prices were used in the case of* certain, cereals—jawari,
bajra, barley, ragee, and gram—wholesale prices of these articles
not being available before 1897, and the difference between wholesale
and retail prices being very small.
WEIGHTING.
Briefly, the procedure adopted in preparing the Atkinson series of
index numbers was to ascertain the relative importance of each
article as compared with the total value of all production in India,
as shown by the table of production for the year 1893. In computing
the index numbers for each month and year for the whole of India,
one or more prices at different important places of production were
taken for each article in proportion to its relative importance to
the whole. Rice, representing three-tenths of the value of all
products of India, was allotted 30 prices out of a total of 100 taken.
Similarly wheat was allotted 5 prices and other grains 5, sugar 4,
jawari 3, ragee, gram, bajra, and meat 2 each, and maize, barley,
potatoes, spices, and ghee 1 each; making a total of 60 prices for food
articles. For raw produce, seeds were allotted 4 prices, cotton 3 ,
hides and skins 3, jute, indigo, opium, tobacco, and timber 2 each,
and tea, coffee, saltpeter, cutch, myrobalans, animal bones, coal,
raw silk, and raw wool 1 each, making 29 in all. For manufactures,
hides and skins were allotted 3 prices, cotton goods, jute goods,
and oils 2 each, and silk piece goods and shellac 1 each, making 11 in
all. Summarized, the division was:
Prices.
A rticles of food.............................................................................................................
Raw produce.................................................................................................................
Manufactures................................................................................................................
60
29
11
T otal.................................................................................................................... 100
No weights are used in constructing the index numbers of exported
and imported articles published by the Department of Statistics.
TABLES OF RESULTS.
The Department of Statistics has published a table on page 1 of
its report, “ Index Numbers of Indian Prices, 1861-1918,” containing
the unweighted index numbers of 28 articles of export, 11 articles of
import, in all 39 articles, and a continuation of the Atkinson series
of weighted index numbers to 1918.
18 See Index numbers of Indian Prices, 1861-1918, p. 19.
286
IKDEX IfUM BEES OF WHOLESALE PRICES.
T a r le 6 7 __ I N D E X NU.M BERS OF P R IC E S IN IN D IA , 1881 T O 1918. i
(A verage prices in 1875= 100.}
Year.
1861..
1862..
1863..
1864..
1865..
1866..
1867-.
1868..
1868..
1870..
1871..
1872..
1873..
1874..
1875..
1876..
1877..
1878-..
1879..
1880=..
1881..
1882..
1883..
1884..
1885..
1886..
1887..
1890..
1891..
1892..
1893..
1894..
1895..
1896..
1897..
1898..
1899..
1900..
1901..
1902..
1903..
1904..
1905..
1906..
1907..
1908..
1909..
1910..
1911..
1912..
1913..
1914..
1915..
1916..
1917..
1918..
Exported
artieles (28—
unweighted).
88
88
93
103
100
no
102
94
im
105
95
101
100
102
95
90
110
114
112
lift
99
95
93
96
91
93
94
98
104
104
103
109
112
110
111
117
124
102
100
124
116
113
103
104
116
139
145
151
133
127
136
145
154
160
155
163
170
199
Imported
articles ( 1 1 —
unweighted).
95
95
113
132
125
126
124
107
97
95
88
91
100
99
90
91
88
84
83
85
79
78
75
80
83
92
91
91
84
84
89
84
87
94
86
8087
96
86
96
88
93
96
105
116
106
99
109
113
117
117
114
146
236
262
289-
Weighted in
General index
dex number
number for all
( 1 0 0 price
articles (39— series)
equated
unweighted). to 100 for 1873.
90
90
98
111
107
115
108
98
105
102
93
98
100
101
94
90
104
106
104
104
96
92
m
91
8?
89
91
96
101
100
98
102
105
102
104
110
113
96
96
116
110
106
99
101
110
129
137
138
124
122
129
137
143
147
152
184
196
225
93
93
97
105
109
124
118
107
118
107
93
98
100
108
96
100
129
138
126
109
99
98
99
107
106
103
104
111
117
117
120
132
129
122
128
131
153
125
121
143
139
128
122
121
135
158
167
179
160
150
155
174
182
187
182
185
186
215
1 These index numbers are for wholesale prices except in the case of the cereals jawari, bajra, barley,
ragged and gram, wholesale prices of these articles not being available before 1897.
In the table printed in the Journal of the Royal Statistical Society
for September, 1909 (pp. 500-502), Atkinson gives, in addition to
the index numbers o f 100 articles of Indian production, index num
bers for II articles of import taken from the data published by the
Department of Statistics. As imported articles are regulated by their
gold price and as the individual articles are not weighted according to
their importance, it is not surprising that the two sets of index num
bers do not agree, though their general trend is practically the same.
It having been suggested, however, that so far as the dweller in India
is concerned the prices of imports now materially affect his annual
expenditure, Atkinson shows in column 6 of his table the index
INDIA---- DEPARTMENT OF STATISTICS.
287
number of 11 articles of import equated to 1868-1876, as given by
the Commercial Intelligence Department, and then adds the index
numbers of these 11 articles of import to those of the 100 articles of
Indian production to which his own index numbers relate, and shows
the total index number in column 7 of his table. In column 8 are
shown Sauerbeck’s index numbers for gold prices; in column 9 the
gold price of silver; in column 10 the gold price of the rupee; in col
umn 11 index numbers of articles of export; in column 12 Sauerbeck’s
gold prices index number of the 11 articles of import shown in column
6 ; and in column 13 Sauerbeck’s gold prices index number of 11
articles exported by India.
This table is reproduced below with the exception of the data con
tained in columns 12 and 13.
T able
6 8
.—IN D E X N UM BERS OF PRICES IN IN D IA , 1870 TO 1908.
[Journal of the Royal Statistical Society, September, 1909, Vol. L X X I I , pp. 500-502.]
Index lumbers (percent*strg.es) of
rupee prices in India (a^ erage of
1868-1 §76=100).
Year.
1
Index
, Index
Sauer
num
Index
Gold
num
beck's
Gold
bers,
num
price
ber of
gold
price
bers
of
sil
ver
1 1 1 ar
articles
prices (60.84d.
of
of 1 1
ticles,
©#
rupee ;
AH
Manu
articles includ (average,
per
; export
Food I Raw
1867(23.34d.
ing 1 1 ar
produce factures prices
of
ounce
, (1868(60
1877
=
1 0 0 ).
(29
(1 0 0
= 1 0 0 ).
import. ticles of = 1 0 0 ).
1876
(U
articles). articles).
articles). articles).
= 1 0 0 ).
import.
4
3
2
5
7
6
9
8
1 0
1 1
1
1870...........
1871...........
1872...........
1873...........
1874...........
1875...........
1876...........
1877..........
1878...........
1879...........
1880...........
1881...........
1882...........
1883...........
1884...........
1885...........
1886...........
1887...........
1888...........
1889...........
1890...........
1891...........
1892...........
1893...........
1894...........
1895...........
1896...........
1897...........
1898...........
1899...........
1900...........
1901...........
1902...........
1903...........
1904...........
1905...........
1906...........
1907...........
1908...........
104 |
85 !
91 !
96
107
92
98
142
155
137
108
93
93
96
107
107
103
103
103
99
103
99
103
96
98
106
106
105
102
102
90
93
95
103
111
106
116
118
123
138
131
111
109
108
106
115
117
118
125
102
100
121
113
133
171
131
122
152
148
131
124
118
139
167
178'
202
104
108
103
100
100
100
98
99
101
100
105
100
88
102
97
97
91
85
90
96
100
103
113
116
118
120
111
114
90
103 |
98
111
120
117
113
113
115
116
128
134
127
102
104
106
112
108
118
121
140
144
122
105
91
95
97
105
94
97
125
135
123
106
96
95
96
104
103
100
101
108
114
114
116
128
125
119
116
127
149
122
117
139
135
124
119
117
130
153
162
174
105
91
95
98
105
94
97
100
93
96
105
104
95
96
93
122
130
119
105
95
94
95
88
87
93
90
89
83
82
79
84
87
97
96
96
102
101
m
113
124
88
116
114
124
143
118
114
135
132
113
121
:
88
85
84
82
76
72
69
70
72
72
72
122
101
110
122
96
95
94
97
83
107
88
88
90
93
98
102
68
94
102
101
109
HI
100
112
112
91
99
90
84
91
96
100
116
115
127
349
158:
168
68
68
63
62
61
62
64
68
75
70
69
69
70
72
77
80
73
m
100
99
99
96
93
87
90
100
100
84
99
99
96
93
87
90
85
84
86
86
86
85
85
83
• 83
80
75
73
70
70
78
74
65
58
48
49
50
45
44
45
46
45
40
41
43
46
51
50
40
85
85
83
83
80
75
73
70
70
78
74
65
64
57
57
61
65
68
69
68
68
69
69
69
66
69
69
68
'
103
97
101
102
108
98
96
105
107
109
106
100
96
97
100
94
93
97
101
106
103
105
115
122
116
115
113
115
102
102
112
110
110
108
114
113
128
136
131
More recent index numbers of wholesale prices compiled b j the
Department of Statistics of India are based an prices at the end of
July, 1914, as 100 and include 75 commodities. Available figures
for this series follow.
288
INDEX NUM BERS OF WHOLESALE PRICES.
IN D E X NUM BERS OF TH E D E P A R T M E N T OF STATISTICS, CALCUTTA, IN D IA .
Raw cotton.
Jute manufac
tures.
Other textiles.
Oils, mustard.
Raw jute.
Oil seeds.
Tea.
Sugar.
100
100
100
100
100
100
100
100
100
100
100
100
147
153
179
158
135
144
132
139
154
142
133
116
149
150
170
142
147
151
163
163
136
118
107
162
159
161
164
164
168
168
164
164
164
156
128
135
116
123
119
119
115
115
132
118
124
104
118
119
99
105
96
89
90
354
351
357
365
364
364
360
347
343
341
333
83
89
91
105
104
90
83
173
166
163
169
171
169
167
179
184
163
152
78
87
90
90
73
74
72
65
64
62
69
407
321
377
511
482
503
477
456
392
348
273
166
160
159
150
149
159
160
170
169
168
149
154
151
156
157
156
151
154
154
155
150
139
184
192
185
183
180
188
185
186
178
173
160
81
80
97
324
306
311
107
104
119
104
149
149
123
116
97
85
80
85
130
124
131
77
70
76
314
352
354
135
119
147
139
129
141
139
148
150
100
100
1920.
Average for the
year....................
March....................
April......................
M ay.......................
June.......................
July.......................
August..................
September............
October.................
N ovember............
December.............
138
127
114
128
131
139
142
158
154
161
161
231
218
222
211
215
233
235
235
237
282
246
229
219
248
244
249
257
245
245
243
242
209
160
116
1921.
January................. 158
February.............. 147
March.................... 147
238
226
242
247
243
255
201
238
100
101
97
110
120
&
Other foods.
Cotton manu
factures.
100
End of July, 1914.
Date.
Cereals.
Hides
and
skins.
100
Manufactured
articles.
i
3
Building ma
terials.
Metals.
[Source: Federal Reserve Bulletin, May, 1921, p 604.]
Index numbers for all of the 75 commodities combined are as
follows:
IN D E X NUM BERS OF TH E D EPAR TM EN T OF STATISTICS, CALCUTTA, INDIA.
[Source: Federal Reserve Bulletin, March, 1921, p. 317, and May, 1921, p. 601.]
Year and month.
End of July, 1914.....................
Year 1920....................................
1920.
January......................................
February....................................
March..........................................
April............................................
May.............................................
June.............................................
Julv.............................................
August........................................
September.......................... .
October.......................................
N ovember..................................
December...................................
1921.
January......................................
February....................................
March..........................................
Index
num
ber.
100
204
218
209
198
200
210
206
209
209
208
206
194
180
178
174
183
ITALY.
INDEX NUMBERS OF THE ANNUARIO STATISTICO ITALIANO.
PUBLICATION.
Index numbers based on the prices of a limited number of com
modities at wholesale are contained in the annual statistical report
for Italy (Annuario Statistico Italiano), issued from the Office o f the
Minister of Industry, Commerce and Labor (Ministro per VIndustria,
il Oommercio e il Lavoro).
ITALY---- AN N U A R I0 STATISTICO ITALIAN 0 .
289
HISTORY.
Since 1886 the Annuario Statistieo Italiano has presented tables
showing fluctuations in the prices of a large number of commodities,
both raw and manufactured, during a series of years. In the earlier
reports these prices were shown for a period extending, in some
instances, back to 1862. In more recent issues the figures have been
limited as a rule to the last five years preceding the date of publication.
Prior to 1912 no index numbers appear to have been computed,
the data being given only in the form of actual average prices. In
the report for 1912, however, was begun the publication of two
series of index numbers based in the one case on the prices of a few
articles of food furnished to the army, and in the other case on a
larger number of articles of the same class supplied to 43 national
boarding schools (convitti nazionali) of Italy. The latter series of
index numbers has been continued in the reports for subsequent
years.
SOURCE OF QUOTATIONS.
The price quotations on which the index numbers are based were
furnished by the directors of the schools and by the minister of war
(ministero della guerra, direzione generate dei servizi logistici e amministrativi).
BASE PERIOD.
The five years 1890-1894 constitute the base period in the series
relating to the boarding schools. In the series for the army the
relatives are based on the period 1900-1904.
PRICES: H O W SHOW N AND COMPUTED.
Only the average annual prices of the different commodities
included in the two indexes are given in the reports. In several
instances data for earlier years are lacking from the figures relating
to the army.
NUMBER AND CLASS OF COMMODITIES.
The table of index numbers for supplies furnished to the army
contains 8 commodities, while that for boarding schools contains 13
commodities. All articles belong to the food group.
DESCRIPTION AND GROUPING OF COMMODITIES.
The following articles are included in the table of index numbers for
the army: Corn, bread (ration), Italian paste, rice, beef (young steer),
coffee (roasted), sugar, and wine. The list of articles supplied to
boarding schools for which index numbers are shown includes bread,
Italian paste, rice, beef, sausage, fish (in oil), eggs, butter, oil, milk,
coffee, sugar, and wine. In the latter series the index for beef is based
on the average of the prices paid for young steer flesh and veal in a
single institution. No further description of the commodities is
furnished.
SUBSTITUTIONS, ADDITIONS, AND INTERPOLATION.
So far as can be determined from the information at hand, no
additions to the list of articles or substitutions of one grade or quality
of an article for another haVe been made. No prices appear to have
been interpolated.
WEIGHTING.
All commodities are given equal weight in the computation of the
general index number for each series.
3 3 2 2 6 °— 21— B u ll. 284--------19
INDEX NUM BERS OF WHOLESALE PRICES.
290
TESTING.
The accuracy of these index numbers is not tested by comparison
with similar data or by other means, so far as the published results
show.
TABLE OF RESULTS.
Table 69 contains the index numbers for the 13 commodities
furnished to boarding schools during the period 1890 to 1916, in
clusive.
6 9 —IN D E X NUMBERS BASED ON PRICES PAID B Y THE N A T IO N A L SCHOOLS
(CONVITTI NAZIO NALI) FOR COMMODITIES N ECESSA R Y FOR T H E NOURISHM ENT
OF THE PUPILS, 1890 TO 1916.
T able
[Armuario Statistico Italiano, 1912, p. 138, and 1916, p. 197.]
(Average prices in 18S©-1894=*=100.)
Commodity.
1890
1891
1892
1893
Bread............... ‘ 99.7 1 0 1 . 1 107.6
Italian paste.. 1 0 2 . 1 1 0 1 . 2 103.1
Rice................. 1 0 2 . 1 101.7 103.1
99.4 104.7 100.5
Beef i ...............
99.8 98.9 1 0 1 . 1
Sausage...........
Fish (in o il)... 1 0 0 . 0 1 0 1 . 0 101.7
Eggs................. 1 0 1 . 1 99.3 100.4
99.5 97.4 1 0 1 . 0
Butter.............
99.3 99.4 1 0 2 . 1
Oil....................
Milk................. 103.8 1 0 2 . 1 97.2
Coft'ec............... 98.8 99.1 96.5
97.4 96.2 98.6
Sugar...............
Wine................ 115.5 1 1 0 . 0 90.7
1894
99.2
99.6
96.5
98.8
1 0 1 .1
98.5
99.0
1 0 0 .2
1 0 1 .1
1 0 0 .6
99.5
lj
90. 7|
100.
General j
index
number 101.4 100.9 100.3, 98.8
1895
1896
1897
1898
1899
1900 ; 1901
1902
92.1 95.5 92.4 94.1 103.9 103.1 1 0 :2 . 8 1 0 0 . 0 96.9
94.2 91.8 91.7 97.9 99.8 1 0 0 . 0 99.0 98.1 95.
96.3 97.5 96.1 101.3 99.1 98.5 95.7 93.9 93. 7
96.4 97.0 95.2 93.8 91.4 90.7 90.6 93.5 93.8
98.8 100.3 101.3 101. 5 1 0 2 . 2 98.2 104.2 104.8 104.3
98.5 1 0 0 . 0 101.4 1 0 2 . 0 1 0 2 . 6 1 0 0 . 8 103.1 1 0 0 . 2 103.6
100.3 99.7 100.7 1 0 0 .9 101.7 1 0 0 . 2 107.8 108.9 1 1 1 . 0
101.7 1 0 1 .-o 1 0 0 . 2 102.4 98.8 97.4i 98.4 1 0 2 . 0 99.9
98.3 97.6 92.8 91.5 105.0 1 0 0 . 7 1 1 0 . 8 1 1 1 . 1 103.8
96.5 97.5 95.4 91.6 93.0 93. 7i 89.2 89,,9 92.7
106.0 105.1 102.7 98.9 93.9 85.3! 853.8 82.4 78.5
107.5 100.9 101.7 100.3 99.8 1 0 1 . 0 1 0 1 . 1 100.9 96.6
93.2 94.6 1 0 2 . 8 91.3 94.3 95.2 94.9 93.8 87.;3
1905
1906
1907
95.3j
96.7
97.4,
1 0 0 .0
1 1 1 .2
101.3
102.7
92.0
75. 6
93.8
89.3
98.4
98.3
98.0
97. 5; 98.9
1908
1909
1910
1911
1912
1913
1914
1915 : 1916
104.8
99.2
95. 5
124.7
139.1
140.1
140.0
113.9
146.6
112.5
84.2
108.2
103.5
95.3
132.6
138.3
134.6
140.9
118.1 I
138.1 i
114. 9,|
92.9 !
104. 8 j
133. 8
109.6
106.4
100.3
123.6
143.6
145.5
142.3
119.9
139.0
113.5
94.5
106.5
106.4
97.1
122.4
142.6
154.1
149.3
119.3
137.9
118.0
90.5
92.5
93.0
129.5
127.9
104.1
143.5
150.8
167.4
164.5
129.6
143.6
91.3 91.3 93. 8
93.0 97.2 103.7 106.8
Bread...............
Italian paste... 89.7 90.5 91.8 94.4 1 0 0 . 6 102.5 1 0 2 . 1
93.3 92.1 93.7 94.3 96.0 98.4 100.5
Rice..................
96.6 101.4 1 0 2 . 6 1 0 2 . 8 1 0 2 . 6 1 1 1 . 2 1 1 2 . 0
Beef 1 ................
Sausage............ 109.1 1 1 1 . 6 109.4 111.7 118.9 118.4 130.6
Fish (in oil). . . 104.2 102.4 104.8 106.9 117.3 127.5 133.6
Eggs.................. 110.9 114.9 115.9 121.4 126.2 131.5 133.4
Butter.............. 101.7 1 0 0 . 6 103.1 104.9 107.4 1 1 1 . 0 • 112.4
Oil..................... 95.2 1 0 0 . 0 94.3 1 0 1 . 0 105.1 145.3 135.8
89.9 91.6 93.4 98.6 99.6 103.8 109.3
Milk..................
Coffee................ | 74.6 74.5 75.7 75.6 77.4 76.4 76.1
Sugar................ 1 92.2 97.4 96.0 96.7 96.8 96.4 97.8
Wine................. ; 89.6 89.3 92.1 98.3 84.5 71.7 77.5
General
i ndex
number
91.5
91.7
93.7
92.6
107.6
115.0
1
97.3
98.6
98.4
96. 8 ! 97.1
I
1
Commodity. | 1904
1903
J102.3
1 0 0 .1
123.5
1 0 0 .2
118.4
129.6
167.3
1 2 1 .1
I
j
1
1
107.5j 109. 8 ( 117.2
93.5
98.5
110.3
134.6
154.3
113.1
195.1
193.9
223. 3
225.7
156.8
180.1
137.1
106.6
138.7
215.6
1
119.7( 119.7
117.7j
1 The prices used represent the average prices paid for beef (steers) and veal in one school.
Statistico ItaUano, 1912, p. 338; 1913, p. 184; 1916, p. 197.
Annuario^
INDEX NUMBERS OF PROF. RICCARDO BACHL
PUBLICATION.
This series of index numbers is published annually in YItalia
Economica, a yearbook of economics edited by Prof. Riccardo Bachi,
a distinguished economist of Italy. The series is also carried in
current issues of the quarterly bulletin o fth e French General Statis
tical Office (Bulletin de la Statistique Generate de la France), in the
monthly bulletin of the International Institute of Statistics (Bul
letin Mensuel de VOffice Permanent, Institut International de Statis
tique) published at The Hague, Netherlands, and in the monthly
bulletin of the Federal Reserve Board, Washington.
ITALY---- PROF. RICCARDO BACHI.
HISTORY.
The publication of these index numbers was begun in 1915, and has
been continued to date in the publications mentioned above.
SOURCE OF QUOTATIONS.
It is stated that the prices are for the most part taken from the
ublished reports of the Genoa bourse, the most important market of
taly. For some articles recourse has been had to official prices
established by the cities of Milan, Rome, Naples, and Palermo.
f
BASE PERIOD.
The years 1901-1905 comprise the base period used in computing
the index numbers up to the end of 1920. In that year a new series
was constructed with the average end of month prices for the year
as the basis for the calculations. The new series was, however,
shifted to the old base by the compiler to preserve the continuity of
the data.
NUMBER AND CLASS OF COMMODITIES.
Until 1920 Prof. Bachi based his original index numbers on the
prices of 40 articles, divided into 5 groups:
1. Cereals and meats.
2. Other foodstuffs.
3. Textiles.
4. Metals and minerals.
5. Miscellaneous.
The following 40 articles were included in the index: Hard wheat,
soft wheat, rice, oats, maize, flour, paste, beef, pork, codfish, olive
oil, butter, cheese, coffee, sugar, lemons, almonds, Latium wine,
Apulian wine, American cotton, Indian cotton, wool, hemp, raw silk,
thrown silk, Cardiff coal, bar iron, cast iron, copper, zinc, lead, sul
phur, petroleum, lumber, brick, lime, copper sulphate, hay, leather.
Two of these, flour and lemons, were later left out of the reckoning
because their prices could not be ascertained with any assurance so
that only 38 commodities have been included in recent months. In
1920 the number of commodities was increased to 76.19
Vegetable foodstuffs:
Wheat, soft.
Wheat, hard.
Rice.
Oats.
Rye.
Maize.
Macaroni.
Potatoes.
Almonds.
Tomato preserve.
Wine, Latium.
Wine, Emilia.
Wine, Sicily.
Olive oil.
Seed oil.
Dried beans.
Coffee.
Sugar.
Cocoa.
Animal foodstuffs:
Halt codfish.
Tunny.
Butter.
Cheese.
Honey.
Eggs.
Beeves.
Calves.
19
Animal foodstuffs:
Hogs.
Bacon.
Chemical products:
Copper sulphate.
Sodium nitrate.
Ammonium sulphate.
Calcium carbide.
Sulphuric acid.
Caustic soda.
Citric acid.
Tartaric acid.
Textile materials:
Cotton, American.
Cotton, Tndian.
Cotton yarn.
Wool, Sardinia.
Wool, Aleppor.
Hemp.
Raw silk.
Thrown silk.
Cocoons.
Minerals and metals:
Coal, Cardiff.
Coal, Newport.
Coal, American.
Iron,, wrought.
Iron, sheet.
Tin plate.
Minerals and metals:
Zinc.
Tin.
Copner.
Lead.
Antimony.
Sulphur.
Building materials:
Bricks.
Lime.
Cement.
Lumber, fir.
Lumber, pitch pine.
Miscellaneous vegetable products:
Hay.
Wheat straw.
Wood charcoal,
firewood.
Miscellaneous industrial materials:
Alcohol.
Tobacco.
Illuminating gas.
Electric power.
Hides, raw.
Hides, tanned.
Kerosene.
Paper.
Soap.
292
INDEX NUMBERS OF WHOLESALE PRICES.
WEIGHTING.
As far as the published information shows, no weights were used
in computing the older indexes. In the new series the geometric
mean is employed.
TABLES OF RESULTS.
The 1916 volume of the Italia Economica gives two tables of
index numbers. The first one is a monthly index number, by groups,
for the years 1914 to 1916, and the second is a yearly index, by
groups, for the years 1913 to 1916. In Table 70 the monthly figures
For the years 1917-1919 have been taken from the monthly bulletin
of the International Institute of Statistics.
T able
7 0 .—M O N TH LY IN D E X NUM BERS, B Y GROUPS OF COMMODITIES, JAN U AR Y,
1914, TO DECEM BER, 3920.
[Data from Italia Economica and Bulletin Mensual de POffice Permanent, Institut International de
Statistique.]
(Average prices in 1901-1905=100.)
Year and month.
1914.
January.............................................................
February...........................................................
March.................................................................
April..................................................................
All com
modities.
Cereals
and
meats.
128.6
119.7
117.9
116.5
118.2
117.0
115.6
135.1
133.7
130.7
129.8
137.4
132.7
127.2
117.0
115.6
115.7
July....................................................................
August...............................................................
September.........................................................
October..............................................................
November.........................................................
December..........................................................
119.4
123.3
127.2
127.1
117.0
116.3
116.0
115.1
115.7
115.9
114.0
114.6
119.6
135.4
139.5
136.0
1915.
January.............................................................
February...........................................................
March.................................................................
April...................................................................
May....................................................................
June....................................................................
July....................................................................
August...............................................................
September.........................................................
October..............................................................
November.........................................................
December..........................................................
132.7
140.7
146.5
152.3
159.0
161.3
164.4
170.0
177.8
186.4
200.4
214.8
1916.
January.............................................................
February...........................................................
March.................................................................
April...................................................................
May....................................................................
June....................................................................
July....................................................................
August...............................................................
September.........................................................
October..............................................................
November.........................................................
December..........................................................
1917.
January.............................................................
February...........................................................
March.................................................................
April...................................................................
May....................................................................
June....................................................................
July....................................................................
August...............................................................
September.........................................................
October..............................................................
November.........................................................
December..........................................................
Other
food
stuffs.
Textiles.
Metals
Miscella
and
minerals. neous.
114.2
1 2 0 .0
118.7
117.4
114.1
113.7
114.4
115.6
116.9
123.4
103.8
97.8
104.7
119.8
1 0 2 .2
148.7
141.5
129.6
138.0
142.1
142.9
148.4
152.3
157.1
155.0
149.5
156.9
158.0
168.2
173.5
168.5
178.2
116.0
115.1
116.8
129.7
126.7
124.6
133.4
141.4
151.1
157.4
164.3
166.2
103.5
124.9
125.0
115.8
116.3
126.3
127.0
140.2
151.8
157.2
171.1
170.3
167.0
186.7
207.9
217.8
241.3
260.3
253.3
253.0
254.8
276.1
316.0
371.1
126.4
129.0
134.2
132.7
147.7
154.5
155.8
162.1
165.8
175.7
193.0
206.2
232.3
240.5
247.8
252.7
254.5
242. 8
242.8
248.1
256.7
260.2
269.5
294.6
179.9
183.5
187.9
185.7
184.3
184.8
186.2
187.3
186.2
176.3
195.5
205.7
177.9
181.7
186.8
192.9
195.8
194.7
189.1
208.4
203.5
198.7
204.6
198.8
190.7
194.1
208.9
224.2
246.3
248.3
275.7
278.8
434.3
448.7
469.7
482.8
491.9
459.7
402.8
398.4
423.7
453.0
465.7
594.2
223.3
227.3
231.9
236.2
232.8
236.5
236.7
238.7
240.1
248.1
243.7
288.1
303.0
328.1
333.5
350.1
362.9
383.1
391.9
416.8
441.6
458.8
459.3
226.0
227.0
227.2
229.8
249.2
255.1
283.6
320.5
322.1
322.9
219.6
226.7
231.2
231.2
244.2
245.4
252.1
254.6
257.1
267.6
288.8
276.5
289.0
299.0
330.3
320.5
339.4
371.0
398.1
411.8
438.7
498.4
514.7
515.5
506.1
543.4
625.4
645.8
716.7
722.5
738.1
762.7
831.2
829.7
860.8
870.7
247.6
257.7
271.0
286.0
271.6
297.1
336.9
337.0
337.3
359.4
378.7
384.0
1 2 2 .0
210.4
2 2 1 .8
2 0 2 .2
213.0
2 1 2 .2
2 0 1 .1
1 1 2 .6
112.4
110.7
117.9
118.1
115.2
115.5
1 1 0 .8
1 1 1 .2
111.7
107.7
108.1
115.5
1 1 1 .6
1 1 1 .2
115.3
114.3
120.9
2 1 2 .0
ITALY---- PROF. RICCARDO BACHI.
T able
7 0 .—M O N T H L Y IN D E X N U M BER S, B Y GROUPS OF COMMODITIES, JA N U A R Y ,
1914, TO D EC E M B E R , 1920—Concluded.
Year and month.
1918.
January.............................................................
February...........................................................
March.................................................................
April..................................................................
May....................................................................
June....................................................................
August...............................................................
September.........................................................
October.............................................................
November.........................................................
December..........................................................
1919.
January.............................................................
February...........................................................
March................................................................
April..................................................................
May................................................................
July....................................................................
August...............................................................
September.........................................................
October..............................................................
November.........................................................
December..........................................................
1920.
January....................................................... .
February..........................................................
March................................................................
April..................................................................
May....................................................................
July....................................................................
August..............................................................
September........................................................
October.............................................................
November.........................................................
December.........................................................
T able
293
All com
modities.
Cereals
and
meats.
457.6
478.8
496.8
505.6
515. 4
523.5
540.6
544.1
545.8
556.4
550.7
467.9
326.0
336.4
353.8
360.9
347.9
335.5
391.7
403.9
413.5
413.9
414.5
360.8
264.3
266.2
274.1
288.9
299.4
326.0
338.7
348.0
360.3
407.3
416.8
418.4
513.8
540.6
562.4
575. 6
577.6
589.6
590.0
582.7
585.8
584.3
584.3
584.3
869.4
939.1
954.6
949.6
950.8
975.1
992.8
978.2
942.0
939.8
891. 7
520.9
388.1
392.5
421.1
435.2
454.3
468.1
468.1
481.4
499.8
502.9
509.1
509.5
410.1
403.9
410.1
417.9
426.3
451.0
456.6
465.5
468.3
492.0
552.9
576.2
370.8
371.3
356.7
365.8
365.1
396.6
408.7
405. 8
385.3
407.2
410.2
420.4
438.7
449.7
455.1
482.1
491.2
501.8
481.9
513.0
521.2
534.7
542.4
544.9
398.4
395.2
399.3
401.7
452.8
455.9
483.2
509.7
517.9
601.9
764.0
793.5
357.3
357.3
418.2
428.2
435.4
506.6
511.8
512.8
535.3
555.0
687.1
706.4
496.9
453.2
425.6
411.4
401.4
396.0
403.7
402.4
402.7
412.0
414.1
477.1
63< 7
701.0
780.0
855.7
830.3
774.7
772.4
795.9
832.2
834.3
829.1
800.6
436.5
438.6
458.5
475.2
530.2
535.0
521.9
534.8
551. 7
536.0
570.6
577.8
578.1
582.2
610.4
712.6
728.4
746.7
742.7
745.6
759.7
733.2
781.8
775.6
937.4
1013.2
1159.6
1283.9
811.9
1036.7
1205.2
1302.5
1317.3
1109.9
1093.0
1157.9
1257.8
1303. 4
1220. 8
1123.1
492.3
522.7
576.0
630.8
619.4
628.4
639.5
637.0
637.9
674.1
689.6
735.6
Other
food
stuffs.
Textiles.
1 0 1 2 .8
895.3
915.4
957.3
1009.0
976.3
920.0
813.8
Metals
Miscella
and
neous.
minerals.
7 1 ___Y E A R L Y IN D E X NUM BERS, B Y GROUPS OF COMMODITIES, 1913 TO
1916.
[Data from Italia Economica.]
(Average prices in 1901-1905=100.)
Year.
1913.....................................................................
1914.....................................................................
1915.....................................................................
1916.....................................................................
All com
modities.
126.0
119.8
167.2
251. 8
Cereals
and
meats.
Other
food
stuffs.
Textiles.
Metals Miscella
and
minerals. neous.
1 2 0 .2
1 2 2 .6
146.1
1 2 0 .6
1 2 1 .6
1 2 2 .6
159.5
188.3
136.4
195.5
116.1
139.1
219.4
250.1
460.4
1 2 1 .8
117.9
113.4
157.8
224.9
294
INDEX NUM BERS OF WHOLESALE PRICES.
The following index numbers for 1920 and to February, 1921,
have been supplied by the American trade commissioner at R om e:
BACHFS IN D E X NUMBERS COMPUTED ACCORDING TO N E W SYSTEM OF GEOMETRIC
AVER AG ES.
(76 commodities; average prices for months of 1920=100.)
Year and month.
Index j
number, j
1920.
January.......................................
February.....................................
March...........................................
April............................................
May..............................................
June..............................................
July..............................................
August........................................
September..................................
October......................................
November.................. ...............
December....................................
81. 25
89.13
96.41
106. 30
10.5. 73
101.18
97. 77
100.13
104. 98
105. 47
107.33
104.97
1921.
January.......................... ............
February.....................................
102.89
98.23
INDEX NUMBERS OF ACHILLE NECCO.
PUBLICATION AND HISTORY.
A volume entitled “ La Curva dei Prezzi delle Merci in Italia negli
Anni 1 8 8 1 - 1 9 0 9 (The Price Curve of Commodities in Italy during
1881-1909), which was published in Turin by Achille Necco in 1910,20
contains four series of index numbers based in each case on the import
or export values of certain important articles of commerce. Com
parative tables showing the price fluctuations in several countries
on a common basis, that of the year 1881, are also contained in the
volume.
A continuation of the two principal series of these index numbers
was published by Necco in La Riforma Sociale for 1911, pages 68-72,
and 1913, pages 621-635; also in a special bulletin entitled “ Prezzi
della Merci in Italia nel 1912” (Prices of commodities in Italy during
1912) , appearing in 1914. In the last-named publication which also
was issued by La Riforma Sociale, the figures for 1910, 1911, and
1912 were supplied. The issue of La Riforma Sociale for SeptemberOctober, 1920, pages 341-355, brings the information up to the end
of 1918.
SOURCE OF QUOTATIONS.
The data used in computing the index numbers were those formu
lated by the Central Commission for Customs Valuation ( Commissione Centrale dei Valori per le Dogane) and published in the reports of
the Ministry of Agriculture, Industry, and Commerce.
BASE PERIOD.
The values of imported and exported articles for the year 1881,
taken as 100 , constitute the bases on which the two principal series
were computed. The other two series appearing in the first volume
20
Societa Tipografia-Editrice Nazionale (gia Roux e Viarenga), Torino. The same information also was
published as a supplement to La Riforma Sociale, Vol. X X I , September-October, 1910.
ITALY---- ACHILLE KECCO.
295
issued in 1910 are based on the method employed by Pantaleoni and
have for the standard of measurement the values of imports and
exports, respectively, in 1878.
NUMBER AND GROUPING OF COMMODITIES.
In each of the two main series of index numbers the groupings
adopted in the collection of the customs revenue have been followed.
Within the 16 principal groups there are approximately 400 different
articles, each of which in turn may comprise several varieties.
The 16 groups of commodities are as follows:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Spirits, beverages, oils.
Colonial products, spices, tobacco.
Chemical products, m edicinal substances, resins, gums, etc.
Coloring and other materials for dyeing and tanning.
H e m p , flax, jute, and other fibrous plants.
Cotton.
W ool, horsehair, and other hair.
Silk.
W ood and straw.
Paper and books.
H ides.
Minerals, metals and their manufactures.
Stone, earthen, glass, and crystal ware.
Cereals, flour, Italian paste, and vegetable products.
A nim als and animal products.
Miscellaneous commodities.
The two index numbers of import and export values computed
according to the method adopted by Pantaleoni are likewise based
on commodities selected from the tariff schedules. There are 19
import and 12 export commodities as follows:
Imports:
1. Petroleum, refined.
2. Coffee, raw.
3. Sugar, second grade.
4. D yes, in dry state.
5. Cotton, raw.
6. Cotton cloth, unbleached.
7. W ool, natural or unwashed.
8. W oolen cloth (com bed wool).
9. W oolen cloth (carded wool).
10. H ides of oxen and cows.
11. Leather.
12. Cast iron (in pigs or plates).
13. Bar iron and steel in sheets.
14. Copper, brass, and bronze.
15. M achinery (not specified).
16. Grain.
Imports— Concluded.
17. Cheese.
18. D ry goods, ordinary.
19. D ry goods, fine.
Exports:
1. W in e in bottles.
2. O live oil.
3. Boric acid.
4. Sulphate of quinine.
5. Soap (com m on).
6. H em p , raw.
7. Silk, raw.
8. Straw braid for hats.
9. Sulphur, raw and refined.
10. Oranges and lemons.
11. Alm onds (shelled).
12. Coral (polished).
WEIGHTING.
In the computation of his two main series of index numbers Necco
adopted the method employed by De Foville in following the changes
in tne import and export values of France from 1826 to 1880. It
consists in weighing the prices of the first of any two consecutive
years with the mass quantities of imports or exports of the second
year. The price of the preceding year in any instance is multiplied
by the mass quantity of the current year, giving what is termed the
“ provisional value.” The price of the current year is then multiplied
by the mass quantitv to produce the “ actual value.” These pro-
296
INDEX NUMBERS OF WHOLESALE PRICES.
visional and actual values of the imports or exports are then summed
and compared to ascertain the increase or decrease which has taken
place in these values as between the two years under comparison.
For example, it might be found that the import values of 1881 were
2 per cent lower than those of 1880; those of 1882, 4 per cent higher
than those of 1881; those of 1883, 3 per cent lower than those of
1882; and so on. Assuming now that the import values of the year
1880 are taken as the base, or 100, the index for 1881 would be 98,
since the import values in 1881 decreased 2 per cent from those of
1880. Again in 1882 the import values increased 4 per cent over
what they were in 1881; that is, 4 per cent of 98, or 3.92. Therefore
the index for 1882 is 98 plus 3.92, or 101.92. In like manner, since
import values in 1883 were 3 per cent lower than in 1882, the index
for 1883 becomes 97 per cent of 101.92, or 98.86.
It is seen that under Necco’s system there is a constantly changing
weight, namely, the quantity of an article imported or exported
each year. No direct relationship exists between the index number
of any one year and that of the basic year, since, as has been said,
the relative importance of a commodity changes from year to year
according to the quantity imported or exported, as the case may be.
Pantaleoni likewise employed a fluctuating weight in determining
the relative importance of the commodities entering into his index
number. Under his original plan there was determined each year
the percentage which the value of each commodity imported or ex
ported, as the case might be, formed of the total value of all imported
or exported commodities. This figure was then used as a weight
for each commodity included in the final index number. To simplify
this rather laborious process from year to year, Pantaleoni sug
gested— and Necco followed the suggestion in extending Pantaleoni’s
indexes— that it would be advisable to ascertain the average import
or export value of each commodity concerned over a period of years
and then calculate the ratio between the average value of each com
modity so ascertained and the total average value of all imports and
exports over the same period and use the result as a weight for each
commodity for each of the years involved in the period under con
sideration. Necco has done this for each of the three periods, 18901895, 1896-1901, and 1902-1908, in the case of imports and for each
of the two periods, 1890-1898 and 1899-1908, in the case of exports.
The result of this method is that there is employed a constant
weight over a limited number of years, a weight which may be termed
the average importance of the particular commodity as determined
by its proportionate value in the total import or export trade of the
country, as the case may be.
TABLE OF RESULTS.
On page 32 of Necco’s original work 21 are given the following
index numbers for imports and exports, respectively, for the years
from 1881 to 1909. The first series in each case is Necco’s own
number computed according to the method of Benini. The two
remaining series are those o f Pantaleoni and Benini reduced for the
sake of comparison to a common-base period, that of the year 1881.
Necco’s figures for the years 1910 to 1912 have been supplied from
21
La Curva dei Prezzi della Merci in Italia negli Anni 1881-1909.
ITALY— ACHILLE NECCO.
297
the “ Prezzi della Merci in Italia nel 1912,” published in 1914, while
those for years subsequent to 1912 have been taken from La Riforma
Sociale, September-October, 1920, pages 353 and 354.
T able
7 2 .—IN D E X NUM BERS OF IT A L IA N IMPORTS AN D EX P O R TS.
(Average prices in 1881=100.)
Imports.
Exports.
Year.
Necco.
1881................................................
1882................................................
1883................................................
1884................................................
1885................................................
1886................................................
1887................................................
1888................................................
1889................................................
1890................................................
1891................................................
1892................................................
1893................................................
1894................................................
1895................................................
1896........................ .......................
1897................................................
1898................................................
1899................................................
1900................................................
1901................................................
1902................................................
1903................................................
1904................................................
1905................................................
1906............................................
1907................................................
1908................................................
1909................................................
1910................................................
1911................................................
1912................................................
1913................................................
1914................................................
1915................................................
J916................................................
1917................................................
1918................................................
1 0 0 .0 0
96.86
93.01
87.42
82.68
81.95
79.53
81.19
82. 58
83.23
79.25
77.43
76.73
71.81
71.04
70.96
70.42
74.49
79.77
86.47
79.65
76.75
77.73
80.05
79.52
84.29
87.96
84.55
85.45
8 6 .55
87.35
89.85
90.05
91.34
128. 96
198.24
358.68
407. 96
Pantaleoni.
Benini.
1 0 0 .0 0
98. 98
94.90
8 6 .73
78.57
74.49
71.43
73.47
73.47
7-8.47
71,43
68.37
66.33
61.22
60.20
60.20
60.20
62.24
6 6 .33
6 6 . 33
65.31
63.27
64.30
66.33
67.35
71.43
73.47
73.47
76.53
1 0 0 .0 0
98.43
94.06
87.64
82.49
80.81
79.61
79.97
81.58
82. 24
78.62
77.04
76.41
71.14
69.47
69.87
69.30
73.23
76.81
83.17
77.37
74.48
Necco.
1 0 0 .0 0
96. 84
91.96
88.08
84.64
84.11
79.62
76.73
80.49
81.72
76.31
76.37
76.18
71.97
72.83
69.02
67.80
69.09
75.55
75.10
72. 73
74. 1 0
76.92
76.07
77.12
79.54
83.72
77. 8 8
79.29
82.12
83.44
83.54
83.80
83.52
95.46
128.58
188.32
253.63
Pantaleoni.
1 0 0 .0 0
92.71
86.46
85.42
79.17
84.37
80.21
69.79
78.12
78.12
67.71
77.08
77.08 '
6 8 .75
70.83
62.50
62.50
64.57
79.17
72.92
69.79
72.92
78.12
72.92
73.96
78.12
86.46
69.79
72.92
Benini.
1 0 0 .0 0
94.44
91.43
85.62
81.95
81.64
77.80
74.40
76.89
79.58
74.25
74.33
74.15
70.59
71.60
67.72
66.40
67.79
74.02
73.58
71.25
72.14
INDEX NUMBERS OF MARIO ALBERTI.
PUBLICATION.
A volume entitled “ II cos to della vita i salari e le paghe a Trieste
nell* ultimo quarto di secolo” (The Cost of Living, Salaries, and
Wages in Trieste During the Last Quarter of the Century), by Mario
Alberti, was published m Trieste in April, 1911, under the direction
of the Museo Commerciale, a nonofficial organization of that city.
A second volume, II movimento dei prezzi e dei salari a Trieste, 1911
(The Movement of Prices and Wages m Trieste, 1911), was published
by the same author in 1912.
HISTORY.
Mario Alberti in his two works has made a study of prices and has
computed index numbers for certain commodities purchased by
contract in Trieste.
In his cost of living study the author traces from the time of ancient
Rome to the present the interest manifested in the question of the
cost of living and the means adopted to study the movement of
INDEX NUM BERS OF WHOLESALE PRICES,
298
prices in various epochs and countries. Chapter IV of this work
deals exclusively with conditions in Trieste. The volume on the
movement of prices and wages also relates exclusively to Trieste.
SOURCE OF QUOTATIONS.
The prices used in the tables of index numbers are contract or semiwholesale prices quoted by the Istituto dei Poveri (Institution for the
Poor) and by the Austrian Lloyd Steamship Co. Those for the
Istituto dei Poveri extend from 1885 to 1911, inclusive, and for the
Austrian Lloyd Steamship Co. from 1892 to 1911, inclusive.
BASE PERIOD.
For the index numbers relating to the Istituto dei Poveri the
year 1885, taken as 100 , is used as a base. For those of the Austrian
Lloyd Steamship Co. the years 1892-1896, taken as 100 , are used
as a base.
PRICES: H O W SH OW N AND COMPUTED.
Tables are presented showing by years the average annual price of
each commodity purchased by the Istituto dei Poveri and by the
Austrian Lloyd Steamship Co. In addition, the simple index, the
coefficient or weight attributed to the commodity, and the weighted
index for the commodity are given. The totals for each year show
the figures on which the simple index and the weighted index for all
commodities taken as a whole are computed.
NUMBER AND CLASS OF COMMODITIES.
The commodities included are:
(a) For the Istituto dei Poveri:
Bread.
Flour.
Paste (macaroni, e tc.).
R ice.
K id n ey beans.
Potatoes.
Oil.
Cheese.
Meat.
Sugar.
W in e.
Vinegar.
Coal.
Soap.
(b) For the Austrian Lloyd Steamship Co.
B read.22
Flour N . G . (national).
Paste (macaroni, e tc.), national,
fine.
R ice, fine, in transit.
K id n e y beans.
Potatoes.
O il, fine, in transit.
Cheese, Parma, in transit.
Beef (cow ).
Sugar.
W in e, D alm atian, in transit.
Vinegar.22
Soap.22
Fresh butter.
Coffee, Santos, in transit.
M ilk.
Lard.
Eggs.
Meal, yellow .
SUBSTITUTIONS AND ADDITIONS.
For 1910 two sets of prices were used in the computation of the
indexes. The report on the cost of living contained prices based
on estimates, while the volume devoted to the movement of prices
and wages contained prices based on actual expenditures. On page
22
prices are average prices paid’by the Istituto dei Poveri.
ITALY— MARIO ALBERTI,
299
21 of the latter the author states that for this reason “ The index
number for 1910 was newly calculated on the base of prices furnished
by the Istituto dei Poveri. ”
WEIGHTING.
The prices of the different articles are weighted by the use of
coefficients which represent the relative amounts of the commodities
consumed. These coefficients are shown for the tables relating to the
Istituto dei Poveri in the following list:
B rea d ...............
Flour, w h ea t.
R ic e .....................
K id n ey beans.
Oil.
26
1
4
2 .5
2
10
2
Cheese.
1-5
20
Sugar. .
W in e . . .
Vinegar
Coal
4
20
3-5
50
1
The coefficients used in the tables for the Austrian Lloyd Steam
ship Co. are as follows:
B rea d .......................................................... .... 26
Flour, w h ea t........................................... .... 1
Paste (macaroni, e t c .) ......................... .... 4
R ic e ............................................................. .... 2 .5
K id n ey b ean s......................................... .... 2
Potatoes....................................... ................. 10
O il................................................................ .... 2
Cheese......................................................... 1 -5
B e e f...................................................... ...........20
Sugar........................................................... .... 4
W in e ............................................................
V in egar......................................................
Soap.
Fresh butter.
Coffee.............
M ilk ................
L ard ................
20
3 -5
1
2
1
30
2
50
M eal, y e llo w .
1
TESTING.
No comparison with other index numbers or other means of test
ing the accuracy of the results obtained has been attempted so far
as the published information discloses.
TABLES OF RESULTS.
Table 73 shows the average price of each of 14 articles for the base
year 1885; the coefficients used in computing the weighted index;
the average price, the simple index and the weighted index for each
of the 14 articles for the years 1910 and 1911, and the totals of the
simple and the weighted indexes for those two years. The prices
in the table are based on reports from the Istituto dei Poveri.
Table 74 shows the indexes, both simple and weighted, for the 14
articles taken as a whole, in yearly periods from 1885 to 1911,
inclusive.
Table 75 shows the average prices for 1892-1896 of 19 articles (18
food articles and soap); the coefficients; the average price, the
simple index and the weighted index for each article' iu the years
1910 and 1911; and the totals of the indexes for 1910 and 1911,
respectively. THe prices are based on the period 1892-1896, taken
as 100 , and were obtained from the Austrian Lloyd Steamship Co.
300
INDEX NUMBERS OF WHOLESALE PRICES.
T a b l e 7 3 ___ IN D E X
Name of article.
Bread............................
Flour, wheat...............
Paste (macaroni, etc.)
Rice...............................
Kidney beans.............
Potatoes.......................
Oil.................................
Cheese..........................
Meat..............................
Sugar............................
Wine.............................
Vinegar........................
Coal...............................
Soap..............................
1
T
NUM BER S BASED ON PRICES OF ISTITI7TO DEI PO VER I.i
1910
Price
in base
period,
1885.
Coeffi
cient.
Price.
32.00
25. 52
36.00
30. 50
15. 52
8 .0 0
1 2 0 .0 0
26
1
4
2.5
2
10
2
137.60
92.00
43.36
20
6 6 .0 0
20
.2
4
17.00
2 .0 0
52. 00 -
.6
50
1
38.17
35.69
48.47
31.82
29.69
8.06
102. 43
213.57
105.00
79.14
46.50
6 .0 0
2.14
56.90
Simple
index.
1911
Weighted
index.
119.28
139.85
134. U
104.33
191.26
100.75
85. 36
155.22
114.13
182. 52
70.45
35.29
107.00
109.42
3,101.28
139.85
538.56
260.82
382. 52
1,007.50
170.72
31.04
2,282.60
730.08
1,409.00
21.18
5,350.00
109.42
1,650
15,535
Price.
37.54
34.19
49.97
33.91
*35.05
11.04
106.70
219.98
138.00
82.13
78.00
6.03
2.04
59.33
Simple
index.
Weighted
index.
117.31
133.97
138.81
111.18
225.84
138.00
8 8 . 92
159.87
150.00
189. 42
118.18
35.47
114.10
3,050.06
133.97
555.24
277.95
451.68
1,380.00
177.84
31.97
3,000.00
757.68
2,363.60
21.28
5,100.00
114.10
1,823
17,415
1 0 2 .0 0
Mario Alberti, II movimento dei prezzi e dei salari nell’ anno 1911 a Trieste, pp. 24 and 25.
able
7 4 . — IN D E X
N UM BERS BASED ON PRICES OF ISTITUTO D EI PO VE R I, FOR EACH
Y E A R , 1885 TO 1911, INCLUSIVE.i
(Average prices in 1885=100.)
Year.
1885................................
1886................................
1887...............................
1888...............................
1889................................
1890................................
1891...............................
1892................................
1893................................
1894................................
1895................................
1896................................
1897................................
1898................................
1899...............................
1900...............................
1901...............................
1902................................
1903...............................
1904................................
1905...............................
1906...............................
1907...............................
1908...............................
1909...............................
1910................................
1911................................
Simple
index.
1 0 0 .0 0
97.21
95.57
94.14
90. 21
92.43
93.80
97.86
96.86
95.43
90.28
91.86
96.93
99. 57
99.87
1 0 1 .2 1
101.78
103.00
102.43
1 0 0 .8 6
108.00
105.00
111.43
116.79
117.14
117.86
130.21
Weighted
index.
1 0 0 .0 0
95.74
93.52
90.54
88.96
94.11
97.55
92.77
92.46
89.69
86.32
90.79
97.40
99.99
102.43
114.95
116.11
108.09
103.31
100.36
104.97
104.36
111.72
117. 55
112.33
108.41
121.53
i Mario Alberti, II movimento dei prezzi e dei salori nelP anno 1911 a Trieste, pp. 24 and 25.
ITALY---- MARIO ALBERTI.
T able
7 5
.—IN D E X NUM BERS BASED ON PRICES O BTAIN ED FROM THE AU ST R IA N
L L O Y D STEAMSHIP CO.i
Name of article.
Bread2 ..............................................
Flour, N. O., national..................
Paste (macaroni, etc.), national,
fine................................................
Rice, fine, in transit......................
Kidney beans.................................
Potatoes...........................................
Oil, fine, in transit........................
Cheese, Parma, in transit............
Beef (cow) 2 .....................................
Sugar................................................
Wine, Dalmatian, in transit.......
Vinegar2 ..........................................
Soap2 ................................................
Fresh butter...................................
Coffee, Santos, in transit.............
Milk..................................................
Lard.................................................
Eggs.................................................
Meal, yellow...................................
1 Mario
2 Index
301
Price
in base
period, Coeffi
cient.
18921896.
27.91G
25. 952
26.00
35.920
36. 880
2 2 . 740
6 . 392
97.908
181.752
94.300
64.996
23.322
12.896
50.092
224. 992
193.300
4.00
2.50
2.00
10.00
2 0 .0 0 0
102.612
6.232
17.364
1.00
2.00
.20
20.00
4.00
20.00
.60
1.00
2.00
1.00
30.00
2.00
50.00
1.00
1910
1911
Weighted Price.
Price. Simple
index.
index.
Simple Weighted
index.
index.
38.17
40.61
136.73
156.48
3,554.98
156.48
37.54
35.22
134.48
135.71
135.71
45.46
43.95
30. 55
7.57
135.40
252.08
105.00
80.99
22. 53
6.00
56.90
231.68
108.87
22.00
187.67
6.60
23.68
126.56
119.17
134.35
118.43
138.29
138.70
111.34
124.61
96. 61
46. 52
113.59
102.98
56.37
110.00
182.90
105.90
136.37
506.24
297.94
268.69
1,184.30
276.58
27.74
2,226.80
498.44
1,932.20
27.91
113. 59
205. 96
56.37
3.300.00
365. 80
5.295.00
136.37
47.86
45.11
31.72
9. 8 6
145.60
240.52
138.00
83.66
44.06
6.03
59.33
259.96
154.38
23.80
160.65
7.18
22.68
133.24
122.32
139.49
154.25
148.71
132.34
146.35
128.72
188.92
46.76
118.44
115.55
79.87
119.00
156.56
115.05
130. 62
532.96
305.79
278.98
1.542.50
297.42
26.47
2.927.00
514.88
3.778.40
28.06
118.44
231.10
79.87
3.570.00
313.12
5.752.50
130.62
2,256
20,431
2,446
24,060
Alberti, II movimento dei prezzi e dei salari nelP anno 1911 a Trieste.
numbers based on prices paid by the Istituto dei Poveri.
JAPAN.
INDEX NUMBERS OF THE DEPARTMENT OF AGRICULTURE AND
COMMERCE.
PUBLICATION.
Wholesale prices, with index numbers for the same, are published
annually in the reports issued by the Department of Agriculture and
Commerce of Japan. These reports, which are printed in English as
well as Japanese, contain various data of a statistical nature relating
to the agricultural, mining, manufacturing, and other industries of
the Empire and its dependencies. No text analysis is given of the
tables included in the reports.
HISTORY.
The quotations of prices, according to N. Hanabusa, director of
the Japanese bureau of statistics, were not matters of record prior to
1886 except for the four standard commodities: Rice, barley, beans,
and sake, for which there were incomplete records for earlier years .23
The latest available report, issued in March, 1920, is the thirty-fifth
of the series.
SOURCE OF QUOTATIONS.
Wholesale prices of the articles for which index numbers are com
piled are obtained from the cities of six statistical divisions of the
Empire by the Department of Agriculture and Commerce. No
statement is made in the reports as to the methods of securing this
information.
23
Bulletin de Tlnstitut International de Statistique, tome X I X , 3® livraison, p. 237.
302
INDEX N UM BEES OF WHOLESALE PRICES.
BASE PERIOD.
The base period for each year from 1901 to 1912, inclusive, is the
year 1900 taken as 100 . In each of the reports for this series of
years, except the one issued in 1905, index numbers are computed for
the years from 1900 to two years earlier than the date of publication.
For example, the report published in 1908 contains index numbers for
the years 1900 to 1906, inclusive. The report issued in 1914 com
prises index numbers for the years 1904 to 1912 only, 1900 being
still retained as the base or 100 . In the thirty-fifth statistical report,
the latest issued, the year 1912 is used as base.
The report of 1905, which is the earliest one available, uses the
year 1887 as a base or 100 and shows average annual prices and
index numbers for each year for the several commodities, from 1887
to 1903, inclusive. In the case of a few commodities,for which data
for 1887 were lacking, a subsequent year was used as the base. There
is no general index number in this volume for the groups of com
modities as in those for succeeding years.
PRICES: H O W SHO W N AND COMPUTED.
Average annual prices are published for each commodity taken
separately for all years subsequent to and including 1900. In the
report of 1905 there is no grouping of the commodities, while in those
for other years the commodities are divided into three main groups—
food, clotiling, and material.
Following the average annual prices of the different articles for the
Empire as a whole, there is a table showing for each commodity the
average monthly price and the average price for the year of that
commodity in each of the principal cities of the several statistical
divisions and in the country at large, all data being for the year prior
to the one preceding the publication of the report.
NUMBER AND CLASS OF COMMODITIES.
In the thirty-fifth statistical report, covering the years 1912 to
1919, only 39 articles are quoted. They are:
R ice, superior.
B arley.
Naked barley.
W heat.
Soy beans.
Sm all red beans.
Salt.
Soy (sauee).
Miso (sauce, soy beans, rice,
W h ite sugar, domestic.
Sake (rice liquor).
Tea.
Bonito (fish), dried.
Beef.
Eggs.
M ilk.
Cut tobacco.
Ginned cotton, domestic.
Cotton y a m s, domestic.
Haw silk, superior.
water).
H em p .
Bleached cotton fabric, domestic.
Calico.
Silk tissue, for lining.
K aik i silk.
Petroleum.
Coal.
Firewood.
Charcoal.
Rapeseed oiL
Hanshi paper.
Balk (beam ) sugi.
Plank, sugi, 4 bu. (0.477 inch) thick.
Sleeper, chestnut.
Pig iron, domestic.
Nails, foreign.
Dried sardines (for manure).
Herring (for manure).
Rapeseed cake.
JAPAN — DEPARTMENT OF AGRICULTURE AND COMMERCE.
303
DESCRIPTION AND GROUPING OP COMMODITIES.
In the report for 1904 and in subsequent issues the commodities for
which average yearly prices and index numbers are given are divided
into three groups: ( 1 ) Food, etc.; (2 ) clothing; (3) materials. Under
food the following articles are listed: Rice, barley, naked barley,
wheat, soy beans, red beans, salt, sake, soy (soy-bean sauce), miso,
tea, bonito (dried fish), beef, eggs, milk, umeboshi (pickled plum),
takuwan (pickled radish), sugar (4 grades), tobacco (2 grades)— a
total of 23 articles.
Under clothing are listed ginned cotton (2 grades), cotton yarns (2
grades), raw silk, hemp, bleached cotton fabric, gray shirting, calico,
silk tissue, kaiki silk— 11 articles in all.
Under materials are listed the following: Petroleum, coal, firewood,
charcoal, rapeseed oil, paper (2 grades), indigo, beams (4 grades),
planks (4 grades), logs (2 grades), shingles, sleepers (2 grades), pig
iron (2 grades), nails, straw, hay, manure fish, (2 grades), rapeseed
cake— altogether 29 articles.
As previously stated, leaf tobacco was discontinued after 1905 and
cut tobacco after 1907.
In the thirty-fifth annual report the 39 commodities quoted are
divided into 5 groups.
The grouping is as follows:
Grain.
Hice.
Barley.
Naked barley.
W h eat.
Soy beans.
Sm all red beana.
Food , etc.
Salt.
Soy.
Miso.
W h ite sugar.
Sake.
Tea.
Bonito, smoked, dried.
Beef.
Eggs.
M ilk.
Cut tobacco.
Clothing.
Ginned cotton.
Cotton ya m s.
Raw silk.
H em p .
Bleached cotton fabric.
Calico.
Silk tissue, for lining.
K aik i silk.
Material.
Balk sugi.
P lank, sugi, 4 bu (0.477 inch) thick.
Sleeper, chestnut.
Pig iron.
Nails (foreign).
Others.
Petroleum
Coal.
Firewood
Charcoal.
Rapeseed oil.
Hanshi paper.
Dried sardines (for manure).
Herring (for manure).
Rapeseed cake.
304
INDEX NUM BEBS OF WHOLESALE PRICES.
SUBSTITUTIONS AND ADDITIONS.
So far as the reports show, no substitutions of a particular grade or
q u a lit y of an article for another grade or quality of the same article or
for a different article have been made at any time. Additions to the
list of articles have apparently been made from time to time as data
became available. In all such cases the average price for the earliest
year for which data were available was taken as the base, or 100 . No
additions have been made since 1901, in which year logs (pine and
su g i24) appear to have been included for the first time.
INTERPOLATION.
No interpolation of prices has been made in any of the reports so far
as can be determined. In cases where prices for a particular month
in a given locality were lacking, the averages for the year and for the
country at large have been based on the data for the remaining
months and localities.
WEIGHTING.
There is no attempt at weighting any of the commodities for which
index numbers have been computed, apart from the inclusion in the
tables of several grades of the more important articles.
TESTING.
No comparison of these index numbers with those for other coun
tries has been made in the reports, nor have other means of testing
their accuracy been employed, so far as the published results show.
TABLES OF RESULTS.
The index numbers for average wholesale prices of four staple
articles, viz, rice, barley, beans, and sake,25 from 1881 to 1909, inclu
sive, are shown in the following table published in the bulletin of the
international statistical institute .26 This table was compiled by the
director of the Japanese imperial statistical bureau, Mr. N. Hanabusa,
and is based on data collected annually during the months of March,
June, September, and December from six principal cities of the Empire
by the minister of agriculture and commerce. Only medium grades
of the articles for which quotations were secured have been included
in the compilation.
24 A genns of evergreen trees of the pine family.
25 A native beer made from rice.
26 Bulletin dePInstitut International de Statistique, tome X I X , 3« livraison, p. 239.
JAPAN— DEPARTMENT OF AGRICULTURE AND COMMERCE.
305
T able 7 6 .—IN D E X N UM BERS FOR FOUR PRINCIPAL COMMODITIES, 1881 TO 1909.
(Average prices in 1881=100.)
Year.
Rice.
1881.......................................................................
1882......................................................................
1883.......................................................................
1881......................................................................
1885.......................................................................
1886......................................................................
1887......................................................................
1888......................................................................
1889......................................................................
1890......................................................................
1891......................................................................
L892.................................................. ...................
L893......................................................................
1894......................................................................
1895......................................................................
1896......................................................................
1897.......................................................................
1898......................................................................
L899......................................................................
1900......................................................................
1901......................................................................
L902.......................................................................
1903......................................................................
1904.......................................................................
1905......................................................................
1906......................................................................
1907.......................................................................
1908......................................................................
1909......................................................................
100
84
60
52
64
56
52
48
61
90
75
77
78
91
90
103
129
145
108
125
127
132
153
143
137
155
173
167
139
Barley.
100
88
63
59
64
55
45
41
47
72
70
64
64
72
73
71
94
117
85
84
72
81
113
137
123
85
106
115
103
Beans
Sake (rice
(Japanese). wine, clear).
100
102
77
61
64
57
58
57
71
78
74
73
79
81
84
93
113
119
123
115
108
104
116
147
146
139
149
134
117
100
89
78
73
94
80
83
76
80
85
84
84
83
91
102
118
146
170
165
192
201
201
205
208
223
239
240
252
251
General
index.
100
90
72
64
77
66
66
61
69
83
78
78
79
87
92
103
129
147
133
147
149
150
164
172
175
180
189
192
180
Table 77, compiled from the 28th, 29th, and 34th statistical reports
of the Japanese Department of Agriculture and Commerce, shows the
index numbers for each of the three groups 27—food, clothing, and
materials— and for the three groups combined, by years, from 1900 to
1917. The indexes for the groups as a whole apparently were
obtained by taking the simple average of the index numbers for all
articles included in the three groups reported.
T able 77.—IN D E X NUM BERS FOR THE TH R EE PRINCIPAL GROUPS OF COMMODITIES,
1900 TO 1917.
(Average prices in 1900=100.)
Year.
1900.............................................................................................
1901..............................................................................................
1902............................................ ................................................
1903..............................................................................................
1904..............................................................................................
1905..............................................................................................
1906..............................................................................................
1907..............................................................................................
1908............................................................................... .............
1909..............................................................................................
1910............................................................................................
1911..............................................................................................
1912..............................................................................................
1913..............................................................................................
1914..............................................................................................
1915..............................................................................................
1916............................................................................. ...............
1917..............................................................................................
Food.
100
98
102
111
124
135
129
135
136
133
132
139
154
155
148
139
147
180
Clothing.
100
100
98
102
109
120
122
129
120
119
124
133
130
132
120
113
146
216
Materials.
100
95
92
94
95
100
101
110
113
107
110
114
119
121
118
122
138
183
All groups
combined.
100
97
97
101
108
116
114
122
122
118
120
126
133
135
129
127
143
188
27
For index numbers ofthe separate commodities of each group see the statistical reports of Department
of Agriculture and Commerce for 1911, 1912, and 1919.
33226°— 21— Bull. 284------- 20
806
INDEX NUMBERS OF WHOLESALE PRICES.
Table 78, taken from the thirty-fifth statistical report (p. 159),
gives index numbers for the five groups for the years 1912-1919.
T a b le
7 8 .—IN D E X NUM BERS B Y GROUPS OF COMMODITIES, 1912 TO 1919.
(Average prices in 1912=100.)
Year.
Grains.
1912..............................................................
1914......................................................................
1915......................................................................
1910......................................................................
1917......................................................................
1918......................................................................
1919......................................................................
Food.
Clothing.
Material.
100
100
100
100
102
86
100
100
100
101
92
89
74
77
108
169
211
97
100
111
116
146
190
200
156
272
99
117
162
229
314
281
Others.
100
103
99
96
111
151
219
262
Grand
average.
100
101
96
94
109
145
200
238
INDEX NUMBERS OF THE BANK OF JAPAN.
PU BLICATIO N A N D H ISTO R Y.
Since 1897 the Bank of Japan has published index numbers of
wholesale prices in Tokyo. These index numbers are published in
monthly bulletins and in yearbooks.
BASE PERIO D .
The year 1887 was used as a base period until February, 1913.
The base was then changed to October 1900.
SO U RCES.
The prices are taken from private sources, being supplied by mer
chants and associations in Tokyo and Yokohama.
N U M B E R AN D C LASS O F C O M M O D IT IE S .
Before 1900 the index numbers were based on the prices of 41 com
modities. In 1900 this number was increased to 56 and has not
been changed since. For a short time in 1904 the commodities
were divided into the three following classes:
1 . Commodities the prices of which were influenced by domestic
economic conditions.
2 . Commodities the prices of which were influenced by export
conditions.
3. Commodities the prices of which were influenced by import
conditions.
This method proved to be too cumbersome for the prompt issuance
of index numbers and was very soon abandoned. Later on another
group subdivision was adopted as follows: Cotton cloth, building
materials, papers, metals, fuels, silks (raw and waste), silk cloth,
foodstuffs, fertilizers, cereals, tobacco, and other goods. This
group plan was also abandoned later, since which relative prices for
the 56 commodities have been given individually.
JAPAK— BAISTK OF JAPAN,
307
DESCRIPTION OF COMMODITIES.
The following is the list of commodities on which the latest index
numbers of the Bank of Japan are based:
Rice.
Barley.
R ye (hadakamugi).
W h eat.
Soy beans.
R ed or white beans (azuki).
W h ea t flour.
Rice bran (fertilizer).
Fish fertilizer.
Oil cake.
Sugar.
Tea.
Salt.
Miso.
Soy.
Sake.
Dried bonito ( Jcatsuobushi).
Kerosene oil.
Charcoal.
Eggs.
Oils.
Tobacco.
Cigarettes.
Raw silk.
H abutai silk.
Silk handkerchiefs.
K aik i silk.
Silk for lining.
Floss silk.
Cotton yarns.
Im itation nankeens.
Cotton shirtings.
Cotton.
Ram ie and China grass.
Flannel.
Muslin.
Italian cloth.
Fuel.
Indigo.
Tim ber.
Steel, bars and rods.
Nails.
Copper.
Stone.
Bricks.
Tiles for roofing.
Portland cement.
Mattings.
Glass plates.
K am i (Japanese paper).
Paper.
Lacquer.
Vegetable wax.
Leather.
Matches.
Coal.
WEIGHTING.
The index numbers are unweighted, being the simple arithmetic
averages of the relative prices of individual commodities.
TABLES OF RESULTS.
The information now issued by the Bank of Japan in the form of
monthly sheets shows, besides the general index numbers for a period
of ten years, relative prices of principal articles for the preceeding 12
months. Relative prices of individual commodities for the month in
question, the month immediately preceding it and the same month
of the previous year are also given.
In the yearbook the relative prices of all commodities are given
individually by months for the years 1900 to date; also the monthly
eneral index numbers are given for the same period, as well as the
Iconomist figures. In addition a table showing the increase in the
cost of commodities during the war in Tokyo and in London is
presented.
Table 79, taken from the June, 1920, issue of the statement showing
“ Index numbers of average monthly wholesale prices in Tokyo,
published by the Bank of Japan, gives index numbers by months
from January, 1911, to June, 1920.
f
308
INDEX NUMBERS OF WHOLESALE PRICES.
T able
7 9 .—IN D E X NUMBERS OF W H O LESALE PRICES IN T O K Y O , JAN U A R Y, 1911, TO
DECEM BER, 1920.
(Average prices in October, 1900=100.)
Month.
1911
1912
1913
1914
January...........................................................
February........................................................
March.......................................... ...................
123
123
123
123
123
129
130
131
132
133
131
133
131
130
132
133
134
134
133
132
132
131
132
130
130
132
133
132
131
130
128
128
127
125
125
125
127
129
125
122
122
August.........................................................
September......................................................
October...........................................................
November......................................................
December.......................................................
1
124
128
127
126
127
1915
120
123
125
127
128
127
125
126
126
127
133
141
121
119
1916
1917
1918
1919
145
168
166
1.53
154
167
153
173
182
150
147
190
206
147
221
151
153 • 214
214
157
212
168
172
216
224
232
238
243
242
245
252
267
274
280
278
277
277
275
267
267
278
295
319
324
332
352
370
381
1920
398
414
425
397
359
327
1317
1311
1305
1299
1293
1272
From Bulletin of Supreme Economic Council, London, vol. 2, No. 4, 1921, p. 10.
Table 80, showing the trend of wholesale prices in Tokyo during
the war, is taken from the January, 1920, yearbook, page 45.
T able
8 0 .—IN D E X
NUMBERS
OF
W HOLESALE PRICES IN
DECEMBER, 1919.
T O K Y O , JULY, 1914, TO
(Average prices in July, 1914=100.)
Month.
1914
January..........................................................................................................
February......................................................................................................
March..............................................................................................................
April................................................................................................................
M ay.................................................................................................................
June................................................................................................................
July.................................................................................................................
August.............................
......................................................................
September....................................................................................................
October..........................................................................................................
November......................................................................................................
December.......................................................................................................
100
101
103
100
97
95
_______________________________
Average
1915
1916
1917
1918
116
178
185
190
194
193
195
106
113
125
134
137
132
132
133
138
145
151
164
176
170
170
170
173
100
112
132
96
98
122
100
123
101
102
101
100
100
100
101
122
120
117
117
120
122
1919
221
219
213
213
221
221
221
235
254
258
265
280
291
303
ISO
218
201
213
219
223
!
NETHERLANDS.
INDEX NUMBERS OF THE NETHERLANDS STATISTICAL OFFICE.
HISTORY AND PUBLICATION.
An index of wholesale prices prepared by the Netherlands Statis
tical Office appeared for the first time in the monthly journal of that
office in June, 1914.28 It grew out of a compilation of tables on
wholesale prices presented by the Statistical Office in its yearbook of
1913, and covered the years 1885 to 1913. It was continued through
1918, in the monthly journal of the same office, after which its pub
lication was discontinued. A new series was begun in the monthly
journal for August, 1920, and has been continued in subsequent issues.
SOURCE OF QUOTATIONS.
Price quotations were obtained from the different wholesale mar
kets, the number and place of these markets not being specifically
mentioned. Monthly price quotations were averaged annually.
23 Maandschrift van het Centraal Bureau voor de Statistiek.
461-466.
The Hague, 1914, vol. 9, No.
6
(June) np.
NETHERLANDS— STATISTICAL OFFICE.
309
BASE PERIOD.
The base period used for the original series was 1893, the average
price of that year representing 100 . In the more recent series the
years 1901-1910 form the base period.
NUMBER AND DESCRIPTION OF COMMODITIES.
There were 12 commodities for which separate index numbers, or
relative prices, were given; in the first published series no general
index for all commodities combined was presented. In the new series
published since August, 1920, the number of commodities has varied
from 48 to 50, relative prices being shown for individual commodi
ties and a geneial index for all articles combined. Both raw and
manufactured commodities are included.
WEIGHTING.
A simple arithmetical average apparently has been employed no,
mention being made of any kind oi weighting.
TABLE OF RESULTS.
The following table shows the index numbers of wholseale prices
for the Netherlands as now published in the monthly journal of the
Central Bureau of Statistics:
T able
8 1 .—IN D E X NUMBERS OF W H O L E SA L E PRICES, B Y Y E A R S , 1911 TO 1920, AND B Y
M ONTHS, JA N U A R Y TO DECEM BER, 1920.
[Sourcc: Maandschrift van het Centraal Bureau voor de Statistiek, Jan. 31, 1921, p. 128.]
(Average prices in 1901-1910=100.)
Year.
L-1910.......................................
1911L
19121
1913!
19141
1915
1916
1917r
1918J
1919)
............................
............................
............................
.............................
.............................
............................
)............................
General
index
(49
articles).
100
Index
of foods
(31
articles).
100
116
117
120
121
120
112
122
170
266
340
454
349
325
171
263
313
338
333290
114
Year and month.
1920.
January................
February..............
March...................
April.....................
May.......................
June......................
July.......................
August.................
September...........
October................
November............
December.............
General
Index
index
of foods
(49
(31
articles). articles).
334
329
331
338
339
339
343
330
328
323
297
266
285
283
291
293
301
307
297
292
290
283
260
NEW ZEALAND.
IN D E X N U M B E R S O F T H E C E N S U S A N D S T A T IS T IC S OFFICE.
PUBLICATION AND HISTORY.
During the year 1917 a system for the regular collection of infor
mation regarding wholesale prices was put into operation by the
Census and Statistics office of New Zealand. A considerable volume
of data was obtained from merchants and traders by means of retros
pective investigations covering the years 1891 to 1917, and it was
found possible to obtain sufficient information upon which to calculate
index numbers of wholesale prices over the period from 1891 on.
SIO
INDEX NUMBERS OF WHOLESALE PRICES.
These index numbers first appeared in the November, 1919, issue of
the Monthly Abstract of Statistics, and. since then have been pub
lished in every issue of this periodical. The yearly figures are repro
duced in the New Zealand Official Yearbook and go back as far as
1891.
BASE PERIOD.
The base for the computation of index numbers is the average of
the period 1909-1913, which is taken as 1,000.
SOURCE OF
QUOTATIONS.
The wholesale prices quoted are obtained from private sources—
merchants, traders, millers, live-stock brokers, etc., in various locali
ties. The index numbers, however, are based, only on the prices in
the four chief centers: Auckland, Wellington, Christchurch, and Dune
din. These cities include one-third of the Dominion’s entire popula
tion and about 70 per cent of the urban population.
PRICES: HOW SHOW N AND COMPUTED.
Average wholesale prices of commodities are given in the Monthly
Abstract of Statistics. These are shown in very great detail, by
localities. The official yearbook gives quarterly prices as well as
the average yearly prices, by commodities.
NUMBER AND CLASS OF COMMODITIES.
The general index number, covering all years since 1891, is based
upon the prices of 106 commodities. From 1909 on, more detailed
information was available and the group index numbers now com
puted represent the fluctuations of 140 articles or grades of com
modities. Both raw and manufactured commodities are included.
DESCRIPTION AND GROUPING OF COMMODITIES.
The 140 commodities used for the computation of the group
index numbers are divided as follows:
I. Agricultural produce.
II. Flour, bran, pollard, and oatmeal.
III. Wool, hides, tallow, butter, and cheese.
IV. General merchandise and crockery.
V. Building materials.
VI. Leather.
VII. Chemicals and manures.
V III. Coal.
The following is the list of articles:
Agricultural produce.
W heat: Tuscan, Hunters, velvet, fowl.
Oats: Garton’s A ; Garton’s B ; Sparrowbills.
Barley: M alting; feed.
Cocksfoot (M . D .).
R y e grass: Italian (M . !> .); perennial (M . D .) .
Potatoes.
Onions.
Maize.
NEW ZEALAND— CENSUS AND STATISTICS OFFICE.
311
Milling products.
Flour.
Bran.
Pollard.
Oatmeal.
Woolt hides, tallow, etc.
W ool:
Merino, m edium to good; merino, inferior.
H alf-bred, m edium to good; half-bred, inferior.
Crossbred, m edium to good; crossbred, inferior.
H id es:
O x, h eav y; ox, m edium ; ox, light.
Cow, heavy, m edium and light, vearling.
Calf.
Tallow.
Butter.
Cheese.
General merchandise.
Tinned fruits: Apricots, 2J-pound tins; peaches, 2J-pound tins; pears, 2|-pound tins.
Dried fruits: Currants, sultanas, prunes.
Herrings; salm on; sardines; coffee; cocoa; tea; sugar, No. 1 A ; golden syrup, 2-pound
tins; treacle; sago; tapioca; rice; salt, com m on; table, 7-pound bags; cream of
tartar; carbonate of soda; vinegar; mustard; pepper; starch; soap (N ew Zea
land); bluing; candles, w ax; tobacco, plug (d u ty p aid ); kerosene; ham ; bacon;
lard; honey; bran bags; corn sacks; wool packs.
Ironmongery.
Iron: Pig; bar; angle and T ; plate; hoop; corrugated galvanized.
Fencing wire, No. 8, galvanized; black.
Nails, 4-inch wire.
Z in c, sheet.
Lead, sheet.
Tinned plates, I . C. coke.
Copper, sheet.
L im e (Auckland and W ellington prices are for hydraulic lim e; Christchurch and
Dunedin for white lim e).
W h ite lead.
Linseed oil, raw.
Bricks, ordinary building.
Textiles.
F o x ’s No. 4 serge; blankets, colonial all wool (qu ality most sold), size 10/4; S. S. S.
galatea shirting (Prestwick’s), 30-inch; F in la y ’s F . T . sheeting, 80-inch; Crewdson’s No. 2 calico, 36-inch; N airn’s linoleum , second quality, 6 feet w ide; W il
liam son’s linoleums, third quality, 6 feet wide.
Leather.
Sole: N ew Zealand— Crop, 1 8-20 lb .; bends, first quality. Sole, imported.
K ip , N ew Zealand; k ip, im ported; split kip, New Zealand.
Chrome calf, New Zealand; chrome calf, imported (E . or A .).
Chrome sides, New Zealand; chrome sides, imported.
Glace kid, New Zealand; glace kid, imported (European or Am erican; Australian).
Crockery.
English china cups and saucers:
London W . <fe G.
Norwich W . & G.
Worcester shape, gold edge, London size.
Worcester shape, gold edge, Irish size. ^
Worcester shape, gold edge, breakfast size.
W h ite granite and semiporcelain dinnerware, 10-in. plate.
Wrliite granite toilet sets, six pieces, size 9.
Tum blers, £ quart, plain heavy bottom.
312
INDEX NUMBERS OF WHOLESALE PRICES.
Fresh fr u it , etc.
Apples: Eating; cooking.
Oranges: Fiji and Island; Sydney.
Peaches: Eating; cooking.
Pears: Eating; cooking.
Bananas.
Appricots: Eating; cooking.
Cabbage; cauliflower; onions; eggs, fresh, preserved.
Coal (at wharf).
New Zealand: Screened; unscreened; small.
New Castle: Screened.
Timber.
New Zealand timbers:
Kauri, heart, first grade.
Kauri, heart, second grade.
Kauri, heart, third grade.
Rimu: Clean heart; heart (bridge quality); ordinary building; second class.
Matai: Heart; ordinary building.
Totara: Clean heart; heart (bridge quality); ordinary building; second class.
Southland beech or birch: Black birch or beech (fagus fusca ); white pine ( kahi-
katea).
Imported timbers:
Oregon, merchantable, saWn; Oregon, select, sawn; Jarrah, sawn; Jarrah, hewn.
Tasmanian hardwoods, sawn.
Ironbark; hew n ; sawn.
Tasmanian palings, split, 6 ft.; sawn, 6 ft.
Jarrah palings, sawn, 6 ft.
Chemicals.
Acid: Boracic (20-Mule Team); carbolic, crude; tartaric.
A lu m , bluestone (M acclesfield); borax.
Bottles: Glass, empty, quart beers; 8-oz. dispensing. Caustic soda; Cod-liver oil,
Norwegian, nonfreezing; cottonseed oil; cotton wool, cheap English; E n o’s fruit
salts; Epsom salts; glycerin (1.260 S. G .); Jeyes’s fluid; naphtha, wood; rectified
spirit, 90 per cent (65 O. P .); resin; saltpeter; soda bicarbonate, B . M .; soda
hyposulphite; sugar of milk, B . P .; sulphide of sodium, solid (concentrated);
sulphur, ground; washing soda.
Manures.
Superphosphate, 39-43 per cent; superphosphate, 36-38 per cent.
Guano: 60 per cent analysis; 40 per cent analysis.
Basic slag.
Kainit; pure bone dust; nitrate of soda; nitrate of lime; sulphate of ammonia; sul
phate of potash.
WEIGHTING.
The system of weighting is the one known as the “ aggregateexpenditure” method. Each article has been given a weight corre
sponding to its average annual prewar consumption in the Dominion
and, further, to its price.
TABLE OF RESULTS.
The monthly, quarterly, and yearly figures appear in every issue
of the Monthly Abstract of Statistics. The Official Year Book
gives, in addition to the above, a table for yearly index numbers
for the period from 1891 on.
Tables 82, 83, and 84 are taken from the May, 1920, and February,
1921, issues of the Monthly Abstract of Statistics and Table 85 from
the Official Yearbook, 1919, page 806.
NEW
ZEALAND-----CENSUS
AND
STATISTICS
OFFICE.
313
8 2 .—M O N TH LY IN D E X NUM BERS OF W H O L ESAL E PRICES IN N E W ZE A L A N D
(A V E R A G E OF T H E FOUR CHIEF CENTERS) B Y GROUPS OF COMMODITIES, JANU
A R Y , 1918, TO DECEM BER, 1920.
T able
(Average annual aggregate expenditure of the four chief centers, 1909-1913=1,000.)
Month.
Group
Group
III:
Group
II:
Wool,
I: Agri Flour,
hides,
cultural bran,
tallow,
pollard, butter,
prod
uce,
and oat
and
meal.
cheese.
1918.
January.....................
February...................
March........................
April..........................
May............................
June...........................
July............................
August......................
September— ..........
October.....................
November.................
December..................
1919.
January.....................
February...................
March........................
April..........................
M ay............................
June...........................
August......................
September................
October.....................
November.................
December..................
1920.
January.....................
February...................
March........................
April..........................
May............................
June...........................
July............................
August......................
September................
October.....................
November.................
December.................
Group
IV :
Group
Group
V II:
General
Group
V:
Chemi
mer
Build
V I:
chan ing ma Leather.
cals
and ma
dise and terials.
crock
nures.
ery.
1, 585
1, 556
1,604
1,702
1, 720
1, 726
1, 784
1, 931
1,992
2,077
2,197
2,269
1,456
1,464
1, 504
1, 553
1, 558
1,561
1, 566
1, 574
1, 580
1, 588
1, 587
1, 586
1,468
1, 471
1, 471
1,466
1, 459
1,461
1, 462
1,462
1,464
1, 465
1,468
1,467
1, 574
1, 587
1,605
1,642
1,663
1, 690
1,718
1, 729
1, 731
1, 750
1, 760
1, 757
1, 996
1, 903
1,719
1, 698
1, 732
1, 754
1 , 812
1,923
1,961
1, 975
1,961
1, 986
1, 585
1, 584
1, 550
1,534
1, 524
1, 517
1, 519
1, 517
1, 512
1,509
1, 519
1, 533
1,468
1,467
1,463
1, 462
1, 476
1, 502
1, 537
1, 552
1, 542
1,567
1, 570
1, 576
1, 747
1, 736
1, 722
1,713
1, 705
1, 711
1,756
1, 797
1,869
1,890
1,930
1, 979
2,185
2,130
2, 083
1, 548
1, 560
1,718
1, 722
1, 724
1,736
1,746
1, 756
1,764
1.761
1, 755
1,747
1, 586
1, 634
2,083
2,174
2, 224
2, 243
2,227
2,206
2, 470
2,498
2,493
2, 509
2, 494
2, 459
2 ,1 0 0
2,055
2,015
1,983
1,939
1,931
1, 855
1,740
1,828
1 ,6 8 8
1,650
1,610
1,546
1,532
1,560
1,582
1,907
1, 770
1,750
2 ,0 1 1
1,992
2,019
2,044
2, 076
2, 111
2,158
2 ,2 2 1
2,247
2, 277
2,324
2,299
Group
V III:
Coal.
1, 895
1,891
1, 912
1, 909
1, 906
1, 907
1,891
1, 885
1, 897
1, 884
1, 910
1, 916
1, 757
1 , 822
1, 873
1,901
1, 988
2,006
2,031
2,089
2 , 088
2,098
2,052
2,052
1,460
1,463
1,462
1,462
1,463
1,463
1, 463
1,463
1,463
1, 516
1,529
1,529
1, 677
1, 676
1, 697
1, 741
1, 759
1, 775
1,808
1,860
1, 879
1, 917
1, 955
1, 961
1, 88 6
1, 986
1, 988
1, 957
1, 938
1, 931
1, 8 8 8
1,848
1, 789
1, 770
1, 760
1, 762
1,788
1 , 828
1, 8 6 8
1, 999
2 , 039
2,123
2,153
2,167
2,158
2,264
2,261
2,267
2,291
2 , 247
2,249
2,240
2,186
2,119
2,094
2, 041
2 , 018
1, 989
1,996
2 , 016
2, 022
2,034
2,050
1,883
1,892
1,902
1,902
1,941
2,027
2,128
2, 196
2 , 282
2,345
2,405
1, 899
1,924
1,925
1, 958
2,042
2,053
1,529
1,529
1,546
1,546
1, 551
1,574
1, 649
1,664
1, 750
1,807
1,809
1,809
2,078
2,181
310
2,338
2, 391
2,405
2,584
2, 587
2, 585
2,605
2,603
2,612
2,441
2, 538
3, 279
3, 476
3,437
3, 368
3,145
2, 901
2,890
2, 897
2, 716
2, 592
2,152
2,164
2,176
2, 229
1, 435
2,463
2, 528
2, 489
2, 501
2,494
2, 426
2, 368
1 , 818
1, 849
1, 935
1,995
2,066
2,092
2,096
2,108
2,153
2,158
2,170
2,178
2,
All
groups
com
bined.
1 ,8 8 8
1 ,8 8 6
1,899
1, 925
INDEX NUMBERS OF WHOLESALE PRICES.
814
f
$ 3 .— Q U A R T E R L Y IN D E X NUM BERS OF W H O L E SA L E PRICES IN N E W ZE A L A N D
(A V E R A G E OF TH E FOUR CH IEF CENTERS), B Y GROUPS OF COMMODITIES, JAN
U A R Y , 1914, TO DECEM BER, 1920.
ABLE
(Average annual aggregate expenditure of the four chief centers, 1901-1913= 1,00$.)
Quarter ending—
Group
Group
131:
II:
Group
Wool,
I. Agri Flour,
hides,
bran,
cultural
pollard, tallow,
prod
butter,
uce.
and oat
and
meal.
cheese.
Group
IV :
Group
Group
V II:
General
Group Chemi
V:
mer
Build
VI:
cals
chan ing ma Leather.
and ma
dise and terials.
crock
nures.
ery.
Grouo
VHl'i
Coal.
All
groups
com
bined.
1914.
March........................
June...........................
September................
December___ *.........
907
969
1,033
1,176
990
1,031
1,131
1,326
1,123
1,159
1,115
1,067
1,078
1,077
1,077
1,125
1,103
1,109
1,123
1,144
1,172
1,180
1,187
1,198
1,064
1,077
1,080
1,082
1,003
1,003
1,003
1,008
1,045
1,067
1,073
1,123
1615.
March........................
June..........................
September................
December..................
1,542
1,621
1,641
1, 517
1,589
1,727
1,605
1,298
1,196
1,305
1, 371
1, 315
1,175
1,186
1 , 216
1,229
1,148
1,192
1,256
1, 273
1,286
1, 349
1,368
1, 388
1,160
1,195
1,244
1,272
1,016
1,014
1,014
1,031
1, 2 2 1
1916.
March........................
June...........................
September................
December..................
1, 480
1, 305
1,461
1,702
1,361
1,273
1, 306
1, 300
1, 337
1, 402
1,436
1, 430
1,271
1,312
1, 340
1, 345
1, 334
1, 448
1, 485
1,508
1,440
1, 458
1, 472
1, 510
1, 375
1, 417
1, 494
1,495
1,082
1,118
1,175
1, 204
1,
1917.
March........................
June...........................
September................
December..................
1, 499
1,503
1,535
1,529
1, 419
1,447
1, 459
1, 436
1, 431
1,452
1,468
1, 512
1,367
1,407
1,463
1, 551
1,564
1,676
1, 838
2,009
1,676
1, 795
1, 853
1,901
1, 574
1,650
1,719
1,753
1, 267
1, 330
1, 432
1,445
1, 450
1, 510
1,593
1,665
1918.
March........................
June..... .....................
September................
December..................
1, 582
1,716
1, 902
2,181
1, 475
1, 557
1, 573
1, 587
1, 470
1,462
1,463
1,467
1,592
1,665
1, 727
1,756
2,007
2,077
2,209
2,300
1, 899
1, 907
1,891
1, 903
1, 817
1,965
2, 070
2,071
1,461
1,462
1,463
1, 525
1,684
1, 759
1,849
1, 944
1919.
March........................
June..... .....................
September................
December..................
1,873
1,727
1, 898
1,974
1, 573
1,525
1, 516
1, 521
1,466
1, 480
1, 544
1,571
2,181
2,050
1, 977
1, 919
1,916
1,535
1, 557
1 ,6 8 8
2,035
1, 887
1,915
2,117
2, 344
2 ,0 2 1
1, 809
1, 842
1,764
1 , 828
1,903
1920.
March............. ..........
June............ , ............
September................
December.................
2,132
2,057
1,951
1,808
1,609
1,727
1, 755
1, 754
1,636
1,602
1,558
1,809
2,190
2,379
2, 585
2,607
2,753
3,427
2,979
2,735
2,164
2,376
2,506
2,429
1, 867
2,051
2,119
2,169
2,054
2,160
2, 264
2,262
1, 735
1, 709
808
1,933
1,
2,160
2,225
2,487
2,487
2 ,0 0 0
1, 207
1,304
1, 284
323
1,314
1,403
1, 449
8 4 .—IN D E X NUM BER S, OF W H O L E SA L E PRICES IN N E W ZE AL AN D (A V E R A G E
OF T H E FOUR CHIEF CEN TER S), B Y GROUPS OF COMMODITIES, 1910 TO 1919.
T iB L E
(Average annual aggregate expenditure of the four chief centers, 1909-1913=1,000.)
Year.
Group
Group
III:*
11:
Group
Wool,
I: Agri Flour,
hides,
cultural bran,
tallow,
pollard, butter,
prod
and oat
uce.
and
meal.
cheese.
1910.............................
I 9 il.............................i
1912.............................!
1913.............................
1914.............................
1915.............................
1916............................
1917.............................
1918.............................
1919.............................
1920............................:
)
1 ,0 2 1
991
1,089
957
1, 021
1, 580
1, 487
1, 517
1, 845
1, 8 6 8
1,987
989
972
970
981
1, 555
1,310
1, 440
1, 548
1, 534
1,711
981
1 . 008
i;029
1,047
1,116
1,297
1,401
1,466
1,466
1, 515
1,651
I
Group
IV :
1 Group
Group
VII:
General
Group Chemi
mer
Build
V
I:
chan ing ma
cals
dise and terials. Leather. and ma
crock
nures.
ery.
969
993
1, 053
1, 055
1,089
957
995
1,037
1,063
1 ,2 0 2
1, 217
1, 444
1, 772
2,148
2, 067
2,440
1, 317
1, 447
1,685
1, 796
2,340
1 ,1 2 0
986
982
993
1,126
1,184
1, 348
1,470
1,806
1,900
2 , 066
2,974
994
986
1,024
1,035
1,076
1 , 218
1,445
1,674
1, 981
1, 958
2,309
Group
V II I:
Coal.
985
985
1 ,0 1 1
1,038
1,004
1,019
1,145
1, 369
1, 478
1,647
2,052
All
groups
combiued.
983
994
1,041
1,032
1,077
1, 269
1, 380
1, 555
1,809
1, 834
2,185
NEW
Z E A L A N D -----C E N S U S
AND
STAT ISTIC S
01T IC E .
315
T able 8 5 .— G EN ER AL IN D E X NUM BERS OF W H O L E SA L E PRICES IN N E W ZE A L A N D
(A V E R A G E OF T H E FOUR CHIEF CENTERS), 1891 TO 1920.
(Average aggregate annual expenditure, four chief centers, 1900-1913=1,080*)
Index
num
ber.
Year.
1891
1892
1893
1894
1395
18®$
1897
1898
1899
1900
1
:
'
|
;
i
j
991
937
925
927
920
943
1,005
982
893
917
Year.
1901......................................
1902......................................
1903........ ‘ .........................
1904......................................
1905.....................................
1906.....................................
1907......................................
1 9 0 8 ..................................
1909......................................
1910.................................
Index
num
ber.
931
975
954
922
994
1,028
1,016
1,006
949
984
Year.
1911.....................................
1912.....................................
1913.....................................
1914.....................................
1915.....................................
1916....................................
1917.....................................
191S..................................
1919.....................................
1920............................
Index
num
ber.
994
1,041
1 ,051
1.077
1,299
1,380
1, 555
1,809
1,834
12,185
From Monthly Abstract of Statistics, Wellington, February, 1921, p. 30.
INDEX NUMBERS OF JAMES W. McILRAITH.
PUBLICATION.
This index, which is based on the wholesale prices of certain
important articles in New Zealand from 1861 to 1910, was published
by the Government of that Dominion in 1911 in “ The Course of
Prices in New Zealand,” by James W. Mdlraith. It is stated in the
introduction to the volume that the author intends to continue the
tables from year to year, the results to appear annually in the “ New
Zealand Official Yearbook.” 29
HISTORY.
The report is the result of a post-graduate research in economics at
Canterbury College. It has two main objects: ( 1 ) To measure the
changes in the general level of prices, year by year, since 1860; and
(2 ) to attempt to ascertain the causes of the changes in the local
price level. In the absence of any official index numbers for New
Zealand the study was undertaken in the hope that it would “ help
all engaged in the solution of those practical problems of social life
in which the changes in the purchasing power of money are an
important factor.” 30
SOURCE OF QUOTATIONS.
The author states that prices for all imported goods and for a few
colonial products have been obtained at Wellington, the geographical
and political center of New Zealand, while for all the cereals (including
flour) and for pastoral products he has used prices at Christchurch, a
city in the heart of the agricultural district of New Zealand. He
further states, however, that while it seemed inadvisable to use
Wellington prices for certain articles for certain years and then to use
the prices from some other city for the same articles for other years,
the nature of things demanded that the principle of continuity must
be subordinated to that of accessibility. The data contained in the
89 An examination of the yearbooks for 1911,1912,1913, and 1919, however, fails to disclose any data bear,
ing on this subj ec t.
so The Course of Prices in New Zealand, James W . Mdlraith, x>. 3.
316
INDEX NUM BERS OF WHOLESALE PRICES.
report were secured mainly from Wellington and Christchurch news
papers: “ The New Zealand Trade Review and Price Current”
(Wellington), “ The Press” (Christchurch), and “ The Lyttleton
Times” (Christchurch) being the principal sources of information.
BASE PERIOD.
The base or standard period is the decade 1890-1899. The author
at first felt inclined to use 1867-1877, because this would have enabled
him to compare his figures more directly with those of Sauerbeck,
but he decided that the period was too early in the development of
New Zealand and that the data were insufficient. He found the
decade chosen a period of comparatively stable prices, the average
being almost identical with that for the 20 years 1886-1905.
PRICES: H O W SHOW N AND COMPUTED.
Wholesale prices for the articles upon which the index number is
based were taken quarterly, during the first week of January, April,
July, and October, or as near as possible to those dates. The simple
average of the four prices was taken as the average price for the year.
From these averages the simple average for 1890-1899 was com
puted and was taken as the base, or 100 , the prices for each of the
years included in the report being reduced to percentages of this
figure. For each year the percentages representing the prices of the
several articles were then added together and the result divided by
the number of articles. The resulting figure is the general index
number for the year.
NUMBER AND CLASS OF COMMODITIES.
The index number from 1887 to 1910 is based on the prices of 45
commodities, both raw and manufactured. Previous to 1887, accord
ing to the tables, the number of articles included was smaller and
varied from time to time, being lowest in 1861 and 1862, when only
33 articles were represented. Since 1875 the index number has in
every year been based on 41 or more commodities. The author states
that his aim has been to make the list include, as far as data were
available, the articles of the greatest importance in the trade of the
country. He considers his selection of commodities superior to the
selections of Sauerbeck and the Economist, because m his judg
ment these indexes contain undue proportions of raw materials.
DESCRIPTION AND GROUPING OF COMMODITIES.
Concerning nearly all of the 45 articles the statement is made that
only the best grade is quoted. Further description of the articles, as
far as given, appears in the following list which classifies the 45
commodities under 8 heads:
Agricultural products (5 articles).
W heat, best on Christchurch market.
Flour, best brands of N ew Zealand roller
flour.
B arley, best on Christchurch market.
Oats, best on Christchurch market.
Oatmeal, first-class N ew Zealand manu
facture.
N E W ZEALAND---- JAMES W . M ILBAITH.
31T
Pastoral products {8 articles).
W ool, best merino, greasy.31
W ool, best half-bred, greasy.31
Beef, best quality, Addington (Christ
church) sales yard.
Mutton, carcasses exported.32
L am b, carcasses exported.32
Bacon, b est N ew Zealand product.
Cheese, best N ew Zealand product.
Butter, best N ew Zealand product..
Liquors (5 articles).
Beer (ale), Bass’ s “ Dog’ s H e a d .”
W h isk y , Teacher’ s (in b on d ).
Port, Superior (in bond).
Claret (in bond).
Brandy, H en n essy’ s (in bond).
Beverages (3 articles).
Tea, Congou, fine
Coffee, ground.
( Cocoa, V an H o u ten ’ s.
Oils (3 articles).
Kerosene
Linseed oil (boiled).
) Castor oil.
Minerals (6 articles).
Iron,
galvanized,
“ O rb”
brand
(26
Iron, bar.
Lead-, sheet.
Materials
Cement, Portland
Soap, N ew Zealand.
M atches, plaids
Candles, Price’s London Sperm.
Z in c.
WTire, black fencing, No. 8.
Coal, New castle (N ew South W ales), on
ship.
(7 articles).
H ops, Nelson (New Zealand).33
Soda, carbonate.
Soda crystals.
Other foods (8 articles).
Sugar, A u ckland (N ew Z ealand), refined
No. I . 34
Currants.
Sultanas (raisins).
R ice.
Salt, Liverpool, fine.
Salmon, in tins.
Pepper, white.
SUBSTITUTIONS AND ADDITIONS.
The substitution of one description of an article for another
description can not, except in a few cases, be traced in the report,
owing to the vagueness of the descriptive terms employed. It
appears quite probable, however, from tne extensive period of time
covered by the tables that a considerable number of such substitu
tions must have been made. In a number of instances additions to
certain groups of commodities have been made since the initial year
of the series. In such cases the index numbers for the added articles
have been carried into the total of index numbers for the year and
this total divided by the number of articles to obtain the level of
prices for that year.
£! The quotations are prices of New Zealand wool at the London wool sales, minus freight from New Zea
land, as reported by the London agents of H. Watson & Co., one of the principal wool-broking firms in
New Zealand.
Prices are deduced from estimated values of carcasses exported, as published in the “ Statistics of New
Zealand.”
S3 This kind is quoted in later decades; Kentish is quoted in earlier periods.
Best imported sugar was quoted prior to the opening of the Auckland refinery.
31 &
INDEX NUMBERS OF WHOLESALE PRICES.
INTERPOLATION.
Much difficulty was experienced in securing data for the earlier
years covered by the report, and for years prior to 1875 there was an
occasional quotation lacking which the author deemed it necessary
to supply by an interpolated figure. He thus describes his method
of interpolation.
“ I examined the price of the particular commodity in preceding
and succeeding years. If those prices showed a continuous rise or
fall, that was strong evidence of the probability of the missing price
being one of an uninterrupted series. To test this probability, I
referred to similar products which would most likely vary in price
in the same manner as the commodity whose price was missing; and
if the price movement in these commodities was in the same direction
as the movement in the latter one, I presumed that the missing price
would most probably vary in the same manner as the price of the
similar articles in the same year. This method was applied chiefly
where there was a causal connection between the fluctuations in the
prices of the two articles, e. g., where both were produced from the
same raw material (as galvanized iron and bar iron), or where one
was raw material and the other the finished product (as wheat and
flour), or where the production of both would be affected by the
same causes, as by drought (in the case of wheat and oats, mutton
and beef) . ” 35
All index numbers based on interpolated prices are inclosed in
brackets.
WEIGHTING.
The author does not attempt to assign definite weights to all
commodities. He has, however, quoted prices on more than one
variety of certain articles considered as of great importance. He
illustrates as follows: “ Thus, I have taken three cereals, and to
emphasize the importance of wheat I have taken flour as well. I
have sought to give wool adequate representation by including two
kinds, greasy merino and greasy half-bred. Iron is represented by
bar iron and galvanized iron, while meat is represented in a similar
manner by beef, mutton, lamb, and bacon . 7' 36
TESTING.
The tables of the report compare the index numbers derived by
the author with—
( 1 ) A series of index numbers based on the median instead of the
simple arithmetic mean.
(2 ) The figures of the Economist and Sauerbeck reduced to the
New Zealand basis: Annual average prices 1890-1899 = 100 .
(3 ) Index numbers indicating the local movements in certain
related phenomena, as, for example, the marriage rate, the bank
ruptcy rate, world gold production, and New Zealand gold production.
The third set of comparisons is made by single years; the first and
second are made both by single years and by decades.
35 The Course of Prices in New Zealand, James W . Mcllraith, p. 29.
36 Idem, p. 23.
N E W ZEALAND— JAMES \V. M ’ iLRAITH .
319
TABLES OF RESULTS.
The first eight tables of the report show for the several groups of
commodities and for each year the index numbers by articles, and
also the prices from which the index numbers were computed.
The next table summarizes the data contained in the preceding
tables by showing the index number for the total of all articles.
Other tables compare the New Zealand results with the figures
of Sauerbeck and the Economist ( 1 ) in their original form, and
(2 ) reduced to the New Zealand basis: Average annual prices 18901899 = 100.^
Tables similar to these show index numbers for decades (e. g.,
1861-1870, 1862-1871, 1863-1872, etc.) instead of for single years.
In the two succeeding tables are shown the New Zealand index
numbers for farm products and for nonfarm products (1) by single
years, and (2) by decades.
The remaining tables show for New Zealand ( 1 ) an annual index
number of prices based on the volume of foreign trade and the tonnage
of vessels carrying such trade, (2 ) an annual index number of prices
based on the volume of foreign trade per head of population, (3) a
comparison of index numbers of prices (a) of farm products and (6 )
of all commodities with the index numbers of the marriage rate and
the bankruptcy rate, and (4) the index numbers (a) of the gold pro
duction of the world, (6) the gold production of New Zealand, and (c)
the price of wheat. The last table also gives the New Zealand wheat
yield per acre, in bushels.
The table showing the comparison of the index numbers for New
Zealand derived by the author with those of Sauerbeck and the
Economist reduced to the same base period, and also with index
numbers based on the median, instead of the arithmetic means, is
reproduced herewith as Table 86.
TABLE
8 6 .—
Year.
1861.........
1862.........
1863.........
1864 ........
186 5 ........
1866.........
1867.........
1868.........
1869 ____
1870 .......
1871 . . . .
1872 .......
1873 .......
1874 .......
1875 .. .
1876 ____
1877
1878
1879
1880..........
1881.........
1882.........
1883.........
1884 . ,
1885 .. .
COMPARISONS OF IN D E X NUMBERS COMPUTED UPON THE SAME BASE
PERIOD.
Sauer New Zea
New Zea “ Econo
beck,
land
land,
m ist/’
(by
1890-1899 1890-1899 1890-1899
= 1 0 0 .0 . median).
= 1 0 0 .0 .
= 1 0 0 .0 .
187
184
164
154
150
154
164
161
148
140
144
135
127
130
125
123
118
115
133
141
171
185
175
174
149
131
130
131
127
139
144
140
135
132
132
124
109
124
116
119
114
109
111
102
184
186
193
195
189
200
149
153
156
159
153
155
152
150
149
146
152
165
168
155
146
144
142
132
126
133
129
127
124
115
109
172
172
176
185
183
187
192
189
163
148
137
154
152
160
148
134
148
133
125
129
125
122
118
113
110
Year.
1886..........
1887.........
1888.........
1889
1890
1891
1892..........
1893
1894
1895
1 8 9 6 ......
1897
1898 ,.
1899
1900 ..
1901 ,
1902
1903 ____
1904.........
1905
190S.........
1907
1908
1909
1910.........
Sauer New Zea
New Zea “ Econo
land,
beck,
mist, ”
land
1890-1899 1890-1899 1890-1899
(by
= 1 0 0 .0 .
= 1 0 0 .0 . median).
- 1 0 0 .0 .
108
103
103
111
107
108
104
100
98
93
96
9797 I
98
10 1
98
100
100
95
98
10 1
107
104
101
103
94
98
95
93
94
104
104
96
98
106
108
116
105
103
106
109
109
109
103
96
96
94
92
94
U
103
114
106
105
105
106
109
116
121
121
109
no
111
112
99
101
109
106
110
110
104
103
102
118
107
102
10
i
111
108
108
102
100
100
98
98
98
96
90
99
98
100
100
95
99
100
101
101
104
101
320
IN D E X N U M B E R S
OF W H O L E S A L E PRICES.
The author has this to say concerning the foregoing table: “ A
general review of this table shows a marked similarity in the move
ments of columns 1 and 3. The fluctuations in both columns are
arallel, i. e., a rise or fall in both is synchronous. The index numers are fairly even in both columns, and particularly so since 1872,
the numbers in column 4— those calculated by the median— being on
the whole slightly lower than those in column 1 , thus indicating that
exceptional variations have been of the nature of a rise more often
than of a fall.”
NORWAY.
E
INDEX NUMBERS OF EINAR RUUD.
PUBLICATION AND HISTORY.
This index of wholesale prices, based on the prices of imports into
Norway, was first published in 1911 in the official journal of the Nor
wegian Labor Office 37 No. 9-10 of that year. It covers a period from
1880 to 1910, but whether or not it has since been continued is not
known, although the prices on which it is based are found for later
ears in the annual summary of commercial statistics issued by the
Norwegian office of general statistics.38
S
SOURCE OF QUOTATIONS.
The summary of commercial and customs statistics alluded to pre
sents the average annual prices of some 135 different commodities
imported into the Kingdom each year, and from this list the compiler
of the index selected 39 articles for inclusion in his series. The
prices of these articles are obtained from a number of manufacturers
and wholesalers in different parts of the Kingdom.
BASE PERIOD.
The base period chosen extends from 1891 to 1900. This period
was selected, it is explained, because it contained both a rising and
a falling tendency of prices, not only in Norway but also in foreign
countries, and therefore, when taken as a whole, may be regarded as
representing average market conditions.
NUMBER AND CLASS OF COMMODITIES.
The 39 articles chosen from the larger list of 135 imports are divided
into five groups: ( 1 ) Food commodities; (2 ) grains and flours; (3)
imports from the East; (4) manufactured products; (5) miscellaneous.
Group I includes beef, pork, cheese, oleomargarine, eggs, and pota
toes; Group II, barley, oats, wheat, rye, also hulled and prepared bar
ley and oats, and flour or meal ground .from the above grains; Group
III includes coffee, tea, sugar (two kinds, loaf and granulated),
tobacco (smoking and chewing), and rice; Group IV is made up of
wool yarn (plain and dyed), cotton yarn (single and several twist),
*7 Sociale meddelelser (fortsaettelse av maanedsskrift for socialstatistik) utgit av Socialavdelingen und er
Departementet for sociale saker, handel, industri og fiskeri. Christiania, lste aargang, Nr. 9-10, 1911,
pp. 136-149.
88 Norges handel, 1908-(Statistique du commerce de la Norvege pendant Fannee 1908-) Utgit av det
Statistiske centralhureau. (Norges officielle statistik, V, 87,116.)
N OEW AY---- EINAR RUUD.
321
cotton cloth or goods (three kinds, printed, dyed and bleached, and
unbleached); and Group V includes petroleum, coal and coke, dressed
skins or leather, pig iron, steel, bar and hoop iron, zinc, lead, and tin.
WEIGHTING.
In the construction of the index a simple arithmetic average is
employed, there being no weighting. The total index is the mean of
the indexes for each of the five groups.
TABLE OF RESULTS.
Table 87, which appears in the journal of the Norwegian Labor
Office for September and October, 1911 (No. 9-10 ) 39 shows in detail
the main results of this series of index numbers, as published.
T able
8 7 __ IN D E X NUM BERS OF W H O LESALE PRICES.
(Based on 7 food commodities imported into Norway, 1880-1910.
Year.
Beef
Pork.
Cheese.
Average prices in 1891-1900=100,)
Butter.
Oleo
marga
rine.
All 7
com
modi
ties.
Eggs.
Pota
toes.
97
97
150
140
133
117
100
100
100
100
107
115
129
124
115
105
95
97
99
98
1880..............................................
1881..............................................
1882..............................................
1883..............................................
1884..............................................
1885..............................................
1886..............................................
1887..............................................
1888..............................................
1889..............................................
141
138
151
154
136
121
103
92
90
90
124
129
149
134
112
102
98
110
122
107
103
113
98
94
96
87
87
96
96
100
98
98
96
101
71
86
97
97
97
97
86
86
91
91
1891..............................................
1892..............................................
1893..............................................
1894..............................................
1895..............................................
1896..............................................
1897............................. ...............
1898..............................................
1899..............................................
90
95
92
100
9595
92
95
105
108
97
98
112
136
115
95
75
76
93
86
104
109
104
104
96
91
96
96
100
100
98
105
101
105
98
97
91
91
105
105
101
112
112
112
107
96
90
84
92
96
91
91
97
97
97
97
102
102
102
102
100
100
100
100
100
100
100
100
100
100
97
101
103
108
101
96
92
92
100
100
1900..............................................
1901..............................................
1902..............................................
1903..............................................
1904..............................................
1905..............................................
1906..............................................
1907..............................................
1908..............................................
1909..............................................
113
113
118
110
103
115
128
133
123
123
115
131
153
132
114
119
136
136
132
166
100
100
104
104
104
104
113
122
113
113
105
105
105
105
105
108
105
108
115
115
96
96
96
96
96
96
101
101
112
112
108
108
108
108
108
108
108
108
108
108
100
83
83
83
92
88
72
88
111
97
105
105
110
105
103
105
109
114
116
119
1910..............................................
131
180
113
129
112
108
83
122
89 Maanedsskrift for socialstatistik.
pp. 141-148.
Utgit av det Statistiske Centralbyraa, lste aargang, Nr. 9-10,1911,
3 32 2 6°— 21— B u ll. 284------- 21
INDEX NUMBERS OF WHOLESALE PRICES.
322
T able
8 7 .—IN D E X NU M BER S OF W H O L E S A L E PRICES—Continued.
(Based on 10 grains and flours imported into Norway, 1880-1910.
Year.
Bar
ley.
1880...........................................
1881...........................................
1882...........................................
1883...........................................
1884...........................................
1885...........................................
1886...........................................
1887......................................... :
1888...........................................
1889...........................................
1890...........................................
1891...........................................
1892...........................................
1893...........................................
1894...........................................
1895...........................................
1896...........................................
1897...........................................
1898...........................................
1899...........................................
1900...........................................
1901...........................................
1902...........................................
1903...........................................
1904...........................................
1905...........................................
1906...........................................
1907...........................................
1908...........................................
1909...........................................
1910...........................................
153
157
137
134
132
119
106
97
102
109
114
134
109
94
78
86
87
87
101
109
115
108
108
98
97
105
109
131
134
126
111
Average prices in 1891-1900=100.)
Pot
All
or
Bar Rye
10
Oat Wheat
Oats. Wheat. Rye. pearl Hulled ley
com
oats. flour. flour. meal. flour.
bar
mod
ley.
ities.
126
136
126
110
105
105
99
84
89
110
99
115
115
105
89
76
89
103
105
102
102
107
107
97
105
105
112
127
126
112
105
156
163
140
156
117
111
107
106
109
113
113
132
115
97
78
82
89
105
105
100
96
96
97
97
109
104
102
118
125
128
111
164
164
130
124
115
104
92
84
84
94
102
148
126
96
79
78
86
86
102
102
97
93
94
93
93
103
100
128
128
121
100
124
129
112
112
97
135
107
79
84
90
96
112
132
120
88
87
87
87
96
98
166
139
133
146
133
126
123
113
119
119
119
136
123
96
80
83
90
88
96
105
105
97
96
94
86
87
93
108
103
101
91
93
93
112
105
105
104
110
121
122
121
106
146
155
146
138
130
130
114
98
106
114
114
142
118
87
83
89
85
85
103
107
99
98
102
98
98
100
106
138
130
118
106
(Based on 6 commodities imported into Norway from the East, 1880-1910.
1900=100.)
141
165
153
143
129
114
127
103
90
101
109
146
135
97
81
82
71
71
103
111
103
96
101
96
95
101
101
132
125
116
92
108
114
108
102
95
95
92
83
86
98
105
111
120
108
95
89
95
95
95
95
95
95
108
102
102
210
108
117
117
109
102
Coffee.
Tea.
Loaf.
1880................................................
1881................................................
1882................................................
1884................................................
1885.................... *.........................
1886................................................
1887................................................
1888................................................
1889................................................
1890...............................................
1891................................................
1892...............................................
1893................................................
1894................................................
1895................................................
1896................................................
1897................................................
1898................................................
1899................................................
1900...............................................
1 9 0 1 ......... ; ....................................
1902................................................
1903................................................
1904................................................
1905...............................................
1906................................................
1907...................... .........................
1908...............................................
1909................................................
1910................................................
96
82
70
70
72
64
74
111
100
117
129
117
117
125
121
122
112
82
70
61
69
63
54
51
56
63
61
52
58
61
70
138
146
131
151
131
126
121
116
106
106
106
106
106
106
106
101
98
95
95
95
94
93
95
95
95
95
95
101
103
101
103
187
194
187
171
139
123
113
113
119
126
116
113
116
119
103
94
94
84
84
87
90
84
74
77
90
97
81
81
87
90
103
Other.177
192
173
162
123
119
108
108
116
127
119
115
123
123
108
92
88
81
85
89
92
81
69
73
89
100
81
81
89
92
108
144
149
133
130
117
115
108
95
98
106
113
108
129
131
104
120
95
100
81
83
85
84
90
87
110
92
107
101
103
103
96
100
92
98
95 • 102
97
•98
107
100
100
101
96
104
112
123
115
123
127
118
117
104
Average prices in 1891-
Sugar.
Year.
156
167
144
139
121
110
110
107
111
108
Tobacco.
80
80
80
76
107
107
107
107
107
107
103
103
103
99
99
99
99
99
99
99
99
99
99
99
96
96
92
92
88
92
96
Rice.
148
119
105
95
95
90
90
90
90
95
100
105
100
138
90
86
90
95
100
100
100
100
95
100
100
100
100
no
100
105
105
All 6 com
modities.
138
136
124
121
111
105
102
108
106
113
112
110
111
118
105
99
97
89
89
89
91
87
81
83
88
92
85
86
88
90
98
N O R W A Y -----E IN A R R U U D .
T a b le
323
8 7 .—I N D E X N U M B E R S O F W H O L E S A L E P R IC E S —Continued.
(Based on wool and cotton yarns and cotton goods (7 articles) imported into Norway, 1880-1910.
Average prices in 1891-1900=100.)
Cotton yarn, un
bleached.
Wool yarn.
Year.
Lndyed
and un
bleached.
1880.......................................
1881.......................................
1882.......................................
1883.......................................
1884.......................................
1885.......................................
1886.......................................
1887.......................................
1888.......................................
1889.......................................
1890.......................................
1891.......................................
1892.......................................
1893.......................................
1894.......
@FedFRASER