NA TURh 



^"^i 



THURSDAY. JUNE 23, 1921. 



Edit <rial and Publishing Offices : 



MACMILLAN 6- CO.. LTD.. 



ST. MARTIN'S STREET, LONDON, W.C.2. 



Advertisements and business letters should be 

 addressed to the Publishers. 



Editorial communications to the Editor. 



Telegraphic Address : PHUSIS, LONDON. 

 Telephone Number: GERRARD 8830. 



"Index-Numbers" and Wages- Regulation. 



IF we want to study the movements of prices, 

 whether within some more or less narrowly 

 defined group of commodities {e.g. foodstuffs or 

 textiles) or over a wider range {e.g. all the com- 

 moner commodities consumed in the United 

 Kingdom), the necessity is soon felt for some 

 means of summarising the diverse fluctuations 

 noted. This can be readily effected by taking 

 some particular price of each commodity as a 

 standard (usually the price in a particular year 

 or the average over a series of years), expressing 

 the price of that commodity at any epoch as a 

 percentage of the standard price — thus rendering 

 the various movements comparable — and then 

 averaging in some way for the whole series of 

 commodities the index-nuvibers thus obtained. 

 The average so calculated is usually itself termed 

 an index-number of prices, with some qualifying 

 expression to show to what it relates — e.g. an 

 index-number of wholesale prices, of retail prices, 

 or whatever it may be. 



When consideration is given to the planning 

 of such an (average) index-number, a great 

 variety of questions at once arises. For example : 

 (i) What commodities shall be included? 



(2) What sorts or qualities of each commodity? 



(3) What prices shall be used? Wholesale prices? 

 Retail prices? Import values? Export values? 

 At what markets? (4) How is the standard price 

 for each commoditv to be determined? For what 

 reference year or reference period? (5) Finally, 

 how are the individual index-numbers to be aver- 



NO. 2695, VOL. 107] 



aged? The answers to be given on these points 

 evidently must depend on the question that the 

 index-number is intended to answer. Only a 

 definite question permits of a definite answer ; and 

 two very distinct questions have dominated re- 

 searches into the movements of prices : {a) The 

 question of the effect of currency changes — e.g. 

 the substitution of paper for gold, or the varying 

 supplies of the precious metals, (b) The question 

 of changes in the "cost of living." 



Questions of the first type present their owu 

 and numerous, difficulties ; but from the point of 

 view of practice there is one simplification, that 

 the answer must be based on wholesale prices, 

 quotations for which can be obtained with com- 

 parative ease. Such questions are certain to arise 

 in a time following important new discoveries of 

 the precious metals, as they did after the gold 

 discoveries of 1848-49, which gave rise to tAO 

 classical researches, those of William Newmarch 

 and of Jevons. 



Newmarch ^ used twenty-two quotations, no 

 fewer than four of which were for cotton. 

 1845-50 was taken as the reference period. At 

 first no summary was attempted; later the index- 

 numbers for the individual commodities were 

 simply added together instead of being averaged, 

 so that the base-figure was 2200 instead of 100. 

 \'irtually, however, the index-number was the 

 simple arithmetic mean of its components. This 

 index-number has been given monthly in the 

 Economist almost without a break since 1869, 

 and remained on precisely the same basis until 

 191 1. The inconvenience of the old basis had by 

 this time become very marked — such vital com- 

 modities as foreign wheat, steel, petroleum, and 

 rubber were not included — and the whole basis 

 was revised. The number of quotations was 

 raised from twenty-two to forty -four, and the 

 base-period was altered from 1845-50 to 1901-5. 

 When the simple arithmetic mean is used for the 

 average the base-period is important, as it deter- 

 mines a virtual system of weighting. 



The work of Jevons- (1863-65) is mainly of 

 importance from his use of the geometric mean 

 as the form of average. Without entering into 

 the reasons that he assigned, some of which are 

 obscure, one reason is clear and important. If 

 the geometric mean be used, the ratio of the 

 index-number for any year B to year A is the 



1 Cf. vols. V. and vi. of " The History of Prices," by Tooke and New- 

 march ; the Mercantile Reports by Newmarch in the Journal of the Royal 

 Statistical Society, vols. xxii. , xxiii., and xxiv. ; and the volumes of the 

 Fconomist from 1864. The summary figure appe.irs to have been first given 

 in iSfg. 



* Reprinted in "Investigations in Currency and Finance" (Macmillan, 

 1864). 



