« AnteriorContinuar »
ing much more pronounced variations than the official ones. Besides being inaccurate, prices of exports and imports are representative of foreign, not domestic trade. Mr. Atkinson has tried to select articles of important domestic consumption. He derives his figures from the prices current of the Chambers of Commerce of Calcutta, Bombay and Madras, from the government publication “Prices and Wages in India," and from private sources.
All the tables agree, however, in showing a tendency of prices to rise in silver countries. To indicate very roughly the general contrast between price movements in gold and silver standard countries since 1873-76, the following table is construed, giving the average index numbers for the four years 1890-93 as compared with 1873–76 as a base :
PRICES IN GOLD AND SIlver STANDARD COUNTRIES.
78 Gold prices fell and silver prices rose about twenty per cent. between the two periods. The divergence between them may be expressed by saying that, in 1890–3, (wholesale) prices in gold countries were two-thirds (78: 117) those in silver countries with respect to 1873-6 as par. This divergence is considerably greater than that between gold and silver themselves as registered in the change in the market ratio." For 1890-3 the value of silver was three-fourths (74: 100) that of gold as compared with 1873–6 as par.
Wages in Silver Countries. While the index numbers for wages given above cannot be greatly relied upon, it is worth observing that they show the usual tendency to lag behind prices, whether in rising or falling.
Thus for India for the four-year periods 1873-6 and 1892-5 we have
1 The index numbers for gold countries are based on those of Sauerbeck for England, Soetbeer, Heintz and Conrad for Germany, and Falkner (Aldrich Report) for the United States. Those for silver countries are from the sources of columns 1, 3, and 4 in the preceding table. The prices for the United States and Japan for 1873–6 were reduced from paper to gold and silver respectively, and that for India for 1893 was reduced from the "rupee standard" to silver.
and for Japan we have
124 In both these tables the index numbers are of course for currency, that is, they are not modified for the depreciation of the paper yen in the inflation period 1873-85, nor for the appreciation of the rupee after the closure of the Indian mint in 1893.
In India, while prices rose 25 per cent., wages only rose 17 per cent., and in Japan, when prices were inflated 58 per cent., wages went up only 30 per cent., but when, in the period of contraction, prices fell 48 points wages only fell 1 point. When finally prices rose again wages did not rise at all.' In short the figures always show a gain to the laborer during falling prices, a loss during rising prices, provided it can be assumed that changes in retail prices have corresponded approximately to changes in wholesale prices.
Price Movements and Interest in India. A third result to be obtained from Mr. Atkinson's figures is fresh confirmation of the conclusion expressed in the writer's “Appreciation and Interest," that interest is high while prices are rising, and low while prices are falling. The opposite has often been assumed.
For India, according to Mr. Atkinson's figures, the most distinctly marked periods of price movements are as follows:
PRICE MoveMENTS AND INTEREST IN INDIA. ?
| These results would be even more pronounced if the index numbers for prices given in “Sound Currency" (loc cit.) were adopted. These include "rent in Tokio" an item of doubtful accuracy. As given it rises regularly from 100 in 1873 to 228 in 1894.
? The figures for the rate of interest are for the Bank of Bengal at Calcutta, and are constructed from the Appendix of "Appreciation and Interest." (Publ.
In this table every period except one exemplifies the law stated. When prices are rising, so that the principal of a debt shrinks in value, the loss to the creditor is partly made up by an increase in the interest ; when, contrarywise, prices are falling, the loss to the debtor is partly cancelled by a reduction in interest.
Stability of Prices. On one point Mr. Atkinson commits a very curious blunder. He concludes from the fact that Sauerbeck's index number for England fluctuated 9 points during 1861–71, while his own for India fluctuated 34, that “ before the demonetization of silver, gold was a more stable measure of value than silver.” But, since, for this period, gold and silver never varied appreciably from each other, one was as good a “measure of value" as the other. In one of Mr. Atkinson's own tables, in which he translates his rupee index numbers into gold index numbers, the two are identical in six of the years 1861-71, and differ by but one point in each of the other five.
This suggests the general problem of measuring variability in prices. It is somewhat remarkable that no systematic treatment of this problem has ever been made, though the materials are ready to hand and the requisite calculations are simple. The best and simplest method appears to be that of the “mean variation." The following table gives the mean percentage variations for the index numbers of five-year groups: MEAN PERCENTAGE VARIATION FOR INDEX NUMBERS OF PRICES.
1862-66 1867–71 1872-76 1877-81 1882-86 1887-91 1892-96 Average. U.S., currency, 22.7 7.7
7.3 U.S., gold, 9.2 2.9 2.5 3.5
4.3 England, gold,
6.7 1.9 4.4 3.5 Germany, Gold,
2.5 Japan, currency,
4.6 Japan, silver,
2.7 2.5 4.6
3.3 China, silver,
1.9 Ind., silver (Offic.),
2.7 Ind., rupee (Atkin.), 8.5 6.7 3.3 10.7 2.9 4.3 3.5 5.7 Amer. Econ. Assoc., 1896.) When this monograph was written, no index numbers for silver countries prior to 1873 were available. Those now published by Mr. Atkinson confirm the surmise on p. 63 as to the antithesis between price .movements in gold and silver countries before and after the rupture of the bimetallic tie.
1 The gold index number for 1866, given as 114, is a misprint for 134, as is obvious from the diagram in the Appendix.
• The mean variation of the numbers 9, 8, 12, 10, 11 is 12 per cent., for the average of the numbers is 10, and their deviations from that average are 1, 2, 2, 0, 1, the mean of which is 1.2, or 12 per cent. of the average.
3 In preparing this table, the writer was assisted by one of his students, Mr. C.F. Gehrman.
According to the last column of this table, prices have been most variable in Japan and the United States. In both cases the reason is obvious-paper money. The mean variation for 1862–66 in the United States was 22.7 per cent., the highest in the table. The next highest is that of Japan for 1877-81.' reduce prices to metallic bases the mean variations for the United States and Japan sink to 4.3 per cent. and 3.3 per cent. respectively.
The lowest averages are for China and Germany. The former is probably quite untrue to the facts, resting as it does on customs values.' Indirect evidence of this is the striking difference in variability between the customs-values figures for India and those of Mr. Atkinson. Discarding, therefore, the unreliable Chinese and Indian “official” figures, the chief international contrast for exclusively metallic standards is that between India's mean variation of 5.7 per cent, and Germany's 2.5 per cent.
It would be rash to conclude that this contrast is due to a difference in the standards. It is probably due chiefly to the alternation of famine and plenty in the Indian food supply as against the steady supply in Europe. Thus for India we find the mean variations for groups of commodities are :3
4.3 This shows that food is more variable and the other two groups less variable than prices in general. To compare the variability of food in the different countries we may construct the following table: India, Mexico.
U.S. Germany. England. Mean Variation, 7.9
1 The Japanese inflation existed 1873-85.
* See above, p. 78. That Mr. Jamieson's figures rest on customs values is stated in his report, (English) For. Off., Miscel. Ser., No. 305, 1893, pp. 13-14.
3 1861-95 is taken instead of 1862–96, as Mr. Atkinson gives no separate index numbers for food, raw produce and manufactures for the year 1896, but only the general index number, which is added in a postscript.
* The Indian figure is based on 15 food products, and is the average of the mean variations for the seven quinquennial periods 1861-95. That for Mexico is for 8 articles for the quinquennium 1892-96; that for U. S. is for 53 articles (gold prices) for the three quin. periods 1877-91; that for England is for 19 articles for the six periods 1862-91, that for Germany is for 29 articles (Kral's) for the three quin. periods 1871-75, 1876-80, and 1880-84 (see Aldrich Rpt. I,
All these figures exceed the variabilities for commodities in general in the respective countries. But the excess in India is much greater than in England, Germany or in the United States.
If we wish to compare the variability of prices under gold and silver standards, perhaps the best way is to calculate the mean variations for each country, first under the standard actually employed by it, and secondly after translating its prices into the opposite standard. Since, previous to 1873, the results would present no contrast, the following table is given for the average of the mean variations for the four quinquennial periods beginning in 1872: MEAN VARIATIONS OF PRICES EXPRESSED IN GOLD AND SILVER FOR
England. Germany. India.
5.3 This table shows clearly that commodities have moved more closely with silver than with gold in all the countries for which reliable figures are available.
Another comparison may be of interest. What has been the variability of prices in each country before and after the breakdown of bimetallism? The averages of the variations for the three quinquennial periods 1862–76, and for the three 1877-91 are:
U.S. (Gold) England, Germany. India,
6.0 The evidence is somewhat conflicting. It is clear, however, that the high figure (4.9) for the United States in the earlier period is due to the extreme variability (9.2) of prices, even in gold, during the war (1862–66). Omitting, then, the war period, the average becomes 2.7 per cent., and we find that for all three gold countries, prices have varied more in the later than in the earlier period, while in India they have varied in about the same degree.
According to a priori theory, if two metals are bound together in bimetallism, they will move more steadily than the less steady of the two partners, but not necessarily more so than the more steady, just as, if two drunken men join arms, their gait will be steadier than that of the more intoxicated, but not necessarily steadier than that of the more sober. It looks a little as if gold, when going alone, were less sure footed than silver. This is, of course, no reason for leaping over a precipice to join arms with silver. It is pertinent to add that part of the "un. steadiness" of gold is a steady downward movement, a variation but not an oscillation.