groups of years.

The first period was that prior to the years 1810-1815, when prices rose. The rise was in part due to the increase of the precious metals, but perhaps chiefly due to the excessive issue of notes and to the fiat money of the war years in the United States as well as over seas. The second period ran from 1850 until after the Civil War. The price rise during the early part of this period was due to the flood of gold produced in the newly discovered mines of California and Australia. The supply of gold exceeded the normal growth of demand for it. The price rise during the latter part of this period in the United States was almost entirely due to the paper money inflation by the excessive issue of the Greenbacks. Finally, prices rose about 50 per cent between 1896 and 1914, and from 1914 to 1920 underwent a leap unparalleled for violence. The rise of the early part of this period was due to increasing gold output from new mines and to the introduction of new chemical processes of mining. The rise of the World War part of the period was due to excessive note issue and bank credit under the abnormal war time conditions.

The foregoing price changes have been indicated by indexes of wholesale prices. Although wholesale prices reflect roughly changes in retail prices, rents, wages, and other price items, nevertheless it is valuable to have an index which combines all these price factors. There is given below such a composite index, based upon wholesale prices, cost of living, rents, and wages, weighted according to the relative importance of each factor in the sum total of expenditures.

AN INDEX OF THE GENERAL PRICE LEVEL COMPARED WITH WHOLESALE PRICES AND WAGES

Reprinted from April 1, 1924, Monthly Review of the Federal Reserve Bank of New York.

Year Index Year Index Year Index Year Index Year

1875 .. 84

1885..

73

1895 .. 72

INDEX FIGURES OF THE GENERAL PRICE LEVEL, 1913 = 100

Index

1906

1918 .. 164

93

1919 .. 186

* Journal American Statistical Association, June, 1924, p. 310.

According to this chart, the general level of prices is more stable than wholesale prices. Wages show a tendency to lag behind both on the fall and on the rise of prices. The cost of living is not given on the chart because in general it practically coincides with the curve of the general price level. In the period during and since the World War, the wholesale price curve rose higher at the peak than the general price curve, but the wholesale price curve also fell lower after 1920 than the general price curve. Fully six years after the armistice, the general price level was still 10 per cent above the last year of the war, whereas the wholesale price level taken by itself was more than 20 per cent below the last year of the war. The general price level shows distinctly greater stability than the wholesale price level alone.

The fluctuations of secular trend have been continuous, and the effects upon business have been highly disturbing. Nevertheless, the range of these fluctuations has been relatively limited. The ups and downs of the price trend have been held within relatively narrow bounds. The only exception to this relatively limited range of fluctuation has been the periods of sharp war inflation. During normal peace times, the range of fluctuation has been conspicuously restricted. The chart of wholesale prices 10 shows that the price level of about 1910-1914 was approximately even with the price level of 1845-1850. If we take these two points as a base line, we find that for nearly a century peace-time prices were never more than 30 per cent above or below the line. The maximum rise and fall of wholesale prices during normal times was always within this range of about 30 per cent above or below the midway line. Even after the monetary chaos of the World War, the price level of the United States in 1924 was only about fifty points above this midway line of comparison.

The same conclusion is borne out by the curve showing the general purchasing power of money. If the forty-year average since 1875 is

10 See above, p. 532.

11 See above, p. 534.

taken as a line of comparison, it is found that the extreme decline was only 12 per cent below the line in the '70's, and the extreme rise was only 25 per cent above just before the World War. The post-war deviation was extreme. Both price charts bring out the fact that in spite of constant and severe fluctuations, prices have nevertheless tended in peace times to keep within a fairly close range of their average or midway point. In the midst of constant secular movement, there has been relative stability. In the midst of constant fluctuations, there has been a relatively narrow limit to the fluctuations. The gold standard has brought instability, but it has been an instability which, except for war, has been anchored to a central line.

As a consequence of this limited range of fluctuations in normal times, the public mind has come to have great confidence in the power of the gold standard to preserve relative stability of prices. The devotion to the gold standard has been more than a mere superstition of the mob. It has been a confidence in relative stability of the purchasing power of And this confidence has been reinforced by the disasters that have befallen the value of money whenever war has forced inconvertible paper money into circulation. War seems to mean violent inflation, gold standard or no gold standard. The United States had inflation during the World War even though it held to the gold standard, but it was an inflation much less severe than that which befell every European country off the gold standard. The gold standard commands confidence because its history during normal times shows relative stability of the purchasing power of money. This confidence cannot be shaken until a better means of regulating price fluctuations can be demonstrated. With all its imperfections, the gold standard has done better than any substitute in securing stability of value. It falls far short of the degree of stability which industry needs, but it comes closer to that degree of stability than any substitute that has thus far been tried. In the present state of political confusion on monetary issues, the public is not without grounds for looking to the gold standard as a safeguard against chaotic instability of the value of money.

The Trend of Gold Money Supply.-The chart on page 537 shows the world's stock of gold money since 1830, in terms of pence per capita.

Although the chart shows a correspondence between rising prices and rising per capita gold supply, nevertheless this correspondence does not mean that the gold changes have their effect on prices immediately and instantaneously. A lag of something like five years exists between gold changes and their full effect on prices. When due allowance is made for this lag, the secular trend of the price level has been definitely responsive to the secular trend of gold supply.

A clear distinction must be made between secular price movements and short term movements. The short term changes are due in large measure to separate causes. There gold supply does not play the determining rôle. But the long term changes reflect decisively the changes in

gold supply. There the gold supply plays definitely the determining

(This chart is constructed of logarithmic or ratio scales.)

For the basis of this chart, see the Harvard Review of Economic Statistics, August, 1921, p. 261. For data on per capita stock of gold money in the United States, see Report of Director of the Mint for 1923, p. 106. The chart as it stands over-simplifies the problem somewhat. For instance, changing habits in the use of gold might be equivalent to an increase or decrease in gold supply in so far as the price level is involved, although the number of ounces of gold remained constant. This and other complications are discussed later in the chapter. The slight over-simplification of the present chart does not impair its fundamental significance.

In interpreting the effect of gold supply upon price trend, it is necessary to bear in mind the fact that not all of the gold supply goes into monetary uses. Of the annual production of gold, upwards of twofifths ordinarily goes to non-monetary uses. During the twenty years. ending in 1919, about 58 per cent of the gold produced was added to the stock of gold money. About 25 per cent went into the industrial arts in Europe and America. And about 17 per cent went into the gold hoard and industrial arts of India, China, Egypt and other parts of the Orient. This division of gold between the money demand and the arts demand has remained fairly constant over a long period of years, and was not materially disturbed even by the World War. Of every $5 of gold produced, only about $3 is added to the money supply.12 The demand for gold in the arts tends to sustain its value. A decline in the. arts demand would weaken the value of gold and tend to raise the general price level.

A certain rate of increase in the gold supply is normal. It is only an increase faster or slower than this normal rate which causes rising or falling prices. A stable long time price level requires, not a fixed gold supply but a normal increase in harmony with the needs of growth of population and growth of production and business generally. Price stability requires increasing gold. What, then, is the normal rate 12 See Review of Economic Statistics, August, 1921, p. 259.

of growth of gold supply which would tend to bring price stability? Cassel has shown that such a rate of growth would have been about 3 per cent per annum for the period from 1850 to 1910. Irregularities in demand would prevent the use of such a rate in any absolute and inelastic sense. It is simply an approximate and long term normal average rate of growth. If this rate of growth be assumed as normal for later years, it follows that to keep the price level stable over long periods, the gold production must each year be approximately three per cent of the existing stock.13 Changes in habit and custom in the use of gold would, of course, throw this estimate off for the time being, but the significant fact is that such temporary deviations have not in the past undermined the fundamental normal rate of growth. It is important to inquire, therefore, How much does actual gold production at any time deviate from the normal rate of growth? When actual production exceeds normal, the secular trend of prices is upward. When actual production falls below normal, the secular trend of prices is downward. Actual gold increase above or below normal increase governs the up or down secular price trend. Prices fell prior to 1850 because actual gold supply did not keep pace with a normal rate of increase. Prices rose from 1896 to 1914 because actual gold supply ran ahead of a normal rate of increase. The relation of actual gold supply to normal rate of increase is of fundamental significance.

The actual supply of the future cannot be forecasted, but some factors which govern the future supply may well be examined. Discovery of new gold mines is unlikely, owing to the fact that geologists have pretty well explored the resources of the earth. Kitchin, in a careful study of gold supply, concludes, "From improved metallurgy little more is to be expected. In general, the principal producing countries seem to have passed their zenith and there are no new fields of importance to be recorded." 14 Manufacture of gold by synthetic chemistry is possible, but the cost of such manufacture is prohibitive. The prospect is for a decline instead of a further increase in annual gold production. Already, the post-war period shows a sharp decline in gold production as compared with the pre-war period. The annual gold production for the eight years ending 1916 was nearly $460,000,000. The annual production during the post-war period tends to fall about $100,000,000 short of this amount. The actual production tends to be less than the normal production required to maintain a stable price. Therefore, declares Kitchin, "But for the war we should be on the eve of a period of falling prices, as was the case in the first half of the last century and from 1873 to 1895." 15

This whole problem of the rate of gold production is linked with the cost of production of the metal. As mines become partially exhausted, the cost of mining an ounce of gold increases. This tendency toward

13 See Gustav Cassel, A Theory of Social Economy, pp. 441-455. 14 Harvard Review of Economic Statistics, August, 1921, p. 257.

15 Ibid., p. 259.