APPENDIX I
TECHNICAL DETAILS
1. The Reserve against Certificates
A. Stabilizing the Dollar Would Destabilize the Present 100% Reserve. To the plan for stabilizing the dollar, as described in Chapter IV, there should be added a proviso of some kind to insure the permanent adequacy of the gold reserve.
We have a 100% Government reserve against our present gold certificates. These certificates are really warehouse receipts, issued at the rate of one dollar for every 23.22 grains of pure gold deposited, and redeemable at all times at this same rate. But, under the plan here proposed, involving, as it does, varying the weight of the gold dollar, there would cease to be an exact equality between the number of dollars of gold in the Treasury and the number of dollars of certificates outstanding. Either might exceed the other; or first one and then the other might be in excess.
Any increase of the dollar's weight decreases automatically the number of dollars in a given physical stock of bullion. A hundred ounces of pure gold contains 2067 dollars of the present weight of 23.22 grains of pure gold. But if the weight of the dollar were doubled, the 100 ounces would contain only half (1033½) that number of dollars. Or if, instead, the weight of the dollar were halved, the same 100 ounces would contain double (4134) that number of dollars. Thus the Treasury reserve (even if there were no variation in its physical amount) would count for more or less dollars according to what a dollar might happen to weigh from time to time.
Suppose that, at the time of adopting the stabilization plan, the Treasury bullion behind the gold certificates contained 23.22 billion grains of pure gold. This mass of gold would, at that time, count as one billion dollars of 23.22 grains each and would be represented by one billion dollars of certificates in circulation. The reserve would then be 100% of the certificates against it. But as soon as the dollar's weight were changed, this exact equality would disappear. Suppose the dollar's weight were raised 1% (from 23.22 to 23.4522 grains). Although, at the instant after this change, there would be the self-same gold in the reserve and the self-same certificates outstanding, yet the number of dollars in the reserve would no longer be a billion but about 990 millions (or exactly, 23.22 billion grains ÷ 23.4522 grains). The gold reserve would then be approximately a 99% reserve instead of a 100% reserve.
On the other hand, a reduction of the dollar's weight by 1% would increase by about 1% the number of dollars contained in a physically unchanged reserve. In this case the gold reserve would become approximately a 101% reserve!
Thus the gold dollar certificates, while they would be certificates exactly like our present gold certificates in that (so far as heretofore provided for) they come into existence only by the deposit of gold and go out of existence only by their redemption in gold, would, at the same time, be very different from our present gold certificates in that they would no longer be true warehouse receipts. Having an indefinite reserve behind them, they would partake of the nature of Government notes.
B. Restabilizing the 100% Reserve. It would, of course, be perfectly possible, although quite unnecessary, constantly to restore the reserve to 100%. When gold was depreciating it would cost the Government thus to replace the depreciation. When, on the other hand, gold was appreciating the Government would reap a profit.
If the reserve became less than the certificates it could evidently be restored to equality either by more gold or by less certificates. The simpler method would obviously be to withdraw from circulation and cancel the requisite number of certificates. Thus, if there were $990,000,000 of gold reserve and $1,000,000,000 of certificates against them, the Government would simply call in and cancel $10,000,000 of certificates obtained through taxes or otherwise. In this case the Government would lose that sum.
Reversely, if the reserve should exceed the certificates, the equality could be restored either by less gold or more certificates. The latter method would be the simpler. The Government would issue and put into circulation the requisite number of new certificates, in making Government expenditures. Thus, if the gold reserve were $1,010,000,000 and the certificates outstanding were only $1,000,000,000, the Government would print and issue $10,000,000 of new certificates. In this case the Government would be making a profit of that amount.
Thus the circulation of certificates would be regulated, by issue or retirement, so as always to be equal to the number of dollars in the reserve. As has been stated, the issue could be through the payment by the Government to the public for expenses of any kind from time to time, and the retirement could be through the payment to the Government of taxes or other revenues from time to time.
But, as promptness of regulation is desirable, it would be best to anticipate such expenditures or receipts so as to make the issue or retirement follow immediately after the appearance of any discrepancy between the reserve and the certificates. Such immediate issue or retirement could best be effected by depositing certificates with banks or withdrawing deposits therefrom. In this way the effect of issue or retirement on the volume of money "in circulation," i.e. outside of Government vaults, would be immediate.
These dealings with banks would not, of course, alter the essential fact that, in the last analysis, the retirement of certificates would be through taxes, or other revenues, while their issue would make possible a reduction in taxes.
C. The Reactions Involved Thereby. These operations of canceling old, or printing new, certificates to make the certificates even with the gold reserve would, as has been noted, be quite apart from the routine operations of redemption and issue in exchange for gold, although, of course, there would be reactions between the two sets of operations.
Thus, if gold is depreciating relatively to commodities, as shown by a tendency of the index number of commodity prices to rise, the consequences would be that: (1) the weight of the gold dollar would be increased, i.e. the price of gold would be reduced; (2) the deposit of gold (issue of certificates) would be discouraged, and the redemption of certificates encouraged, both operations tending to reduce the volume of certificates in circulation; (3) as the gold reserve would fall below 100%, some of the certificates in the Government's possession would be destroyed instead of being put back into circulation, thus further lessening the volume of certificates.
The third of these operations would thus reënforce the second in effecting contraction, would help bring down the rising index number to par, and would obviate, or reduce by that much, the need, at the next adjustment period, of a further increase of the dollar's weight.
If gold were appreciating, the opposite conditions would obtain, namely: (1) the dollar would be reduced in weight; (2) the deposit of gold (issue of certificates) would be encouraged and redemption discouraged; (3) new certificates would be created and issued to bring the total volume of certificates up, so as to equal the reserve. The effect of (3) would be to reënforce (2) in expanding the currency and bringing up the sinking index number; so that the need at the next adjustment period of a further decrease of the dollar's weight would be lessened.
In short, if we thus keep the reserve constant (relatively to the certificates), we thereby lessen the variations which have to be made in the dollar's weight.[1]
D. The Definite- and the Indefinite-Reserve System Contrasted. The last sentence indicates only one of several interesting contrasts between the two forms of the stabilization system—the first, or indefinite-reserve system (described in A above), in which the reserve is allowed to drift, and the second, or definite-reserve system (illustrated in B), in which the reserve is regulated.
Under the "indefinite-reserve" system the only inflow and outflow of certificates would be through the deposit and withdrawal of gold, just as at present; whereas under the "definite-reserve" system there would be, in addition, an inflow and outflow of certificates through special issues or cancellations to keep the total outstanding volume of certificates in tune with the gold reserve.
Under the "indefinite" system the only regulator of the price level consists in adjusting the weight of the dollar. Under the "definite" system there is the added regulator of directly adjusting the volume of certificates.
Both regulators, however, act on the price level by influencing the volume of certificates. The indefinite system does so indirectly. Under this system, as noted in Chapter IV, §9, when the dollar's weight is decreased, i.e. the price of gold is increased, the deposit of gold is encouraged (as compared with what it would otherwise be) and its withdrawal comparatively discouraged, and, as we know, each deposit or withdrawal of gold implies an issue or cancellation of certificates.
In short and practically, the "indefinite" system depends for its stabilizing effect on affecting or preventing the international movements of gold which would otherwise happen, whereas the "definite" system dispenses with the need of interfering with the gold movements as they now occur.
The "indefinite" system is always subject to the risk of a breakdown, whereas the "definite" system is not. Under the "indefinite" system the reserve might sometime sink to zero and redemption become impossible, whereas under the "definite" system the adequacy of the reserve is always safeguarded.
The "definite" system would act more promptly to stabilize the price level than would the "indefinite," because, for one reason, the change in the circulation would be more prompt. The instant any change in the dollar's weight is made there is a change in the number of dollars of the reserve, and the volume of certificates is readjusted to this changed reserve immediately. Under the "indefinite" system, on the other hand, the circulation would be affected somewhat more slowly and only as the flow of gold deposits and withdrawals became changed.
E. Stabilization in Small and Large Nations Compared. The displacement of gold caused or averted by the operation of the "indefinite" system would react on the value of gold per unit of weight. Practically, however, this effect would be negligible unless the stabilization system, in the "indefinite" form, were in almost universal use. Any one country,—at any rate, any one small country, like Switzerland,—could employ the "indefinite" system without appreciably disturbing the gold market; for any displacement of gold which such a country could cause or avert would be too trifling (in relation to the vast reservoirs of gold outside its own circulation) to affect the value, i.e. the purchasing power, of gold in the markets of the world.
But if a large country,—or at any rate a large number of large countries,—should adopt the stabilization system in the "indefinite" form, any change caused in the movements of gold from, or into, their circulation might be so great as to glut or drain the small outside reservoirs, i.e. the gold in the arts and the gold in the circulation of any countries not employing the system.
An acceleration of the movement of gold from the country or countries having the system or a retardation of the movement of gold to them, such as would be caused by an increase in weight of their monetary units, would tend sensibly to depreciate the world's gold and so require a further increase of weight of the monetary units, while the reverse tendency would have the reverse effect.
The result would be that, in the process of compensating for the tendency of prices to change by a given percentage, the dollar's weight would be eventually changed by a larger percentage than would be the case if the definite reserve system were used.
Thus, suppose the system to have been started in 1900 and consider the situation in 1915. The price level would then have been kept unchanged, but there would have been an increase of the dollar's weight of more than 30% although there was only a 30% rise of prices to be overcome.
F. A 50% Minimum Reserve. The maintenance of the 100% Government gold reserve, as described in "B," is only one of several possible solutions of the reserve problem. It is the one which would fit in with the idea implied in our present system of gold certificates, namely, the idea that the certificates are circulating proxies for gold.
But there are other ways of solving the problem. One is simply to let the reserve alone so long as it remains in excess of a specified safe minimum and to replenish it only when, if ever, it falls below that minimum, i.e. to keep the indefinite-reserve system until a definite-reserve system became necessary. When, if ever, the reserve should fall to that minimum, say 50%, the principles described under "B" for maintaining a 100% reserve would thereafter apply. If the reserve became insufficient, in other words, if, at any time, the number of dollars of certificates outstanding were in excess of double the number of dollars of reserve, the excess of certificates would be retired.
The 50% limit would be reached if, for example, the present gold reserve remained unchanged in physical amount but, after a time, the dollar's weight grew to be double what it now is.
G. How Soon Might the "Indefinite" System Reach Its Limit? The time required for such a change, should such a change ever occur, would depend, of course, on the rate of change in the dollar, following the rate of depreciation of gold. Let us take a case which is extreme for rapidity of depreciation in times of peace. Suppose that gold should depreciate at the rate it did between 1896 (the low ebb of prices) and 1914 (the coming of the Great War). During this period the price level rose in the United States at the average rate of 2½% per year, which, let us assume, would require a yearly increase in the dollar's weight of about 2½%. From this figure we can calculate the time required to double the dollar's weight, and to reduce by half the number of dollars in a given physical reserve. This would be twenty-eight years.
This result assumes a gold reserve unchanged physically. As a matter of fact, the reserve would increase slightly. While the effect of the system would be to keep gold out of the country, this effect would stop short of sending it out, for that would contract the certificate circulation and, unless some special opposing cause intervened, reduce the price level. The displacement effect would stop at the point which would maintain the price level, and this, in a growing country, would admit of a slight inflow which would bring the twenty-eight years mentioned up to thirty or thirty-five.
H. A Constant 50% Reserve and a Variable Surplus. A third method would differ from the second, as described in "D" above, only from a bookkeeping point of view. There would be some advantage in separating off any surplus gold above the legal 50%. This "surplus" would then be considered as a secondary reservoir out of which the "reserve" proper could be maintained at a constant level of 50%. Reversely, whenever this "reserve" should tend to exceed 50%, the excess would overflow into the "surplus." The "reserve" proper would then be maintained at an unchanged ratio at all times.
We may, for convenience of thought, suppose the "reserve" and the "surplus" to be kept physically apart in two separate vaults in the Treasury and every week, or every day, the Treasury accounts to be squared off and gold physically transferred between the two rooms, in whichever direction it might be needed to keep the "reserve" at 50% and no more. We should then have a "reserve" the amount of which (in dollars, not weight) would always be 50% of outstanding certificates, and a "surplus" which would represent all above 50%, the percentage varying up and down owing to changes in the declared weight of the dollar as well as to the deposits of gold and the brassage receipts. Under this system of bookkeeping, the "reserve" would need to be eked out (or rather, the excess certificates would need to be retired) at the expense of the Treasury only when, if ever, the "surplus" should disappear entirely.
By this method the system would start off with a large bookkeeping profit for the Government.
I. Putting the Surplus to Work. A fourth method, and one which appeals to me as probably the best, is very similar to the third, described in "H," but makes more than a mere bookkeeping use of the "surplus." By putting the idle and otherwise useless "surplus" to work, a profit could be earned and the day of exhausting the "surplus" could be postponed still longer, if not indefinitely.
So far, we have supposed both the "reserve" and the "surplus" to be kept in physical gold. But only the "reserve" proper need be so kept. Part, or all, of the "surplus" representing the profit with which the system starts might be invested in, i.e. exchanged for, Government bonds and so virtually made to earn interest; for any bonds thus brought into the "surplus" fund would, of course, earn interest for that fund just as truly as though they were in private hands, the general fund of the Treasury paying over into that "surplus" fund the interest which would otherwise have to be paid to private holders of these bonds.
This system would recognize the needless waste involved in a 100%, or other high "reserve." In these days of economy such a "reserve" is, as one economist has said, an "expensive luxury" and one almost peculiar to the United States. In fact we are already dispensing with it, in part, by virtually converting gold certificates into Federal Reserve notes.
J. Reactions Therefrom. It should be pointed out, however, that the very operation of converting the idle gold surplus into an active bond surplus would, theoretically at least, have its own effects on the value of gold. It might either lower or raise that value. The net effect would be the resultant of two opposite tendencies.
Of these two the tendency to lower that value will be explained first. This tendency can be clearly pictured if we follow the processes involved, step by step. These processes are best followed if pictured, not as one immediate sale of the entire gold surplus for bonds, but as a gradual sale, extending over a number of adjustment periods.
The gold thrown on the market to buy up bonds would mostly find its way out of circulation, i.e. go abroad or into the arts. This outflow would be indirect; for presumably the original bondholders would not wish personally to deal in the gold bullion supposed to be given them for their bonds. They would hand back this bullion to the Government in exchange for gold dollar certificates, just as though it were new gold from the mines.
The result would be substantially the same as though the bonds were bought not with gold, but with newly created certificates, the gold remaining in the surplus.
But this extra output of certificates would not remain in circulation but would disappear again and be canceled; for they would tend to raise the index number and lead to an increase of the dollar's weight, i.e. a decrease of the price of gold, and there would be a decreased inflow of gold from the mines and imports and an increased outflow into the arts and abroad, i.e. a decrease in certificates issued by the Government to miners and gold importers and an increase of certificates received by the Government from jewelers and gold importers.
The upshot is, substantially, that, while gold would not find a welcome among bondholders, it would, as gradually cheapened at successive adjustment periods, find a market abroad, or in the arts. That is, in effect, the gold displaced by the bonds, after being bandied about between the Government, the sellers of bonds, and the gold exporters and jewelers, would go abroad or into the arts, being displaced from the "surplus" by bonds.
But, the dollar being now heavier than before, the gold "reserve" of 50% would be proportionally depleted, even assuming the physical gold in the "reserve" to remain unchanged, and so would have to be replenished from the "surplus."
In short, the original "surplus," in the process of being converted from gold into bonds, would tend to shrink in value. It would so tend to shrink both because, if the gold was forced on the market, it would have to be sold at some sacrifice and also because the resultant impairment of the "reserve" would rob some of the "surplus" before it could all be sold.
We are now ready to describe the opposite tendency by which the conversion of the gold surplus into a bond surplus would work toward a higher value of gold. The reactions so far described take no account of an important indirect effect on the price level from reducing the volume of outstanding bonds. As recent war experience has illustrated, bond issues tend toward inflation so that bond retirements would tend toward contraction. This effect would be considerable, not only because bonds are used as a basis for circulating bank notes but also because, as the most convenient form of collateral security for private loans, they are often made the basis for such loans and so for deposits subject to check. The withdrawal into the Treasury of bonds would thus tend to contract the note and deposit circulation and thereby offset, in part or in whole, the expansion from the issue of gold dollar certificates.
After the supposed initial replacement of the gold surplus by bonds, if the reserve needed replenishment from time to time, or, expressed differently, if the certificates outstanding needed to be reduced, the surplus fund would simply reconvert some of its bonds into certificates which would then be canceled up to the point necessary to restore the 50% ratio of reserve to certificates in circulation. The maintenance of this ratio by the method here described would cease to be possible only when the surplus fund should be exhausted. It could then buy up no more certificates and recourse would need to be had to taxation, as previously explained. (See "B" above.)
K. The Interest on Surplus Would Save Taxes. As long as the surplus earned any interest the expense of replenishing the reserve would be borne in part, or in whole, by this interest earned. It is roughly estimated that interest earned on this surplus would extend the 35 year period, in the imaginary example mentioned under "G," to about 50 years. If, instead of the very high rate of 2½% per annum for the accretion of the dollar's weight, we were to assume a 1½% per annum rate, which would itself be a high rate, the net result of the calculation, under the same assumptions, is that the investment of the surplus would about meet the loss indefinitely. In this case the "indefinite" system would last indefinitely and require no taxation.
L. The Future. These calculations are, of course, purely illustrative. No one can guarantee, for instance, that great gold depreciation is not in store for us from extraneous sources.
If this depreciation should be so rapid as to make the cost to the Government of stabilization a matter of real moment, — if, for instance, science should find a way to extract profitably the enormous amounts of gold now held in the southern clays, or in sea water, or in the alluvial deposits said to be at the mouth of the Sacramento River, — then the time would clearly have arrived for dispensing with the gold standard altogether, just as we would dispense with a standard based on decaying fruit and adopt something more retentive of value.
At any rate, the expense to the Government of such a future possible cataclysm is no argument against stabilization; for, as we have seen, if depreciation comes, its cost must be borne by the people anyway, whether the dollar is stabilized or not. The argument is rather the other way; for the private individual ought not to be forced to take a gamble as to the value of the money he carries any more than as to the variations of any other unit.
Instead of taking such drastic action as to give up the gold standard, we could, if we chose, make gold itself less perishable in value by limiting its production, just as diamond mines limit the production of diamonds in order to maintain their value. Such a control of gold production would simplify the problem of stabilization; for it would be a partial stabilization itself. It would be the first rough adjustment, by hand as it were, of a scientific instrument, made in advance of the finer adjustments by means of a micrometer, — i.e. the index number.
If, on the other hand, gold should, without any Government interference, become scarce and appreciate, the monetary system would be earning a handsome profit for the Government.
M. Summary. Each change in the dollar's weight changes the number of gold dollars in the reserve and disturbs the ratio of reserve to certificates.
If gold appreciates, or does not depreciate too far, the reserve will always be adequate to insure redemption. But as there is always the possibility of an excessive depreciation of gold, some method of insuring an adequate reserve should be provided. Of several possible methods, the best seems to be: to fix a definite ratio of reserve to certificates, such as 50%, which shall always be maintained; to appropriate at the outset, to the profit of the Government, all surplus above the 50% and invest it in Government bonds; likewise, in the future, to appropriate, to the profit of the Government, any further surplus which may accrue and to defray, at Government cost, any expense involved in bringing the reserve up to the 50% limit.
These Government gains and losses are not new but merely transformed. They are the same gains and losses which, under our present system, are felt by the public as individuals.
The maintenance of a definite ratio of reserve to certificates compels both of them to expand or contract in unison and leaves the currents of gold between countries and between the arts and the currency substantially as they are at present.
An "indefinite" reserve system would disturb those currents somewhat, mitigating a flood of gold into the country or an ebb of gold out of it, and, if the country be a large one, affecting the value of gold.
2. Speculation in Gold
A. Preventing "Overnight" Speculation. The best "brassage" fee. In the text (Chapter IV, §10) it was briefly stated that a small fee should be charged by the Government for the deposit of gold.[2] This fee would correspond somewhat to the old "brassage" charge for coinage and may, for convenience, be so called. Its object, however, is not primarily to defray the expense of the mint office but to prevent speculation in gold, injurious to the Government.
Without some such safeguard the Government would be in danger of sustaining loss every time a prospective change of the dollar's weight and the price of gold was known. For instance, if, at any time, the Government stood ready to buy or sell gold at, say, $20.00 per ounce and if it were known that to-morrow that price would be raised to $20.10, speculators could to-day buy, of the Government, gold bullion at $20.00 and sell it back to-morrow at $20.10, thus pocketing a profit of 10 cents an ounce overnight at the expense of the Government. Were this operation allowed or made possible, it would be costly to the Government Treasury and might temporarily deplete its gold reserve.
The opposite speculation would, were it not prevented, accompany a drop in the official price. Speculators who possessed stocks of gold could conceivably sell to the Government to-day at, say, $18.00 and buy back to-morrow at $17.90, likewise profiting 10 cents an ounce at the expense of the Government. This last operation would also be costly to the Government, though it would (during the period of operation) increase its gold reserve.
But the "brassage" requirement would effectually protect the Government from either sort of speculation. The Treasury would be put thereby in the usual position of any merchant or broker, charging, at any time, a slightly higher price than it pays at that time and making a profit. This profit, or brassage, would be the Government's fee for its services in maintaining the monetary system. Wedged between the two Government prices, it would remain a fixed percentage, say, 1%, so that the pair of prices would rise or fall together.
In order that this margin should always fully safeguard the Government it should be provided in the plan that the extent of any one shift in the pair of prices, whether that shift be upward or downward, must never exceed the margin or brassage fee. Thus, if the fee is 1%, no one shift could be more than 1%.
Evidently such a limitation would effectually stop any embarrassing speculation. Thus suppose the fee to be 1% on the Government's deposit — or buying — price, which to-day is, say, $18.00. Then the pair of Government prices, to-day, will be:
| For buying gold | . . . . . | $18.00 |
| For selling gold | . . . . . | 18.18 |
Suppose that to-morrow both prices are to be raised 17 cents, almost to the limit of 1%, or so as to be:
| For buying gold | . . . . . | $18.17 |
| For selling gold | . . . . . | 18.35 |
Clearly a speculator who tried to profit on the rising market would fail; for he would have to give to-day $18.18 and would get to-morrow only $18.17, actually losing 1 cent an ounce. Evidently, at best (i.e. if the shift were not 17 but the full 1%, or 18 cents), he would come out only even.
Reversely, if the pair of Government prices are marked down nearly to the limit, say, by 16 cents, or from
| a buying price of | $18.00 |
| and a selling price of | 18.18 |
to
| a buying price of | $17.84 |
| and a selling price of | 18.02, |
clearly the speculator cannot profit by the fall. To attempt it would mean to let the Government buy his gold to-day at $18.00 and sell it back to him to-morrow at $18.02, causing him actually to lose two cents an ounce. Evidently at best (i.e. if the shift were not 16 cents but the full 1%, or 18 cents) he would come out only even.
It is true that this limitation imposed on the shift, up or down, of the pair of official prices, while it would effectually stop injurious speculation, might, in some cases and for the time being, prevent the full adjustment in the dollar's weight (and in the price of gold) called for by the index number.
Thus the index number might indicate a change of 2% in gold prices, while only 1% was permitted under the restriction mentioned. Suppose the redemption-price of gold to be $18.00 and the deposit-price 1% less, or $17.82. Suppose, further, the index number at some adjustment period to be found to be 2% above par. Then, it is true, the price of gold could not be changed more than 1%, or 18 cents. Instead of being reduced the full 2%, or 36 cents, as would be the case if there were no restriction, the redemption price would, on account of the 1% restriction, only be reduced from $18.00 to $17.82 (instead of to $17.64) and the deposit price from $17.82 to $17.64 (instead of to $17.46).
The sacrifice in efficiency of the system here implied is, however, insignificant,[3] assuming, of course, that the fee or margin is wisely chosen in reference to the adjustment period. And even if the sacrifice of efficiency were greater, the superiority of a merely partial adjustment over the present unyielding system of no adjustment at all would be very great.
If the adjustment is to occur monthly, or bi-monthly, a brassage of 1% would seldom hamper stabilization appreciably; if quarterly, 2% might better be used, as giving more latitude. I incline, however, toward a bi-monthly adjustment period and a 1% fee, the prices being quoted, let us say, for the first Wednesday of January and of every alternate month and the resultant index number proclaimed and made effective on the Wednesday following or earlier. Calculations with a 1% fee applied to the actual prices of 1900-1914 seem to justify this choice.
B. Speculation beyond One Adjustment Period. The only kind of speculation thus far considered is the "overnight" variety based on a foreknown change in the Government prices of gold; and this is the only kind against which any safeguard is necessary.
We may, however, add here a short statement as to other kinds of speculation in gold, in order, chiefly, to show that no provision against them is needed. It is conceivable that a speculator might buy (or sell) gold to-day in order to resell to (or to rebuy of) the Government after, not the next, but some succeeding adjustment. Assume, for simplicity, that the adjustment is monthly and limited to a 1% increase or decrease. While the brassage fee of 1% would effectually prevent a speculator from buying (or selling) in January in order to resell or rebuy in February, it would not prevent him from an operation extending from January to March or later months in the hope that the first shift of gold prices,—that, say, on February 1,—might be followed by others in the same direction.
First consider bull operations. Thus, if the Government is buying gold on January 31 at $20.00 an ounce and selling it at 1% more, or $20.20, and if the speculator knows that on the following day, February 1, this pair of prices will be advanced the full limit of 1% and hopes that on March 1 it will be advanced another 1%, while it is true that he could make no profit by selling in February, he could, evidently, if his highest hope were realized, make a 1% profit by selling in March, and he may, if he chooses to take the risk involved, speculate in that hope. That is, he may buy gold of the Government in January, planning to resell it to the Government in March or later, the minimum period for the turnover being a month and a day.
But will he? Seldom, if ever, and for several reasons! In the first place, opportunities for such gain will be few and far between. The maximum gain possible will be 1% a month and that maximum will seldom continue. So far as statistics are available to tell us, gold has very seldom appreciated for a year more than 5% and never as much as 10% (except once in greenback days when gold was not the standard).
In the second place, against these small possible gains which might present themselves from time to time, the speculator would have to reckon with large, if not prohibitive, expenses. Prominent among them would be interest. If the rise per cent in the price of gold is less than the rate of interest he will be a loser anyway. If he has to pay 5% for "carrying" his load through the year and, at the end, the price of gold has risen 5%, he has not even earned interest. For when he closes his operation and sells gold back to the Government, the brassage charge of 1% must be paid, and besides these two expenses are expenses for cartage of the gold from the Government vaults and back, storage charges in the interim, and insurance.
Another obstacle is the difficulty of assembling the yellowbacks necessary to begin such a bull operation. If they are kept ready in advance, the interest expense involved in "carrying" them would be much more than merely the interest for the period of the speculative operation.
Finally, of course, the risk of failure in such an operation has always to be reckoned with. The only case in which such speculation would be reasonably likely to succeed would be when, in any month, the 1% rise in the price of gold were inadequate completely to meet the fall of the index number, so as to create the presumption that it, the price of gold, would be raised again in the following month. If a rise of 1% were announced for February 1, which was several per cent less than the fall in the index number, there would be a high probability that a month-and-a-day bull operation would yield a gross profit of 1%, the net profit, if any, being what is left of that 1% after deduction of expenses. But this is scarcely an attractive speculative proposition.
We conclude: (1) that for short periods, like two or three months, the expenses,—e.g. the expense for the cartage of gold away from and back to the Government vaults and the expense, time, and labor of preparation in order suddenly to assemble the yellowbacks (or else to "carry" them for long beforehand), would be prohibitive; and (2) that for long periods, like a year, the risk would be prohibitive. It is clear, then, that speculation of the sort here discussed would be conspicuous by its absence.
The effect of any such speculation, so far as it did exist, would, of course, be to cause expense to the Government or rather deprive it of the profit it would have made if the gold which the speculator held for a rise had been held by itself ; the temporary withdrawal of gold from the Government reserve should also perhaps be counted as a slight disadvantage to the Government.
But this is only a small part of the picture. As we have seen in the previous section, the Government, during such a period of gold appreciation as we have supposed, would itself be in the very position of the bull speculator and on an immeasurably larger scale. It would, as it were, be holding for a rise its entire gold reserve. Its percentage of reserve would be gaining and might, conceivably, even grow to exceed a 100% reserve. The speculator's losses, if any, would therefore simply be a negligible offset against the Government's own gains from the rising tide of gold value.[4]
But let us return to the contention that such bull speculation would be practically non-existent. The price movements needed for it seldom occur, and when they do occur are not foreseen. In fact, if price movements were so well foreseen, the evils which this book proposes to remedy would not be very serious!
Somewhat the same considerations apply to the opposite sort of speculation, that of the bear operator. But this type of operation,—first selling gold to the Government and then buying it back at a lower price some months later,—would amount to lending the Government temporarily an addition to its gold reserve. It would be helping the reserve when, because of its depreciation in value, it would need help. Practically, the advantage to the Government from such an operation would be small; for the possible bear operations would be limited to very small dimensions by the fact that only small amounts of idle gold bullion are available, i.e. could be, at any moment, found in stock outside the Treasury and so be capable of being immediately deposited there for the period of the supposed bear operation.
C. Unofficial Prices of Gold. We have spoken only of the official pair of prices of gold. These are like the two "gold points" in foreign exchange or the two limits used in the gold exchange standard. As distinct from these two official Government prices of gold bullion, the actual price in the open market might be at any point within these two limits, just as the price of foreign exchange may be any price within the "gold points."
The market price could never lie outside these limits. It could never exceed the redemption-price; for no one would pay more for gold than the price asked by the Government. Nor could it fall below the deposit-price; for no one would take less for his gold than he could get for it from the Government.
But, within the range set by these two official prices, the market price could float unhampered. Thus, if the limiting, or official, prices were $18.00 and $18.18, the market price might be $18.10. There would, so long as this intermediate price ruled in the market, be no actual redeeming or depositing. For no one would sell bullion to the Government for $18.00 an ounce when he could get $18.10 in the open market, nor would any one buy bullion of the Government for $18.18 an ounce when he only needs to pay $18.10 for it in the open market.
Evidently, therefore, the deposit of gold would only take place when the deposit-price, i.e. the lower limit, ruled the market; and its redemption would only take place when the redemption-price, i.e. the upper limit, ruled the market. The buying and selling of gold within the two official price limits would thus not directly concern the Government.
D. Conclusion. Our main conclusion is that speculation in gold, whether or not the Government be involved as a buyer or seller, would, if the brassage safe-guard be used, not embarrass the Government finances nor affect the smooth working of the plan for stabilizing the dollar. Moreover, such speculation would be negligible, probably more so than speculation in silver to-day.
3. Selection of the Index Number
The method of stabilizing the dollar set forth in this book consists in periodically readjusting the weight of the gold dollar so as to make its purchasing power correspond to an ideal composite dollar of commodities. The criterion for this adjustment is an index number of prices. Consequently the selection of the right type of index number is one of the essential details of the plan.
Many different methods of averaging and of weighting and many different selections of commodities, sources, and periods of price quotations have been used or suggested, making many sorts of index numbers. The selection of the best index number is a fascinating subject. It is a curious fact, however, that index numbers of different types usually agree with each other remarkably well, whatever the formula for calculation, the method of weighting, the number of commodities, etc.
In Chapter I some diagrams illustrating this important fact were given. Any reader unconvinced as to the correctness of this conclusion has only to consult the literature on the subject indicated in Appendix VI (especially the writings of Mitchell and Edgeworth) in order to be reassured.
Nevertheless, there are always some differences between the movements of different index numbers, and occasionally these differences are large. Therefore, just as, in determining the physical yardstick, it is worth while to eliminate the effects of temperature and other disturbing factors in order to obtain a unit as nearly perfect as practicable, so it is worth while to construct an index number as nearly perfect as possible.
I shall therefore indicate the points which seem to me of most importance.
The chief factors of an index number are: (1) the agency authorized to calculate the index number; (2) the markets and sources of quotations; (3) the kinds of prices (i.e. wholesale or retail, etc.) to be quoted; (4) the list of goods to be included; (5) the frequency of calculation; (6) the formula for calculation.
Out of the wide range of choice presented under each of these six heads I would, tentatively at least, make my own choices as follows:
(1) The calculation of the index number might well be put in the hands of the Bureau of Standards which now has charge of standardizing every unit other than the money unit. An alternative is the Bureau of Labor Statistics which now publishes excellent index numbers of prices for other purposes.
(2) The markets should be the chief public markets of the United States such as those now used by the United States Bureau of Labor Statistics; and the sources, government agents, standard trade journals, and books of business houses.
(3) Only wholesale prices should, I think, be used. We could not profitably use retail prices or prices of labor (wages) or the prices of securities or the prices of real estate or rents.
There are several reasons for the restriction to wholesale commodity prices, especially: (a) the greater ease of fixing or standardizing definite grades of wholesale commodities than of any of the other classes of goods mentioned; (b) the greater importance of wholesale trade and the fact that most important contracting parties are more concerned with wholesale prices than with retail; (c) the greater sensitiveness of wholesale prices to the influences which affect price levels; (d) the fact that stabilization of the wholesale index number will carry with it the stabilization of the level of retail prices far more promptly and fully than vice versa.
The last two points are worth a little elaboration. It is well known that certain prices are sensitive and others insensitive to the various market influences. For instance, the wholesale price of silver is so responsive to every market wind which blows that rarely are the quotations on two successive days alike; while, on the other hand, street railway fares have only begun to budge from the traditional five cents after having stood stock still through more than two decades of upheaval of prices of most other goods and services. As our index number is designed to register promptly the effects of an increase or decrease of money in circulation, an index number made up of prices, almost unchangeable like street railway fares and the price of postage stamps, would be, to that extent, like a painted clock, a false and useless indicator.
In short, the prompter the indications for needed adjustments, the prompter the adjustments, and the history of prices has repeatedly and clearly shown that wholesale prices respond the most promptly of all classes of commodity prices. They rise or fall before retail prices, just as retail prices do before wages, and wages before salaries.
Not only are wholesale prices a prompter and better index of the purchasing power of money but, if the level of wholesale prices is stabilized, the level of retail prices will be stabilized also. It is true that when the wholesale level changes the retail level lags behind. But the lag depends on the change; that is, other things equal the lag is most, absolutely at least, when the change in wholesale prices is most and least when that change is slowest. If the level of wholesale prices did not change at all the level of retail prices would likewise keep fairly stable; for there can be no lagging behind when there is no movement behind which to lag. When a fisherman moves his pole back and forth, the line and sinker follow, lagging behind. But if he ceases to move the pole, the line will hang more nearly plumb.
(4) What has just been said paves the way for the selection of the particular commodities to be included. Just as prices of commodities at wholesale are more sensitive or responsive than those at retail, so some wholesale prices are more responsive than others. In other words, for one reason or another, the prices of certain commodities, even at wholesale, are more or less resistive to change, i.e. they change only sluggishly and after the pressure to change them has accumulated. Steel rails, for instance, remained $28 a ton for many years.
With these requirements in mind, the next consideration is that the list of commodities be as general as possible. I would myself prefer a more general standard than food, although almost any standard based on a number of commodities would be superior to the gold standard based on one alone. If a very general standard were adopted, it is quite true that the "cost of living" in any restricted sense (such as the cost of food alone) could change somewhat, though not greatly. Furthermore an index number must serve not simply the purpose of stabilizing the value of money to wage earners but serve the purposes of transactions generally. For a large share of those transactions, a general wholesale index number is the best.
I have had a special index number calculated for me through Mr. Bell of the United States Bureau of Labor Statistics, derived from the same data and calculated by the same methods as those used by the Bureau but excluding the articles sluggishly changing, i.e. most frequently remaining unchanged in successive months. The resulting index number is, therefore, presumably more promptly responsive to any influences affecting it than any other index number of wholesale prices which has been constructed. The list of commodities on which it is based includes 75 commodities and 155 series of price quotations as follows:
| Farm Products | ||||||
| Cotton | ||||||
| Flaxseed | ||||||
| ||||||
| Hides | ||||||
Food | ||||||
| Beans | ||||||
| Butter | ||||||
| Coffee | ||||||
| Eggs | ||||||
| ||||||
| Glucose | ||||||
| Lard | ||||||
| Meal, corn | ||||||
| ||||||
| Milk | ||||||
| Molasses | ||||||
| Sugar | ||||||
| Tea | ||||||
| Vinegar | ||||||
Cloths and clothing | ||||||
| Boots and shoes | ||||||
| Carpets | ||||||
| Cotton flannels | ||||||
| Cotton yarns | ||||||
| Denims | ||||||
| Drillings | ||||||
| Ginghams | ||||||
| Leather | ||||||
| Linen shoe thread | ||||||
| Print cloths | ||||||
| Sheetings | ||||||
| Shirtings | ||||||
| Silk | ||||||
| Tickings | ||||||
| Women's dress goods | ||||||
| Wool | ||||||
Fuel and lighting | ||||||
| Alcohol | ||||||
| Coal | ||||||
| Coke | ||||||
| Petroleum | ||||||
Metals and metal products | ||||||
| Bar iron | ||||||
| Copper | ||||||
| Lead, pig | ||||||
| Lead pipe | ||||||
| Nails | ||||||
| Pig iron | ||||||
| Silver | ||||||
| ||||||
| Tin | ||||||
| Zinc | ||||||
Lumber and building materials | ||||||
| Brick | ||||||
| Lime | ||||||
| ||||||
| Shingles | ||||||
Drugs and chemicals | ||||||
| Alcohol | ||||||
| Alum | ||||||
| Glycerin | ||||||
| Opium | ||||||
| Quinine | ||||||
Miscellaneous | ||||||
| Cottonseed meal | ||||||
| Cottonseed oil | ||||||
| Jute | ||||||
| Paper | ||||||
| Rope | ||||||
| Rubber | ||||||
| Soap | ||||||
| Starch |
(5) The frequency of calculating the index number, which means the frequency of adjusting the dollar's weight, depends on a number of circumstances, including the time required to calculate the index number and that required for the effect of each adjustment to be felt.
The time required for calculation should be trifling. Judging from the expeditiousness with which some of the commercial index numbers are now calculated, and with which our Government weather maps are published, I believe that, with the aid of the telegraph, an index number could easily be calculated within two or three days after the date for which the prices are quoted.
How quickly the index number responds to a change in the money supply has never been fully demonstrated. Professor J. Shield Nicholson, plotting English war currency and index numbers of prices at quarterly intervals, found that the behavior of the price level seemed to correspond to that of the currency in the previous quarter rather than to that in the same quarter, thus suggesting a lag between cause and effect of one quarter of a year. His figures did not admit of a closer analysis. A lag about half as great seems to exist in the United States between the changes in the money in circulation (i.e. in pockets, tills, and banks other than Federal Reserve Banks) and the index numbers of Dun or Bradstreet or the United States Bureau of Labor Statistics.
The specially responsive index number which I have had calculated seems to show a still shorter lag, namely, about one month. Perhaps the most sudden and unmistakable single instance of a right-about-face of prices succeeding that of money is that in the autumn of 1915. In August of that year the money in the United States shot up suddenly and rapidly. In September, one month later, the price level likewise shot up suddenly and rapidly and has scarcely receded since. The lag is here one month.
It is interesting to note some other cases sufficiently analogous to be illuminating on this point. The closure of the Indian mints in 1893 showed the same promptness of influence on the value of the rupee.[5] The rate of exchange on London in New York has often changed from the maximum to the minimum inside of a fortnight. Again, Canadian and American price levels, as worked out by the labor bureaus of the two countries, correspond with each other year by year with extreme precision. Even month by month, judging by a careful comparison for twenty-four months, the agreement is very noticeable. The price levels of different countries tend to approximate each other like two connected lakes, through the overflow of currency from one to the other, back and forth. That the adjustment should be so delicate and prompt as between countries whose centers average hundreds of miles apart and whose trade currents are obstructed by high tariffs is not only surprising but extremely significant.
If this estimate of a month and a half be near the truth, a monthly or, at most, a bi-monthly adjustment of the index number would usually give sufficient time for any adjustment to make itself felt in the index number before the next adjustment was made.
Some such period of waiting for the effect of one adjustment to work itself out before another adjustment is made is advisable so as to avoid, as far as possible, occasional cases in which the new adjustment might prove to have been in the wrong direction and need to be recorrected later.
(6) In my book, the Purchasing Power of Money (Chapter X and Appendix to Chapter X), I have discussed at length the question of the best formula for calculating an index number. The merits of forty-four formulæ are there considered. On the whole, I favor a weighted arithmetical average like that adopted by Dr. Meeker, Commissioner of Labor Statistics, and used in the index number of the Bureau of Labor Statistics. This system was used in calculating the special index number of "responsive" commodities to which I have already referred.
As this last-named index number is the one I would, at present, most favor, it is given on the opposite page and the regular index number of the Bureau of Labor Statistics is given for comparison.
4. Selection of the Par
We may distinguish three classes of contracts, past, present, and future, i.e. those both made and fulfilled in the past, those made in the past but to be fulfilled in the future, and those to be both made and fulfilled in
| Index Number of Responsive Commodities 1913=100 | Index Number of Bureau of Labor Statistics 1913=100 | ||||||
|
101.57 | 99.33 | |||||
|
99.65 | 99.32 | |||||
|
98.52 | 98.34 | |||||
|
99.28 | 100.53 | |||||
|
102.74 | 102.18 | |||||
|
103.51 | 101.04 | |||||
|
100.74 | 99.53 | |||||
|
99.18 | 98.92 | |||||
|
97.17 | 97.79 | |||||
|
96.47 | 99.17 | |||||
|
102.43 | 102.90 | |||||
|
98.24 | 97.70 | |||||
|
101.13 | 98.02 | |||||
|
102.57 | 98.69 | |||||
|
103.93 | 100.32 | |||||
|
102.11 | 100.51 | |||||
|
99.57 | 98.47 | |||||
|
106.26 | 102.36 | |||||
|
114.47 | 109.75 | |||||
|
115.96 | 113.58 | |||||
|
119.50 | 117.77 | |||||
|
119.37 | 118.75 | |||||
|
130.48 | 126.94 | |||||
|
149.47 | 142.88 | |||||
|
156.09 | 149.75 | |||||
|
163.28 | 159.77 | |||||
|
192.52 | 180.62 | |||||
|
199.47 | 184.96 | |||||
|
196.46 | 181.69 | |||||
|
197.44 | 182.22 | |||||
|
206.14 | 185.41 | |||||
|
207.48 | 187.45 | |||||
|
201.71 | 190.71 | |||||
|
209.94 | 198.12 | |||||
|
223.21 | 206.65 | |||||
|
217.16 | 206.14 | |||||
|
213.02 | 202.01 | |||||
|
198.40 | 200.45 | |||||
|
214.19 | 206.55 | |||||
|
229.83 | 216.37 | |||||
|
|||||||
|
|||||||
the future. In the start-off, i.e. in the selection of the par or price level which the new system would undertake to maintain, only the middle of these three groups need be considered.
It is true that the chief purpose of the new plan is to provide for the third class, future contracts; for these include the numberless contracts of generations yet unborn. But for this purpose any price level whatever would serve for the par as well as any other, even if it were ten times as high or as low as the present price level.
Nor do the contracts of the past concern us. They have been written off the books and are beyond recall or correction. Nor can those who suffered losses or made gains on past contracts be selected out and indemnified or assessed damages to-day. And, if these past victims could be found, the adjustments they would require could not be accomplished by selecting any particular price level such as that existing at some particular date in the past. A reversion to standards from which we have drifted far will only make bad matters worse. Two wrongs do not make a right. Bygones must be bygones.
To urge going back to an antiquated price level was a fatal mistake in the 16 to 1 proposal in the '90s which aimed to go back to the "dollar of the daddies" and the price level of 1873.
To-day those who talk of pre-war prices as "normal" might almost as well talk of the price of 1896 as "normal." They do not stop to think that most of the adjustments have been made nor of the injustice which a reversion to an obsolete standard would do to the contracts of the present.
The war debts both in this country and in Europe, for instance, have been, for the most part, contracted at high price levels. If we should drop back to the 1913 level of prices it would almost double the burden of our national debt, for the government would have to repay dollars almost twice as big in purchasing power as the average of those which it borrowed at the five Liberty Loan dates.[6]
In considering Europe's burden of debt we must remember the unacknowledged premium on gold and the grave circumstance, of similar significance, that the price upheaval in Europe was even more serious, far more serious, than with us.
In the absence of any more exact estimate let us assume that the average price level in western Europe is threefold that of 1913 while ours is only two-fold.
Conformable to this situation we may further assume that to resume specie payments and get back to and maintain the old pars of exchange, European price levels must drop relatively to ours by about one third; for, if our present price level be maintained, Europe's would have to fall in the ratio of 3 to 2.
This means that the purchasing power of her money must appreciate in the ratio of 2 to 3. Such an appreciation would alone add 50% to Europe's burden of debt as compared with what it is at present prices.
Now, if we in America insist on reverting to our pre-war level—if, that is, we double the present purchasing power of our dollar, Europe's price level, in order to get back to the normal relation to ours, must be cut in three and her war debt virtually tripled. Even without war debts Europe would be ruined economically if her money units were thus tripled in purchasing power within a generation. Even an enhancement of 50% would be almost unbearable and would probably fan social discontent into revolution. To see that this is a grim fact we need only to recall how between 1873 and 1896 business men and farmers in America struggled to swim against the ebbing tide of prices. Yet our burden of debt was negligible compared with Europe's to-day, we were not as economically exhausted as Europe is, and the fall of prices was not so great as that we are assuming.
Under these circumstances we may well ask: Is it reasonable to expect Europe to drop her price level back to the old relation to ours or should it not be fixed at some intermediate level?
If the latter course is to be adopted so that the old relations between the various national price levels are not to be resumed and the old pars of exchange not to be reestablished, the stabilization plan as proposed in this book would afford the appropriate method for maintaining a new set of levels. For we can, by reducing the weight of European gold coins relatively to ours, enable each European nation to adopt its own price level at any desired point. If, on the other hand, we rehabilitate the old units, European price levels must go through a painful fall relatively to ours.
As to individual debts, we long ago abandoned, as impractical, the theory that a bankrupt must pay the uttermost farthing or go to prison. If there ever was a time when the modern theory of treating bankrupts should be extended to nations it is now. In fact we have already applied it to fixing the indemnity of Germany according to her ability to pay rather than according to the damage she did.
Similar considerations apply to the reconstruction loans we are making to Europe. If after loaning, in the near future, billions of dollars to Europe we double the purchasing power of the dollar, we are not only putting ourselves in the position of an unjust (and much to be hated) Shylock but the pound of flesh we would thus exact of Europe would drain her life blood and weaken her usefulness to us as a customer. The sound policy, which we are now adopting, of giving Europe long and easy credits should be carried out in fact as well as in name and this implies that we should not permit any undue appreciation of our dollar.
For various reasons, therefore, in starting the new and permanent level of prices, we cannot, very well, advocate any drastic departure from the level at which we happen to be when the start is made. In short, we ought not to start with a serious jar.
This does not mean that we must adopt the exact level of the moment.
We must take care to do justice as between the then existing debtors and creditors. To these particular debtors and creditors this question of the start-off is vital.
We cannot now say, of course, what the price level will be when the new system shall begin. All that can now be done toward deciding what the start-off should then be, i.e. what par or particular price level is thereafter to be maintained, is to point out the principles which should guide us.
If the time of adoption of the plan should happen to come after a long steep rise of prices, such as in 1919, 1873, 1865, or 1814, it is clear that the price level then existing would be too high to afford a just and proper starting point and that a somewhat lower level ought to be selected to which we should deliberately descend. Otherwise most outstanding debts would have to be paid in terms of a dollar of less value than the dollar contemplated when the debts were contracted, before prices were so high.
On the other hand, if the time of adoption of the plan should happen to come after a long steep fall of prices, such as in 1896, 1849, or 1821, the price level then existing would be too low to afford a just and proper starting point and a somewhat higher level ought to be selected to which we should deliberately ascend. Otherwise most outstanding debts would have to be paid in terms of a dollar of greater value than that contemplated when the debts were contracted, before prices were so low.
But in such cases of a rapidly changing price level, with existing contracts originating at many different previous levels, it is impossible to select any one price level well adapted to them all. If we are to apply a single correction to them all, it must be an average. We must cut our Gordian knots as we did when we resumed specie payments after the Civil War and as we always have to do in readjusting monetary standards. To strike such an average, the price level selected should, I believe, extend back of the moment when the system starts to the center of gravity, as it were, of the outstanding contracts and understandings now in existence which would be affected by the new law.
We can strike this proposed rough average of justice by making a calculation as to the past duration of existing contracts of different kinds. The contracts to pay money are the important factors to be considered. I have made a very rough estimate, largely a guess, of the average duration of the existing indebtedness which would be affected, — railroad bonds, mortgages, bank loans, and other obligations, — which seems to indicate that it is one year, or in that neighborhood.
When the proper time comes, a judicial commission to make a special intensive expert investigation of outstanding contracts might be created and the start-off then fixed in the light of the facts found, and of common sense.
If the average thus selected should effect substantial justice — which implies that this recent average price level is not far from the price level at the moment the system is launched, nor far from the price level for any other moment during the past year at least, — nothing more need be done to secure justice on existing contracts.
But if the case is otherwise — if, for instance, the average price level as calculated should differ say by more than 5% or 10% from that of any date within a year previous to the launching of the new plan, we might perhaps better give up the idea of making a single average correction to apply to outstanding contracts. Instead, special legislation "scaling" or adjusting debts might be adopted, as was sometimes done in the case of Colonial paper money. If this solution were chosen the price level for the start-off need not be changed at all from the level then existing.
Under any ordinary circumstances the price level does not vary more than 5% in a year. Probably by the time the plan could go into force the present, or recent, troubled course of prices may be sufficiently tranquilized as not to require any special legislation for scaling debts nor to afford much discrepancy between the then existing price level and the average price level for a preceding period of several months at least.
Such a debt-scaling law is, of course, not involved in the proposal to stabilize the dollar. In fact, if debt-scaling is really needed after stabilization it is far more needed without it and not once only but at many times.
But once the Gordian knot is cut and the new price level is steadily maintained, all elements still unadjusted would gradually become adjusted — wages, salaries, rents, railway rates, etc. In the long run it will be better to adjust these laggards to the price level than to adjust the price level to them. Even labor discontent would, I believe, be more successfully combated to-day by a rise of wages without a rise of the cost of living than by the reverse adjustment.
5. What Shall Be Done with Existing Gold Coins
The question is sometimes asked: How are existing gold coins to be retired, as they are assumed to be in Chapter IV? The answer is: by putting a premium on the retirement of the coins or a penalty on their retention, or both. To retire the Philippine peso (and replace it by another of less weight) a slight premium was offered to holders of the old coin up to a specified date, after which the coin was not to be received by the government except at a discount.
It may be worth mentioning that neither the retirement of existing gold coins nor the cessation of their coinage in future need be insisted upon. By a slight modification of the plan, we could permit gold coins and coinage to continue. In fact in the formulations of the plan which I usually made before the war, gold coins and coinage were retained. I then thought that the custom of handling gold coins was so firmly intrenched in some places, notably Great Britain, that the plan would be more welcome if gold coins were retained even if only as token coins.
Since then, the war itself has brought about the very retirement which we are discussing and has conquered most of the popular prejudice which stood in its way; and, from motives of economy, all nations, including Great Britain, will probably now prefer not to return to the general use of gold coins. It has therefore seemed best not to cumber the present text with the description[7] of what now proves, apparently, to be an unnecessary complication.
There is, however, a third plan possible, intermediate between the plan of the present text (in which gold coins are retired and their coinage ceases) and the plan formerly put forward (in which both coins and coinage are retained).
This intermediate plan is to authorize the retention of the existing gold coins but to stop the coinage of new coins — though retaining, of course, the unrestricted deposit of gold bullion in return for the issue of gold dollar certificates.
This third plan would seem to me to be preferable in practice to either of the other two as it would dispense with the need of recalling the few gold coins now outside of government vaults and would not involve any special difficulties no matter which way the value of gold should change.
Thus if, at any time, the gold coins were worth more than their contained bullion, they would continue to circulate as token coins, each eagle of 258 grains entitling the holder on demand to a ten-dollar certificate or ten dollars of gold bullion (of more than 25.8 grains per dollar).
On the other hand if, at any time, they were worth less as money than the contained bullion, they would be melted by the owners, disappear as coin, and be deposited with the government as bullion in return for certificates. Any gold coin in the government vaults would likewise be melted.
But the process would stop there, limited by the amount of gold coin available. There would be no "endless chain" of redemption of certificates at one rate and recoinage at another such as would (as explained in my article in the Quarterly Journal of Economics, February, 1913) have to be guarded against in the second of the three plans.
In spite of the slight practical superiority of this third method of handling existing coin, I have preferred in Chapter IV to present the first method as simpler to understand, and less confusing to the reader. With so few coins as are now in circulation it is really almost immaterial which of these two methods is adopted.
6. What Shall Be Done Concerning the "Gold Clause" in Existing Contracts
One of the questions which will have to be faced when stabilization is adopted is: What should be done with the numerous bonds and other contracts containing a "gold clause" to the effect that the contract calls for payment in "gold coin of the present weight and fineness"?
This clause had its origin in the nineties when the "free coinage of silver at 16 to 1" was agitated. It was intended to safeguard the creditor against payment in silver dollars which, it was justly feared, would be greatly depreciated in purchasing power, if the 16 to 1 proposal were adopted.
The statute enacting stabilization ought to include a specific settlement of this gold-clause question.
Otherwise, it would be left for the courts to interpret, and long and costly litigations would be sure to result. Pending a decision by the Supreme Court the status of all gold-clause contracts would be uncertain. In attempting legally to resolve this uncertainty there would be two widely different views possible. It might plausibly be argued that in the gold clause, "coin" was specified only for its convenience to handle, as compared with bullion. According to this interpretation gold-clause contracts ought, under stabilization, to be reckoned in terms of gold bullion, and when the gold dollar became greater or less than 23.22 grains of pure gold, contracts containing the gold clause would still have to be measured in dollars of bullion of 23.22 grains each.
But, on the other hand, it might with almost equal plausibility be argued that the word "coin" must be taken literally and that the creditor had the right to require the delivery of such coins or their equivalent. If such were the interpretation and (as supposed in Appendix I, § 5) gold coin were not retired but continued in existence as token coins, i.e. at a value above that of the contained bullion, the technical fulfillment of the gold clause by the payment of these "over-valued" coins, or their equivalent, would coincide with the use of stabilized dollars to which they would be equivalent (as explained in Appendix I, § 5).
This interpretation, insisting on "coins," would, however, encounter difficulties if gold coins were abolished entirely (as suggested in the text) or if, though not recalled by law, they were all melted into bullion because the bullion in them had come to be worth more than their face value (as supposed in Appendix I, § 5).
All these technical controversies would be avoided if, in the statute establishing a stable dollar, the gold clause in existing contracts were abrogated entirely and unambiguous requirements were substituted to meet the new situation and carry out the real object of the gold clause.
It should be pointed out that abrogation, though beyond the power of our individual states under Article I of our Federal Constitution, is apparently quite within the power of the Federal Congress.[8]
Having thus abrogated the gold clause in all contracts outstanding at the date of the stabilization law, Congress could replace that clause by whatever provision it chose.
The provision which, on the whole, seems to me the fairest from various standpoints is to make all such contracts exactly like all others, i.e. payable in stabilized dollars.
That such a requirement would, even technically, reinstate the gold clause—under at least certain circumstances (such as the retention of gold coin as "token coin")—might well be argued, as has just been shown.
But the only justification worth while for such a law is that it would do justice and by doing justice we would, in a broad sense, be carrying out the intent of the gold clause. This clause was never intended to introduce a hazard into contracts but to take one away, not to enable one of the contracting parties to mulct the other but to prevent it. In a broad sense, therefore, the substitution of stable dollars for "gold coin of the present weight and fineness" would carry out the spirit if not the letter of that clause under the new conditions. Stabilization would supersede the gold clause as a more perfect way of attaining the same general object — contractual justice.
And not only would complaint over such substitution be unjustified but it would rarely, if ever, be made or thought of for the very simple reason that we would go on in our habit of thinking in terms of dollars.
Under stabilization the debtor for $10,000 would still expect to draw his check for exactly $10,000 and the creditor would expect to receive exactly that sum. In 99 out of 100 cases the question of whether, under the gold clause, the check ought perhaps to be drawn for a larger or smaller sum than the face of the obligation would never enter the head of either party.
On the other hand, if exceptional treatment were given to contracts having the gold clause, so that these were not to be fulfilled in stabilized dollars, there would be great complaint. For then the only way to discharge a gold-clause contract to pay $10,000 would be to pay something more or less than $10,000 according as the price of gold had risen or fallen. If, because of a raking up of the gold clause, a debtor owing $10,000 is informed by his creditor that he has to pay, say, $10,842.79 the $842.79 will obtrude itself like a sore thumb and seem to the debtor, as it really would be, the exact measure of an injustice.
On the other hand, if the discrepancy between the stabilized dollar and "gold coin of the present weight and fineness" were in the other direction and a debtor tendered what he owed under a $10,000 debt subject to the gold clause by offering to pay $9,500 (which we shall suppose is the equivalent of ten thousand dollars of 23.22 grains of pure gold each) the creditor would always feel, and justly, that he had been robbed of $500 by a wrong interpretation of a clause originally inserted to safeguard him against just such injustice.
Moreover, if the gold clause were not thus assimilated to the new dollar great confusion would be introduced from the double reckoning. Probably the most extreme instance would be that of the insurance companies, the assets of which are invested largely in gold-clause bonds but the liabilities of which to their policy holders are payable in "lawful money." If the dollars used for measuring both assets and liabilities are to be made different, these companies might become either bankrupt or greatly enriched as a consequence.
It seems clear, therefore, that the solution here offered of the gold-clause problem is the justest, simplest, and most smoothly working of the various solutions which might be considered.
If, however, Congress should conclude that it was necessary to provide further against the possibility of any complaint, it could leave the contracting parties some choice in the matter. That is, the Act to stabilize the dollar could serve notice that stabilized dollars would be understood unless objections were raised by either contracting party between the date of the Act and the date on which the new system was to be put into effect. For cases where such objection was actually raised, the law could provide that the two parties to the contract might come to an agreement and further that, in case of their failure to do so, the creditor should have the right to choose, in advance, between the stabilized dollar and a dollar of bullion of the present weight and fineness. When this choice was made, it could not, of course, be altered afterward, even if, as would be quite possible, the creditor should find that he had chosen against his own interests.
The result would undoubtedly be that even the few contracting parties who would raise the question of the gold clause would find an easy way to settle it, while none of those who failed to raise the question could ever maintain that they had not been given a fair chance, for they had at one time been virtually told to speak then or else forever after hold their peace.
7. Bank Credit and the Plan
A. Misconceptions. It should be pointed out that the plan proposed in this book, by maintaining the purchasing power of the gold dollar, necessarily maintains also the purchasing power of all other dollars, so long as these other dollars are kept interconvertible with gold dollars.
This implies that due provision for redemption, in gold, of paper money and bank deposits must be maintained by suitable legislation or regulations, such as are usually afforded by sound currency and banking laws and practices. That is, the stabilization plan presupposes sound banking though not any special form of sound banking.
In this connection some curious misconceptions have arisen, such as the notion that to stabilize the gold dollar can apply only to gold and not to credit or can only correct such instability as has its origin in gold and not such as has its origin in credit, in commodities, or elsewhere. These views overlook the fact that all dollars are interconvertible.
One friend of the plan fell into an opposite error in that, instead of finding any limitations on the power of the plan to effect stability, he assumed that it would dispense with the need of any restrictions whatever on the inflation of paper or credit! We could, he thought, "run the printing press" ad libitum and, for instance, pay the cost of the Great War thereby, without suffering the penalty of high prices!
Of course the process of stabilizing the dollar has no such magic power to take the place of sound currency and banking. If, with one hand, we were to stabilize the gold dollar and, with the other, we were to inflate paper or deposits, we should be pulling both ways at once and if the conflict were continued long enough inflation would, in the end, exhaust and defeat stabilization. The inflation, tending to raise prices, would necessitate an increase in the dollar's weight which would involve a proportionate decrease in the number of dollars in the reserve. The reserve would also be depleted by the increased tendency to redeem certificates in the heavier dollars, the certificates displacing the gold and driving it abroad.[9]
All would go well and the price level and purchasing power of the dollar be approximately maintained, so long as redemption could also be maintained. But if the inflation were persisted in far enough, the constant increase in the credit superstructure and decrease in the gold base (i.e. in the number of dollars in it) would ultimately break down redemption. Thereafter the gold dollar would cease to exist as a factor in our monetary system, leaving only irredeemable paper and deposit dollars in actual use. After this breakdown the paper and deposit dollars would depreciate.
B. The Effect of War on Bank Credit. During the Great War, as in other great crises, the exigencies of Government finance caused, in almost all countries, an expansion of paper and credit almost regardless of the effect on prices or on redemption. At such times the pressure for inflation is almost, or quite, irresistible. The paramount object is then financing the war rather than maintaining monetary standards, and any stabilization plan might have to be temporarily suspended as one of the emergency measures of war, just as the English Bank Act is temporarily suspended during a crisis. Stabilization could be maintained provided the war could be financed without recourse to inflation, i.e. could be paid for out of taxes and loans from savings. Inflation, which is really a forced loan, puts the otherwise unpaid cost of the war on the shoulders of those of "fixed" incomes, in the form of a high cost of living.
In the future, we have reason to believe, no such world crises are in store. But should they come, and stabilization were suspended, we would, of course, be no worse off than if there had been no stabilization. (See also Appendix II, § 2, D.)
C. Maintenance of Redemption. Thus stabilization, to be successful, implies the maintenance of redemption. The typical or ideal, though by no means the only efficient, type of a redemption-law is one which keeps deposits and paper money more or less proportional to bank reserves (of gold bullion dollar certificates) together with a Government reserve law (as described in § 1) which keeps the volume of gold bullion dollar certificates proportional to the volume of gold dollars in the Government reserve. Under such conditions all parts of the circulating medium tend to expand or contract in unison and a change in weight of the basic gold dollar carries with it a control of the whole mechanism of exchange, cash, and credit.
Bank credit, paper, and the gold reserve (in dollars) would then expand or contract as needed (by the requirements of trade, etc.) to keep the price level constant.
D. The Rôle of Bank Discount. It would be going somewhat outside the scope of stabilization plans to discuss, in detail, the banking procedure for keeping the credit superstructure more or less proportional to the redemption base of gold or gold certificates.
Suffice it, in this connection, to call attention to one factor in the case, the importance of which is seldom realized—the rate of bank discount.
Under almost any sensible banking system the rate of discount is one of the regulators of the volume of credit relatively to reserve. If there is undue expansion of credit relatively to the reserve, the rate of discount is raised to curb it. If, on the other hand, there is a plethora of reserve, the rate of discount is lowered to stimulate an increase of credit. As the expansion and contraction of credit are directly related to the price level, the rate of bank discount is thus concerned very vitally with the price level.
The greatest of banks, the Bank of England, is a model in this respect. It alternately defends and releases its gold reserve, which is the basic gold reserve of England, by raising and lowering the bank rate.
The report, after the Armistice, of the Lord Cunliffe Committee on Currency, Banking and Foreign Exchange shows clearly how the bank rate keeps the English price level in tune with world price levels. Speaking of this long-established system the report says:
"When, apart from a foreign drain of gold, credit at home threatened to become unduly expanded, the old currency system tended to restrain the expansion and to prevent the consequent rise in domestic prices which ultimately causes such a drain. The expansion of credit, by forcing up prices, involves an increased demand for legal tender currency both from the banks in order to maintain their normal proportion of cash to liabilities and from the general public for the payment of wages and for retail transactions. In this case also the demand for such currency fell upon the reserve of the Bank of England, and the bank was thereupon obliged to raise its rate of discount in order to prevent the fall in the proportion of that reserve to its liabilities. The same chain of consequences as we have just described followed and speculative trade activity was similarly restrained. There was, therefore, an automatic machinery by which the volume of purchasing power in this country was continuously adjusted to world prices of commodities in general. Domestic prices were automatically regulated so as to prevent excessive import."[10]
Professor Knut Wicksell of Sweden has, for many years, advocated a more extensive use of this regulative function of the rate of bank discount as a means of preventing cycles of credit and prices. Mr. Paul Warburg, formerly of the Federal Reserve Board, has suggested that the index number of prices should be one of the data scrutinized by the Federal Reserve Board to help guide it in fixing the rate of discount. Senator Shafroth proposed that the Federal Reserve Board should fix discount rates in such a manner as to regulate credit with the object of stabilizing the level of prices.
This adjustment would not of itself, however, be sufficient to keep the price level stable; for while it controls the credit superstructure, it does so only relatively to the metallic base and if this base is uncontrolled relatively to the needs of business, the credit superstructure being proportional to the base, that credit superstructure is equally uncontrolled relatively to the needs of business.
But, given both a stabilization of the base and any sound banking system, that is, any system which makes credit expand or contract with an expansion or contraction of reserves, we can secure complete stabilization.
8. International Aspects of the Plan
A. The Mint Price. It goes without saying that the plan would have a wider usefulness if adopted by all nations than if adopted by only one, or a few. But, if at first its general adoption were not found feasible, the question remains: Would the plan work and work well if adopted, say, by the United States alone?
Many persons have imagined that a single nation could not make the plan work, that the money problem, being essentially an international one, requires concerted action, that it is therefore imperative that there should be "the same mint price of gold" throughout the world, otherwise gold would flee entirely from or to the nation which should alter the present "uniform" price.
We shall see that these ideas are mistaken. In the first place let us see clearly the "fallacy of the mint price." Superficial reasoning, starting from the fact that our mint price ($20.67 an ounce of pure gold) and England's mint price (£3. 17s. l0½d. for gold fine) are now "the same," concludes that, if our price were lowered 1%, i.e. to $20.46, while the English price remained unchanged, all our gold would be sent to England to take advantage of the "higher" price there.
But $20.67 would then cease to be "the same" as £3. 17s. l0½d. and $20.46 would become "the same" as £3. 17s. l0½d. ! The reason is that comparisons between English and American prices are based on the "par of exchange" and this par would change. At present the par is $4.866 of American money for £1 of English money; but this par of exchange is based on the relative weights of the dollar and the sovereign! Consequently a change in the weight of the dollar and the price of gold will change proportionally the par of exchange. If the dollar's weight is changed 1% so that the mint price becomes $20.46 (instead of $20.67), the par of exchange will become $4.82 (instead of $4.86⅔).
It is true that each increase in the weight of the gold dollar in America—in other words, each fall in the official American price of gold—would at first tend to discourage the minting of gold in America. The miner might send more of his gold to London, where the mint price had not changed, and "realize" by selling exchange on the London credit thus obtained. But the rate of exchange would soon be affected through these very operations by which he attempted to profit, and his profit would soon be reduced to zero; the export of gold to England would increase the supply of bills of exchange in America drawn on London and lower the rate of exchange until there would be no longer any profit in sending gold from the United States to England and selling exchange against it. When this happened it would be as profitable to sell gold to American mints at $20.46 per ounce as to ship it abroad; and $20.46 in America would be the exact equivalent, at the new par of exchange ($4.82), of the English mint price of £3. 17s. 10½d.
Consequently, although the new mint price of $20.46 is in figures lower than the old, yet, as it is in heavier dollars, it would still be "the same" as the English mint price of £3. 17s. 10½d.
It is clear that this sameness of mint price as between the two countries really means nothing of economic consequence, for the reason that all prices of gold are in terms of gold. At bottom the basic fact is simply that exchange is at par when an ounce of gold in America will, in the exchange market, buy the right to an ounce of gold in England.
This obvious fact is concealed, or "camouflaged," by measuring gold in America in terms of dollars, and gold in England in terms of sovereigns; but the dollar and the sovereign are merely units of weight, like the ounce, with definite ratios to the ounce and to each other. Of course the price of gold in America (in terms of itself) is "the same" as the price of gold in England (in terms of itself) when either is translated into the other by means of the par of exchange (or ratio between the two units).
This would be self-evident if the numbers were a little simpler. Thus, if the dollar were exactly a twentieth of an ounce of pure gold and the sovereign exactly a quarter of an ounce, the mint price in America would be $20.00 an ounce and in England, £4 an ounce; and the par of exchange would be , or $5 per £. Naturally, then, £4 an ounce would be "the same" price as $20 an ounce when we translate £'s into dollars at $5 per £, i.e. $20=5×$4, or 20=×4. Such sameness of price would evidently still exist if the dollar were doubled, i.e. were made a tenth of an ounce. The mint price in America would then be $10 per ounce which (the par of exchange being or $2½ per £) would be "the same" as £4 an ounce; for $10=$2½×4, or 10=×4.
To turn from theory to experience, if those against whom I am reasoning were correct, everybody would now take his gold to the Mexican mint where he could get twice as many dollars as he can get from the United States mint! Obviously the fallacy lies in the fact that Mexican dollars are half as heavy as ours.
B. Gold Reserves and Price Levels as Internationally Related. So much for the effect of our individual action on the international exchanges. The second effect to be emphasized is the release of the United States from the danger of alternate gold famines and feasts. At present foreign countries may deluge us with gold or drain it away. The only effectual stop to the inflowing tide comes from the rise of our price level and our only important defense against the continued ebb of gold is from the fall of our prices. Thus is our gold reserve now at the mercy of Europe. Their banking and currency policies, over which we have no control, their trade and tariffs, their wars, all affect our gold supply. Thus the Great War, by dumping the gold of the belligerents on neutral countries (including, in 1915–1917, the United States), inflated prices in these neutral countries and a reflux of gold may deflate them whenever Europe deflates her currencies sufficiently.
The only methods used in the war to safeguard against these floods and ebbs of gold were: (1) — as against a flood — the action of Sweden, Holland, and Spain virtually stopping the free coinage of gold and (2) — as against an ebb — the "embargo" on the export of gold adopted by many countries, including the United States.
These were attempts at a partial control of a nation's gold supply by stopping the inflow of gold into the nation's circulation or its outflow therefrom.
But the stabilization plan would afford a complete control of the amount of gold, measured in dollars, without forbidding or much affecting the inflow or outflow of gold measured in ounces!
Had we had stabilization in 1915 we would have been protected against gold inflation, from which we have suffered so grievously. When the gold began to flow in and prices to rise, our gold dollar would have been enlarged. Also the number of gold dollars in the country would have been kept from increasing, despite the increase in the physical amount of gold. Finally the price level would be kept from rising.
Likewise we would have been defended against a drain of gold and would have needed no embargo. If gold began to leave us and prices to fall, gold dollars would be lightened, their numbers would be thereby kept from decreasing, despite the decrease in the physical amount of gold, and the price level kept from falling. If, then, the United States should "go it alone," we would be emancipated from the present involuntary "entangling alliance" of our currency with foreign currencies.
Implied in the last would be the emancipation of our price level from its entangling alliance with foreign price levels. The price level of each country now depends on that of those other countries which have the same monetary standard. The "High Cost of Living," one of the manifestations of inflation, communicates itself from one country to another having the same standard and no one country can avoid the common contagion so long as it has the common unstable unit.
In short, under our present system our money, credit, and price level are far more internationally entangled than they would be if we had stabilization. So long as we let the gold standard drift we are helpless to protect ourselves from the effects of our neighbors' acts on that standard. The close of the war makes us especially liable to the influence of changing currency policies of Europe, policies as yet unknown and unknowable.[11]
C. Exports and Imports. As to the effect on international trade in commodities, these effects would be complex and somewhat varied according to circumstances, though not, probably, important in magnitude.
Suppose that the United States had a stable dollar and other gold standard countries had not. Suppose further that gold units tended throughout the world to depreciate and therefore that we were obliged successively to increase the weight of the dollar, i.e. to decrease the price of gold, and thereby to lower the rate of foreign exchange as measured in American dollars.
Under these circumstances the price level in the United States would remain stationary, the price levels in other countries would rise, and the rates of exchange between the United States and those countries would change accordingly, e.g. exchange on London would decline.
Normally, or in the long run, the change in the exchange between two countries is proportional to the divergence of their price levels. Thus, let us assume that prices in England gradually increase until they have doubled while those of the United States remain the same, and that the exchange on London falls correspondingly from $4.86 to $2.43 per pound sterling.
Under these assumptions imagine an American exporter who now finds that, while the American prices with which he is concerned are about the same, the English prices he can get for his goods are doubled. He receives a bill of exchange for £200 where before he received one for £100. But when he sells the £200 bill at $2.43 per pound he receives the same $486 which he used to get when he sold the £100 bill at $4.86.
Evidently if the changes in price level and the changes in the rate of exchange thus correspond to each other, there is neither gain nor loss.
So far as gains or losses do exist they are only differential and due to the failure of the price and exchange movements to correspond as exactly as is assumed above. That is, there is here, as always where price movements occur, some lagging behind of certain elements. These evils are evils of transition and tend to disappear as the transition, i.e. the price movement, disappears or the movement is reversed. Whatever harm is done is due not to a changed[12] price level, but to a changing price level.
If, as seems to be usually the case, the rate of exchange is adjusted more promptly than the price level, the exchange will reach $2.43 before the price level has doubled and the exporter will receive less than £200 and, so, less than $486. In this case he would have suffered somewhat from English inflation which, presumably, he would not have suffered had there been no stabilization and had American prices but kept pace with English prices. On the other hand, if the pound sterling should appreciate, the American exporter would gain slightly.
Reversely, the American importer would gain a little from stabilization when foreign price levels rose and lose when they fell.
We see that stabilization in one gold standard country alone would expose importers and exporters to the chance of certain slight differential gains and losses, one of the two classes always gaining from the maladjustment while the other is losing. This evil of introducing a new risk to importers and exporters is offset, however, by the removal of the old risks connected with their dealings within the United States.
Furthermore, since the war, there is no common gold standard anyway! Currencies are in chaos, both relatively and absolutely. A stabilized dollar could well be resorted to as a common denominator in foreign trade, just as the old "trade dollar" was resorted to. If international contracts were drawn in stabilized dollars we would be freed from all the uncertainties of roubles, marks, lire, francs, etc. These uncertainties would then fall only on the countries employing such units.
But even if foreign trade were somewhat disadvantaged by stabilization, we must remember that usually over nine tenths of American trade and doubtless a larger fraction of American contracts are within the borders of the United States so that, to the great bulk of Americans, stabilization would be an unmixed blessing.
It is unfortunately true, however, that, to most people, international trade looms up, out of its true perspective, as a far bigger factor in a nation's economic life than it ever really is. As every teacher of economics knows, the average citizen, untutored in economics, is a victim of the old mercantilistic fallacy and still imagines that the old mercantilistic phrases—"favorable balance of trade" and "unfavorable balance of trade"—which have been handed down to us are to be taken literally. Often it is even assumed, absurd though it obviously is, that the only gain which a country as a whole can get is in an excess of exports over its imports and an accumulation of money. This is not the place to consider such elementary errors. Any textbook on economics exposes the fallacy; and the lessons of our recent experience with an accumulation of gold should make it clear that an accumulation of money in a country simply debases the purchasing power of that money.
D. Spreading the Gold Points. There would then be no real international inconvenience introduced by the stabilization plan unless we count as an inconvenience the fact that the "gold points" of exchange would, under certain conditions, be wedged a little further apart (by the amount of the brassage) than at present. Even this would not happen so long as conditions were such that in both of the countries gold is flowing into circulation and not out (or, in both, out and not in) so that the price of gold within each country remains continuously at the lower (or continuously at the upper) of the two limits set by the brassage and discussed, in another connection, in § 2 above.
Under these conditions, a periodical shift in the official prices of gold would not widen the gap between the gold shipping points; it would merely raise or lower them both in unison. Nor would the reversal of the golden stream from an inflow into circulation to an outflow from circulation widen that gap, provided the reversal took place simultaneously in both countries. Only when it happened that gold would flow into circulation in (say) the United States and out of circulation in (say) England, would the gold shipping points between the two countries be spread apart by the amount of the brassage.
By proper international arrangements as to exchange, even this occasional result could be avoided. The international exchange could be itself stabilized at Government expense as has been done during the war.
E. The Adoption of the Plan Would Spread. Thus, on the merits of the question, there is little or nothing to be said against stabilization by one country alone, while its advantage to the country adopting it would be very great indeed. In this connection I may call attention to a recent dispatch from London which says: "English capitalists are certain that the country which first succeeds in reorganizing its currency will be able to obtain a large share of international business."
Sooner or later the perception of the advantages of stabilization would probably lead to the general adoption of the stabilization principle. This might come about either at once by concerted action or gradually by individual action.
With a league of nations, joint action in such matters will be far easier than ever before; and we must not forget that there was joint action once in the case of the "Latin Union" which maintained bimetallism. In this case France, Belgium, Switzerland, Greece, and Italy joined in a uniform standard of currency based on gold and silver. The present exigency will create a powerful motive toward some such action.
The war has upset the monetary standards of the whole world and has brought forward the questions of resumption, deflation, high cost of living, and price movements generally. All of these are related to the more fundamental question of a standard of value, of which that of stabilization is an unescapable part.
Monetary standards already constitute an international question because, under our present system, any disturbance in the price level in one country necessarily affects the price levels of the rest.
If the stabilization plan were adopted internationally, there should, of course, be a common index number. This would not sacrifice greatly the accuracy of adjustment for any one nation; for we have already seen that the index numbers of different countries having the same monetary standards are very similar and we know that, with the future development of international trade, there will come about an even closer harmony of price movements.
In case joint action could not be secured at the outset, individual action by one country, especially if that country were the United States, would, almost certainly, lead to the general adoption of the plan.
Objectors point out that this was not true of bimetallism. Their argument is that if an agreement on international bimetallism could not be secured we cannot hope to secure anything so ambitious as an international standardization of monetary units and that, therefore, we need not trouble ourselves about attempting the impossible. But, as one will see by reading H. B. Russell's book on "International Monetary Conferences," when the proposal to resume bimetallism was made there was a special obstacle which would not exist in the case of the stabilization plan.
This obstacle was the realization, based on the experience of the Latin Union, that when any nation or nations have bimetallism in successful operation, all the other nations enjoy its benefits as much as if they had it themselves but without the trouble or responsibility of operating it. For instance, the Latin Union had, as an intermediary between the gold standard countries and the silver standard countries, virtually held together the rupee and the pound sterling in a fixed ratio to the great benefit of England without effort on her part. Under such conditions, for a long time after bimetallism broke down in 1873, almost every nation wanted some other nation to restore it but wanted, if possible, to avoid doing so for itself! In modern slang each would "let George do it."
In the case of the standardized dollar, on the other hand, if one nation should break the inertia of custom and adopt the plan, and if it were soon seen that this nation was getting benefits from it while all the other nations lacked these benefits and, in fact, were being somewhat injured by the upset in their exchange pars, these other nations would soon want to come in, as the only way to escape the evils and secure the benefits. The case would be analogous, not to the reluctant attitude toward international bimetallism, but to the "scramble" of nations to get on to the gold standard. After the breakdown of bimetallism in 1873 commercial nations turned, one after another, to the gold standard in order to secure the advantage of a stable rate of exchange on London and other important commercial centers.
9. Numerical Illustrations under Various Assumptions
A. The Standard Hypothetical Case. A mental picture of the actual operation of the stabilizing process can best be obtained from illustrative numerical examples, such as are considered in this section.
There are five factors determining the stabilization process: the "brassage" charge, which serves as the limit on any single adjustment of the dollar's weight, the amount of "adjustment" of the dollar's weight for a given deviation from par of the index number, the amount of the "influence" which said adjustment has on the index number, the "lag" of time elapsing between the adjustment and completion of its influence, and the prior "tendency" of the price level to rise or fall, were it not combated by the stabilization process.
Our first example will be called the "standard hypothetical case." In later sections the several conditions will be separately varied from those of this standard case.
The standard hypothetical case assumes the five factors to be as follows:
(1) Brassage charge: 1%.
(2) Adjustment rule: 1% for each 1% of deviation from par of the index number (no one adjustment to exceed the brassage).
(3) Influence thereof on index number: 1% for each 1% of adjustment.
(4) Lag of said influence following the adjustment causing it: 1 adjustment interval.[13]
(5) Tendency of price level: were it not for stabilization the price level would at first increase 1% each interval; afterward, it would decrease 1% each interval.
The fifth assumption implies that, were it not for stabilization, the index number would be:
At beginning of the 1st interval 100.
At the beginning of the 2d interval 1% above 100 or 101.
At the beginning of the 3d interval 1% above 101 or 102.01.
At the beginning of the 4th interval 1% above 102.01 or 103.0301.
At the beginning of the 5th interval 1% above 103.0301 or 104.060401.
Etc., increasing as by compound interest.
Not to put too fine a point on these figures, we may omit decimals and use the figures 100, 101, 102, 103, 104, etc., until the "compounding" produces an appreciable effect. When, for instance, the index number is in the neighborhood of 150 the 1% increase will make the next index number greater by about 1½; and when it is in the neighborhood of 200, the 1% increase will make a difference of about 2. Thus, if we assume that (were it not for stabilization) the course of prices would rise 1% each adjustment interval from 100 to 200 and then fall, the index numbers would run approximately as follows: 100, 101, 102, 103, 104, . . . 150, 151½, 153, . . . 198, 200, 198, 196, . . . 150, 148½, 147, . . . .
Under the fifth assumption, we may distinguish four types of price movements—the four which could take place in actual experience,—a rise, a fall, a reverse after an upward movement, a reverse after a downward movement.
We are now ready to calculate[14] what, under the five assumptions formulated, the stabilized course of the index number will be.
At the start, the index number being 100 or par, no adjustment in the dollar's weight will be made. Consequently, during the ensuing or first interval, the index number will be subject only to the assumed tendency to rise 1%, so that, at the beginning of the next adjustment interval, it will be 101, just as though no system of stabilization existed.
At this adjustment date, therefore, there is a deviation from par of the index number of +1%. This leads (by assumption 2) to an adjustment of the dollar's weight of 1%.
The influence of this adjustment will (by assumption 4) be felt during the ensuing interval and be registered at its close. That influence is (by assumption 3) 1%. If there were no other force, therefore, than this par-ward influence, the index number would then return to 100, or par.
But there is another force; namely, the tendency of the index number to rise 1% during this (second) interval. This force restrains the index number from returning to par and keeps it at 101. In short, the downward and upward forces neutralize each other so that the index number remains unchanged at 101.
Summarizing thus far, we may schedule the events as follows:
At beginning of 1st interval: index number 100; no adjustment of dollar's weight.
During 1st interval: no influence from adjustment, but only unhindered tendency of index number to rise, +1%.
At beginning of 2d interval: index number, 101; adjustment of dollar's weight, +1%.
During 2d interval: influence of aforesaid adjustment on index number, —1%, neutralizing tendency of index number to rise, +1%, leaving, ——
At beginning of 3d interval: index number unchanged at 101.
But the deviation from par being still +1%, the adjustment in weight at the adjustment date now reached (the beginning of the 3d interval) is again +1%, which will again strive to bring down to par the index number during the 3d interval by 1%, and again be foiled by the 1% rising tendency.
The same reasoning gives precisely the same result for each subsequent adjustment interval, as long as the 1% upward tendency continues.
That is, in each case, the new index number is the last index number (101) minus the 1% influence toward par, due to adjusting the dollar's weight, plus the 1% tendency to rise.
Thus, at each successive milestone, the formula for finding the new index number in terms of the old is 101 - 1 + 1 = 101, as long as the 1% upward tendency exists.
The sequence is:
| Index Number[15] | Influence of Adjustment on Index Number |
Tendency of Index Number, if Unstabilized | |
| Beginning of 1st interval | 100 | ||
| During 1st interval | 0 | +1% | |
| Beginning of 2d interval | 101 | ||
| During 2d interval | -1% | +1% | |
| Beginning of 3d interval | 101 | ||
| During 3d interval | -1% | +1% | |
| Etc., repeating. |
When the downward tendency begins, the price level in the first adjustment interval will fall from 101 to 99. The reason is that, during this interval, the 1% influence exerted by the adjustment in the weight of the dollar is reënforced by the assumed tendency to fall 1%. That is, the index number after the first interval of fall will be 101 - 1 - 1 = 99.
The index number, 99, is now 1% below par, i.e. the deviation is now -1%. The dollar will, therefore, be reduced in weight 1%. The influence of this adjustment will now be 1% upward, counteracted, however, by the 1% tendency to fall, still assumed to exist. That is, the next index number will be 99 plus the 1% influence less the 1% tendency, or 99 + 1 - 1 = 99; and it will, thereafter, remain 99 as long as the tendency to fall continues.
Assuming, to fix our ideas, that the reversal from an upward to a downward movement occurs at the point at which the index number would have reached 200 had there been no stabilization, the index numbers in successive adjustment intervals are given (omitting decimals) in the following table as they would be, both without and with stabilization.
| Without Stabilization | With Stabilization |
| 100 | 100 |
| 101 | (100 + 1 =) 101 |
| 102 | (101 - 1 + 1 =) 101 |
| 103 | (101 - 1 + 1 =) 101 |
| 104 | (101 - 1 + 1 =) 101 |
| 150 | (101 - 1 + 1 =) 101 |
| 151½ | (101 - 1 + 1 =) 101 |
| 153 | (101 - 1 + 1 =) 101 |
| 154½ | (101 - 1 + 1 =) 101 |
| 198 | (101 - 1 + 1 =) 101 |
| 200 | (101 - 1 + 1 =) 101 |
| 198 | (101 - 1 - 1 =) 99 |
| 196 | (99 + 1 - 1 =) 99 |
| 194 | (99 + 1 - 1 =) 99 |
| etc. | etc.etc. |
From the standard hypothetical case, just calculated, experimental departures will be made in order to determine what set of rules will serve best in controlling price movements, as they are actually experienced.
B. Changing the Assumption as to the "Lag."
(a) Assumptions same as in standard case except: lag changed from 1 to 2 adjustment intervals.[16]
The index number, being uninfluenced by stabilization, will follow the assumed tendency for two adjustment intervals, and run: 100, 101, 102.
That is, at the start, or beginning of the first interval, there is no deviation from par and so no adjustment in weight; at the beginning of the second interval there is an adjustment in weight of +1%; but, as the lag between this adjustment and its influence on the index number is now assumed to be two adjustment periods, the following index number is unaffected and remains 102.
The par-ward influence (assumed as 1%) of the 1% adjustment made at the beginning of the second interval will, under our present assumptions, be felt during the third interval. During that interval this par-ward influence will strive to bring the index number down 1% from 102. But the assumed upward tendency of 1% keeps the index number at 102. At the beginning of the third interval, the 2% deviation would cause a 2% increase in the weight of the dollar, were it not for the brassage charge limiting any one increase in the dollar's weight to 1%, which will therefore be the increase effected. This 1% increase in the dollar's weight, made at the beginning of the third adjustment interval, influences the index number during the fourth interval to pull it downward; but the upward tendency will keep it still at 102, Thus the formula will be: 102 (the index number at any adjustment date) – 1 (the influence of the adjustment at the preceding date) + 1 (the tendency to increase) = 102.
The following table shows the results:
| Index Number[17] | Influence of Adjustment |
Tendency if Unstabilized | |
| Beginning of 1st interval | 100 | ||
| During 1st interval | 0 | +1 | |
| Beginning of 2d interval | 101 | ||
| During 2d interval | 0 | +1 | |
| Beginning of 3d interval | 102 | ||
| During 3d interval | -1 | +1 | |
| Beginning of 4th interval | 102 | ||
| During 4th interval | -1 | +1 | |
| Etc., repeating. |
Upon reversal of the assumed price tendency the stabilized index number falls to, and remains slightly below, par.
(b) Assumptions same as in standard case except: lag changed to 3 adjustment intervals.
Following the same reasoning as under "a," we find the index number rising to 103, and then remaining at 103, the influence, thereafter, of the 1% adjustment being exactly neutralized so long as the 1% tendency to rise continues.
(c) Conclusion as to lag.
In the preceding examples the stabilization process is very simply and effectively applied, the restraining influence sooner or later (depending on the ratio between the lag and the adjustment interval) taking effect and thereafter, while unable to restore the index number to par on account of the steady upward (or downward) tendency, keeping the index number constant at a point slightly above (or below) par.
We see that the greater the lag in proportion to the adjustment interval, the greater is the range of the index number from par. Yet, even if the lag is many times the adjustment interval, the index number keeps near par.
Thus, if the adjustment interval is two months and it is assumed that the effect of any adjustment were not felt until six times that interval, or a whole year, the index number would at most deviate only 6%, assuming the other conditions unchanged from the standard case.
As a matter of fact the lag is not great.
Our experience during the war and other evidence mentioned elsewhere (Chapter II, § 8 and Appendix I, § 3) show that the influence of inflation or contraction is apparently rather prompt, the lag being probably less than two months, and possibly less than one month for an index number composed of the most responsive commodities.
It is desirable that our adjustment intervals should not be too short compared with the lag, say not shorter than a quarter of the lag.
On the other hand, the adjustment interval might be taken longer than the lag. For such a case the calculations and results would be the same as where the lag is one entire period. The influence of the adjustment would then be complete some time before the following adjustment date arrived; but since no index number is calculated during the interval, our calculations would not be affected.
Ideally, i.e. to secure the greatest attainable degree of closeness to par, the adjustment interval should be as nearly equal to the lag as possible. If the interval is shorter than the lag the influence is not felt fully until another adjustment, perhaps in the opposite direction, is made. A daily adjustment would therefore not help but hurt the closeness of the approximation. If the interval is longer than the lag, the price level is left for the balance of the interval to vary uncorrected. We would be neglecting the opportunity to correct it promptly.
C. Changing the Assumption as to the "Tendency."
(a) Assumptions same as in standard case except: tendency increased from 1% to 2% per adjustment interval.
Although the assumption hitherto made (of a 1% change in price level during every adjustment interval) implies a very rapid change (if the adjustment interval is two months), we shall now assume a movement twice as rapid.
In this case, the index number will be 102 at the end of the first adjustment interval. This deviation calls for an increase of 2% in the dollar's weight, but the brassage charge limits this increase to 1%. Hence, at the end of the second interval the index number is acted upon by two forces, the restraining influence (from the increased weight of the dollar) of —1% and the tendency to a further increase of +2%. The net result is +1%; that is, the index number becomes 103. At the next adjustment period a similar conflict between a 1% decrease and a 2% increase causes the index number to become 104, and this process continues. In short, instead of increasing by 2% each adjustment interval, the index number increases by 1%. The stabilization process, under these circumstances, cannot altogether control the price tendency, as long as this continues upward, but can decrease it by half. On the reverse movement, after passing par, the movement below par is similarly retarded by stabilization.
If, however, the brassage limitation permitted a larger adjustment, the restraint would, of course, be more effective. We shall see this clearly after the effects of different amounts of brassage are shown.
(b) Conclusion as to tendency.
We conclude that the greater the tendency of the index number to vary, the further the index number will deviate from par before being arrested—especially if the tendency exceeds the brassage—but that, unless the tendency to change is very great or long continued or both, the index number will still stay close to par.
D. Changing the Assumption as to the "Brassage."
(a) Assumptions same as in standard case except: brassage changed from 1% to 2%.
The results are exactly the same as in the standard case. The higher brassage makes no difference because it was already high enough not to limit the adjustment, the tendency and lag also being as assumed.
(b) Assumptions same as in standard case except: brassage changed from 1% to 2%, and also: tendency, first upward and later downward, changed from 1% to 2%.
Under these conditions, the restraining influence exactly neutralizes the tendency and the index number is stabilized at 102 (during the upward tendency) and at 98 (during the downward tendency).
(c) Assumptions same as in standard case except: brassage changed from 1% to 2% and also: lag changed from 1 to S adjustment intervals.
In this case, the stabilization process results, while the price tendency is upward, in a movement of the index number between 1% below par and 5% above par. At reversal, the index number at first drops as low as 93, but recovers and (during the downward tendency) fluctuates between 1% above par and 5% below par.
(d) Conclusion as to brassage.
We conclude that, in general, the greater the brassage the greater the freedom of the index number to vary. It is freer to approach toward par; but it is also freer to depart from par, if the lag is very great, i.e. if the adjustment interval is made a very small fraction of the lag.
Practically, the brassage should be between, say, 1 % and 3%. Within such limits it makes remarkably little difference to the result whether the exact figure is near one extreme or the other and any figure within these limits is adequate to secure a close approximation of the index number to par except under most extraordinary conditions such as those existing in a World War.
E. Changing the Assumption as to the "Adjustment."
(a) Assumptions same as in standard case except: adjustment changed from 1% to 2% (per 1% deviation).
The results are exactly the same as in the standard case. The larger adjustment would make no difference because the brassage limitation would prevent it from taking effect.
(b) Assumptions same as in standard case except: adjustment changed from 1% to 2% and also: brassage changed from 1% to 2% or above.
A deviation above par of 1 % would then call forth a 2% increase in the weight of the dollar. The influence of this 2% adjustment would be to decrease the index number by 2%, which influence, however, would be partly neutralized by the assumed upward tendency of 1%. The net result would be a fall of 1% which would bring the index number back to 100 at the next adjustment date. This would call for no adjustment in the next period, and the index number, being acted upon only by the upward tendency, would become 101. Thus it would continue to alternate between 100 and 101.
(c) Assumptions same as in standard case except: adjustment changed from 1% to ½%.
We find the following results:
| Index Number[18] | Influence | Tendency | |
| Beginning of 1st interval | 100 | ||
| During 1st interval | 0 | +1 | |
| Beginning of 2d interval | 101 | ||
| During 2d interval | -½ | +1 | |
| Beginning of 3d interval | 101½ | ||
| During 3d interval | -¾ | +1 | |
| Beginning of 4th interval | 101¾ | ||
| During 4th interval | -⅞ | +1 | |
| Beginning of 5th interval | 101⅞ | ||
| During 5th interval | -15⁄16 | +1 | |
| Etc. |
The index number increases but never reaches 102.
(d) Conclusion as to adjustment.
We conclude that the nearer the adjustment is to the deviation the better stabilization will work—always assuming, of course, that the influence of the adjustment is as in the standard case.
F. Changing the Assumption as to "Influence."
(a) Assumptions same as in standard case except: influence decreased from 1% to ½% (per 1% of adjustment).
We have hitherto assumed that an adjustment of 1% in the dollar's weight would influence its purchasing power 1%. But this need not be assumed and would not be strictly true in practice, especially if the number of dollars, both of money in circulation and of deposits subject to check, were not kept strictly proportioned to the number of gold dollars in the reserve (as by the method described in Appendix I, § 1 and § 7).
The calculations, in the present case, are very similar to those of "E (b)" above.
Calling the original price level 100%, the index number at the end of the first adjustment period will be 101%. The dollar will now be increased by 1% which, according to our present supposition, would tend to lower the price level only half as much, i.e. ½%. As, during the second interval, the price level tends to go up 1% the new index number will be 101-½ +1 or 101½. The excess of 1½% above par will now call for a corresponding increase in the dollar's weight; but the brassage limitation holds it to 1%.
Accordingly, the next adjustment date will see an increase in the dollar's weight of 1% and the price level will be l01½-½+1 or 102. The next increase in the dollar's weight will be again limited to 1% and the index number will be 102-½+1 or 102½, and so on.
Evidently, as the brassage is 1% the power of the system to stabilize will be limited to ½% per adjustment interval.
(b) Assumptions same as in standard case except: influence changed from 1% to ½% and also : brassage changed from 1% to 2% or more.
The results are, evidently :
| Index Number[19] | Influence | Tendency | |
| Beginning of 1st interval | 100 | ||
| During 1st interval | 0 | +1 | |
| Beginning of 2d interval | 101 | ||
| During 2d interval | -½ | +1 | |
| Beginning of 3d interval | 101½ | ||
| During 3d interval | -¾ | +1 | |
| Beginning of 4th interval | 101¾ | ||
| During 4th interval | -⅞ | +1 | |
| Etc. |
The stabilization now keeps the index number within 2% of par, the figures being identical with those of "E (c)" above, although the conditions as to the adjustment and its influence are different.
(c) Assumptions same as in standard case except: influence changed from 1% to 2%.
The index number will alternate between 100 and 101 as follows :
| Index Number[20] | Influence | Tendency | |
| Beginning of 1st interval | 100 | ||
| During 1st interval | 0 | +1 | |
| Beginning of 2d interval | 101 | ||
| During 2d interval | -2 | +1 | |
| Beginning of 3d interval | 100 | ||
| During 3d interval | 0 | +1 | |
| Etc. | 100 |
(d) Conclusions as to influence.
We conclude that the adjustment of the dollar may be greater or less than the influence it has on the index number without greatly lessening the efficiency of stabilization.
G. General Conclusions on Variations from the Assumptions of the Standard Case. We have seen that the stabilization device is such as to adapt itself, in a remarkable degree, to widely varying conditions.
The brassage charge may be anything from, say, 1% to 3% without greatly affecting the results and also (under any ordinary conditions) without impairing greatly the efficiency of stabilization.
The adjustment of the dollar's weight may be anything from, say, ½% to 2% per 1% of deviation without very greatly impairing the efficiency of stabilization,—at least under reasonable assumptions as to the other factors (influence, tendency, lag, and brassage).
The influence of the adjustment on the index number may be anything from, say, ½% to 2% per 1% of adjustment without greatly affecting the results,—at least under reasonable assumptions as to the other factors.
The lag may vary widely relatively to the adjustment interval. Practically this means that the frequency of adjustment may (other things equal) be anything from, say, a quarter of the lag to many times the lag without greatly restricting stabilization.
The tendency of prices to rise or fall may be permanently rapid and temporarily very rapid without often pulling the index number more than 1 or 2% from par,—assuming the other factors which affect stabilization (brassage, adjustment, influence, lag) to be as in the standard case. And no matter how great the tendency of prices to vary, almost all of this tendency can be eliminated if those other factors are adapted to the situation.
Practically, the problem is to secure the most ideal adaptation of these other four factors to the tendency as it exists.
The tendency (barring extraordinary times such as those of the Great War) has seldom averaged for long more than 4% per annum, which is more than the average rate in the long, and almost unprecedentedly rapid, peace-time movement from 1896 to 1915.
In any one year the movement seldom reaches 12% or an average of 1% per month. In the whole pre-war period, 1890-1915, of 25 years for which we have figures of the United States Bureau of Labor Statistics this happened only twice, the figures then being 13% and 14%.
We have monthly figures beginning only with 1900. From these we find that, beginning with January, 1900, and taking every other month up to the end of 1915, the successive jumps of the index number by bimonthly intervals were not over 1% in two cases out of three, were not over 2% in nine cases out of ten, and were not over 3% in 31 cases out of 32.
Our problem, as already stated, is how best to deal with such a tendency by selecting, as ideally as is open to us, the other four factors.
First consider the ideal brassage. This is scarcely capable of exact formulation. Evidently 3% would permit a full adjustment in almost all cases. But, as the calculations in "H" below will show, even a 1% brassage will be adequate for all practical purposes and other calculations which I have made show that there is remarkably little difference in the results between 1%, 2%, 3%, and 4% brassages.
To fix a figure, let us call the ideal brassage 1½%.
The ideal adjustment is, evidently, that which will tend exactly to correct the deviation on which it is based, thus bringing the index number back to par (except as further deviated by further tendency, and this of course is apt to be in either direction).
This ideal adjustment depends on what influence that adjustment has on the index number. If the influence is less than in the standard case the adjustment might advantageously be greater and vice versa. For instance, if the adjustment is 2% per 1% of deviation, this will just correct the deviation when the influence of that adjustment is ½% per 1% of adjustment. For an influence of ½% per 1% of adjustment (i.e. of 1% per 2% of adjustment) makes an influence of 1% per 1% of deviation, which is the ideal.
As a matter of fact the conditions as to adjustment and influence assumed in the standard case are, doubtless, approximately true to life. At any rate if the "definite" reserve system (described in Appendix I, § 1, B, F) and the method of regulating the volume of bank credit (favored in Appendix I, § 7) are adopted so that the entire volume of circulating media is controlled as a whole in direct proportion to the percentage change of the dollar, a 1% adjustment in the weight of the dollar would have a 1% influence.
Even to employ the "indefinite" reserve system would, as we have seen in Appendix I, § 1, D, not greatly change the situation, unless or until a very large part of the world adopted that system. In that case there would be some advantage in increasing the adjustment to l½% per 1% of deviation or even to 2%, the exact ideal figure depending on the results of an investigation of the repercussive effect of adjusting the weight of the dollar on the value of a given weight of gold.[21]
We come next to the ideal lag relatively to the adjustment interval; or, to express it in more practical terms, the ideal length of the adjustment interval relatively to the lag, or the ideal frequency of adjustment.
As we have seen, the ideal frequency is not the greatest possible frequency, but is such a frequency as will make the interval equal to the lag.
The lag for Dun's index number is probably about 1½ months. The lag for the index number of responsive commodities described in Appendix I, § 3, is probably less than 1 month. The ideal frequency is therefore probably somewhere between a fortnight and a month and a half. In the calculations of "H" below it is conservatively taken as two months.
The influence being as indicated, the adjustment should evidently be 1% per 1% deviation.
It will be seen then, that (1) the tendency is beyond our control; (2) the lag measured in months is under control only to a small extent as we may choose the index number but, measured relatively to the adjustment period, is fully under control; and (3) the influence may be assumed to be 1% per 1% of adjustment, provided we have a proper reserve system for the certificates and a proper banking system for deposits (as explained in Appendix I, § 7).
Practically, therefore, these three factors (influence, absolute lag, and tendency) must be taken as we find them and we can merely choose the best brassage, adjustment, and frequency of adjustment.
These we find to be, in round numbers, substantially those of the standard case.
In the following subsection we shall see what the results would be as applied to the historical facts since 1900, taking the brassage as 1% and the frequency of adjustment as bi-monthly, both somewhat more conservatively than the ideal.
H. The Stabilization Process Applied to the Actual Course of Prices.
(a) The assumptions suitable for practical conditions.
We pass now from the highly theoretical calculations just given to the practical question of how close to par the actual index number would keep under stabilization. The best answer can probably be reached by applying the same sort of calculations as those above to the actual price movements experienced since, say, 1900, the year from which the monthly index number of the United States Bureau of Labor Statistics dates.
We shall assume, as the best adjustment period, two months. This, as has been observed, is more than the length of probable lag between any adjustment and its influence on the price level, as explained in Appendix I, §3. To be still more conservative, however, we shall assume that only two thirds of the influence from the adjustment is felt within the first adjustment period of two months and that the remaining third is felt in the ensuing period.
We shall assume the brassage to be 1%. Probably 1½% or possibly 2% would be better, but the above examples and various other calculations applied to the actual price tendencies in the period mentioned show substantially the same degree of closeness to par under brassage charges varying from 1% to over 4%.
We shall assume that (except where limited by the brassage) the adjustment of the dollar's weight is 1% for every 1% deviation from par of the index number, and that the influence of this on the index number is 1% for each 1% adjustment.
These assumptions may be put in the following form:
(1) Brassage: 1%.
(2) Adjustment: 1% for each 1% of deviation from par of the index number (subject to the condition that no one adjustment shall exceed 1%, the amount of the brassage).
(3) Influence: 1% for each 1% of adjustment in weight of the dollar.
(4) Lag: ⅔ of this influence felt within the first adjustment interval of two months and the remaining ⅓ in the second adjustment interval.
(5) Tendency: What it actually was according to the index number of the United States Bureau of Labor Statistics between 1900 and the present.[22]
Assumption (5) means that, instead of considering purely hypothetical cases, we are now to study what would have happened if we had had stabilization started January, 1900.
This affords a very severe test; for the period taken is one of unusual variability of the price level before the war (although of less average variability than the 1% every two months, assumed in the standard hypothetical case).
(b) Calculation of stabilized index numbers. The following table shows the first stages of the calculation:
| 1900 | I Stabilized Index Number[23] |
II Influence of Adjustment of Dollar's Weight |
III Tendency (Percentage Change of Actual Index Number) | |
|---|---|---|---|---|
| Two Thirds of the Influence felt in First Following Interval |
One Third of the Influence felt in Second Following Interval | |||
| Jan. 1 | 100 | |||
| During Jan. and Feb. | +1.35 | |||
| Mar. 1 | 101.35 | |||
| During Mar. and Apr. | -.67 | -1.33 | ||
| May 1 | 99.35 | |||
| During May and June | +.43 | -.33 | -1.88 | |
| July 1 | 97.57 | |||
| During July and Aug. | +.67 | +.22 | - .64 | |
| Sept. 1 | 97.82 | |||
| During Sept. and Oct. |
+.67 | +.33 | + .92 | |
| Etc. | 99.74 | |||
Let us follow the above calculations in detail, taking the index numbers cited from the bulletin of the United States Bureau of Labor Statistics. Changing them by simple proportion so that the price level of January, 1900, when the system is supposed to have been adopted, shall be 100, the index number for March 1, 1900, is found to be 1.35% above this par of January. This is the signal for raising the weight of the redemption bullion 1%, since the brassage will not permit the full increase of 1.35%. This 1% increase in the weight of the dollar, by assumption (3), affects the index number by 1%. Also, by assumption (4), ⅔ of this influence is felt in the following adjustment interval (ending May 1) and ⅓ in the next (ending July 1).
The May index number will then combine the effects of the ⅔ of 1% or .67% downward influence as well as of the downward tendency during this interval which is -1.33. The stabilized figure for May is, therefore, 101.35 - .67 - 1.33, or 99.35.
This figure is below par, and calls, in turn, for a decrease in the weight of the dollar. In this case, however, the brassage limitation does not come into play. The deviation is -.65, the adjustment -.65, and the influence +.65 of which two thirds, or +.43, follows in the next interval, and the remaining third, +.22, follows in the interval next but one. The July stabilized index number is found from that of May as follows: 99.35 + .43 - .33 - 1.88 = 97.57.
The stabilized and unstabilized index numbers are:
| Unstabilized | Stabilized | |
|---|---|---|
| Jan. 1 | 100.00 | 100.00 |
| Mar. 1 | 101.35 | 101.35 |
| May 1 | 100.00 | 99.35 |
| July 1 | 98.11 | 97.57 |
| Etc. |
Figure 12 gives, for comparison, the curves from 1900 to 1918 for this stabilized index number and for the actual course of prices in that period.
Except for the period when the war begins (as it does at the close of 1915) to produce a great effect on the price level, stabilization works almost perfectly,[24] keeping the index number within 2% of the original par during two thirds of the time, within 3% of par six sevenths of the time, and within 4% all of the time.[25] During this same period, on the other hand, the unstabilized index number wandered from the starting point 30%.
Beginning with the fall of 1915, however, the upward tendency becomes too strong and, in spite of the stabilization mechanism, the stabilized price level rises in the diagram 86% above par. This, of course, is a small rise as compared with the rise which actually occurred, as the index number rose 200% above the original starting point.
The deviation from par of the stabilized index number would be slightly less if the brassage were 2% and less still if it were 3%, etc. Yet I doubt whether the brassage should be increased much, if any, above 1%, (1) because presumably we do not now need to provide against a contingency so remote as a repetition of such a situation as that caused by the Great War, and (2) because, if another such situation should develop, a partial stabilization is the most we could expect. The fiscal necessity of the Government is then so paramount a necessity that inflation is probably unavoidable. If the Government itself succeeds in avoiding direct inflation, the people, in subscribing to bonds by borrowed money, will bring about an indirect inflation.
Fig. 12. The Index Number with and without Stabilization
The lower curve shows how the stabilization process would work (under the assumptions of subsection H) as contrasted with the upper curve which shows the actual course of prices.
Until the influence of the war was strongly felt, the stabilized index number would have kept usually within 2% of par; whereas the unstabilized index number rose 30%. Afterward, stabilization, limited in its influence (on account of the brassage) to 6% per annum, only partially restrained the soaring index number.
Of course the foregoing figures do not pretend to give, with absolute exactness, what would have happened under stabilization, for the reason that the five hypotheses do not state the exact conditions which would obtain. Thus the tendency is of course ever changing. The influence of the adjustment in weight of the dollar would doubtless be distributed somewhat differently from the distribution assumed in the figures. But the stabilization process, by its very nature, adapts itself to whatever situation is presented and relentlessly pursues and ultimately eliminates each deviation as it occurs. The figures give us as good a picture as we can secure, until the actual plan is inaugurated, of what the general behavior of the index number would be.
This behavior would usually be very stable. For to keep the price level within two or three per cent of par as is here done, is, for all practical purposes, to keep it perfectly stable. The only evils of instability which are really felt are the cumulative evils of a long sustained rise or fall. In particular, as we have seen, demonstrations of popular unrest, like populism during falling prices and I. W. W.ism during rising prices, develop only after the fall or rise has proceeded both long and far; and this could not happen with the price level closely tethered to par.
10. A Tentative Draft of an Act to Stabilize the Dollar
Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled :
(Replacement of Unstable, by Stable Dollar)
Sec. 1. That at three o'clock, Eastern time, in the morning of January 1, 1921, the gold dollar of the United States shall cease to be a constant quantity of gold of variable purchasing power, and thereafter shall be a variable quantity of standard gold bullion of approximately constant computed purchasing power.
Said quantity of standard gold bullion, constituting a gold dollar at any given time, shall be ascertained and fixed, from time to time, by the computation and use of index numbers of wholesale prices as hereinafter set forth.
Provided: That the gold dollar shall remain 25.8 grains of standard gold until some other quantity is fixed under this Act.
(Computation of Index Number and Its Deviation from Par)
Sec. 2. That for the purpose of computing approximately the fluctuations of various wholesale prices in the United States after the year 1920, and of computing index numbers such as will approximately measure the average of such fluctuations, and of computing therefrom the approximate fluctuations in the purchasing power of gold, the Bureau of Standards (or Bureau of Labor Statistics) hereinafter referred to as the Computing Bureau, shall proceed as follows:
(a) From the list of commodities and the quantities thereof marketed at wholesale in the United States in 1909, heretofore compiled by the Bureau of Labor Statistics from data of the Census of 1910 and other data and published in Bulletin No. 181, Wholesale Prices Series No. 4, the Computing Bureau shall, immediately after the passage of this Act, make up a list of selected commodities comprising about 100 commodities (not less than 75 nor more than 125) deemed by it to be the most suitable (as to importance and otherwise) to be used for computing the said index number.
(b) Immediately after December 25, 1920, the Computing Bureau shall compute, from the best accessible data, the average price of each of these commodities for the year 1920 (to December 25). (c) From the several average prices, so computed for 1920, and the quantities so listed for 1909 by the Bureau of Labor Statistics, the Computing Bureau shall compute an ideal composite "goods-dollar" for reference purposes consisting of such quantities of the several selected commodities, proportional to the quantities so listed by the Bureau of Labor Statistics, that their aggregate value, at the average prices so computed for 1920, shall equal one hundred cents. (This selection of the price level of 1920 as the base or par is, of course, merely illustrative. See Appendix I, §4.)
(d) From average wholesale prices computed from price quotations taken on the first Wednesday (or, if that day be a holiday, the next business day) of the months January, March, May, July, September, November of 1921 and each year thereafter, the Computing Bureau shall speedily compute the value, in cents, of the composite "goods-dollar," and such value in cents shall be the index number of prices for that date.
(e) The Computing Bureau shall compute the deviation from par of such index number by subtracting one hundred cents from said index number. Thus if the index number is $1.01 the deviation is 1 cent or 1% above par, and if the index number is $0.98 the deviation is 2 cents or 2% below par.
(Transmission Thereof to Bureau of the Mint)
Sec. 3. The index number, deviation percentage, and all the data from which they are computed shall (unless delayed by unavoidable causes) be transmitted by the Computing Bureau to the Bureau of the Mint, within one week from the day to which the data relate.
(Calculation of the Correction of the Dollar's Weight)
Sec. 4. That the Bureau of the Mint, upon receipt from the Computing Bureau of such percentage deviation, shall forthwith calculate a percentage correction or adjustment to be added to, or subtracted from, the then weight of the dollar. Said adjustment (provided it shall never exceed the "brassage" charge of 1% described below) shall be equal to the percentage deviation.
(Proclamation Thereof)
Sec. 5. The Bureau of the Mint shall then forth-with give public notice that, on and after the day next following such notice, and until changed by further like notice under this Act, the number of grains of standard gold so computed shall constitute the gold dollar of the United States; and thereupon the number of grains of standard gold in the gold dollar of the United States shall be fixed as prescribed in such notice.
(Unrestricted Issue of Certificates for Gold (Free Coinage))
Sec. 6. That after December 31, 1920, the Bureau of the Mint shall receive, subject to a "brassage charge" of one per cent and subject to such conditions and limitations as are now provided by law touching the receipt of gold bullion to be coined, all gold bullion offered to it and shall pay for the same with "gold bullion dollar certificates" described hereinafter at the rate of one dollar for the number of grains of standard gold in the dollar as then last fixed by or under this Act and (as to any balance less than one hundred dollars) in lawful money.
(Unrestricted Redemption of Certificates in Gold)
Sec. 7. That after December 31, 1920, the Mint Bureau shall receive all gold bullion dollar certificates tendered to it and shall forthwith pay for the same, dollar for dollar, in standard gold bars at the rate of one dollar for the number of grains of standard gold in the gold dollar of the United States (as fixed by or under this Act for the time of such receipt) and (as to any balance less than five ounces of standard gold) in lawful money.
(Details)
(Conversion of Coin into Bullion)
Sec. 8. That after the passage of this act no gold coin shall be struck by the United States. The Secretary of the Treasury shall provide, by rules and regulations to be issued within three months after the passage of this act, for the conversion before January 1, 1921, of gold coin of the United States owned or acquired by the United States into bars of standard gold each containing not less than five ounces, and for like prompt conversion of all like gold coin thereafter acquired by the United States.
(To facilitate the withdrawal of gold coin from circulation into the Treasury through the Federal Reserve and National Banks)
Provided: That the United States, under such rules and regulations as the said Secretary may prescribe, shall receive all standard gold coin of the United States offered to it and pay for the same in lawful money at the rate of ten dollars and one cent of lawful money for every ten dollars of standard gold coin so offered from the date of this act to December 31, 1920, inclusive, and at the rate of one dollar for every dollar of standard gold coin offered to it thereafter. Such payment shall be made in the gold bullion dollar certificates herein authorized and (as to any balance less than one hundred dollars) in lawful money.
(Conversion of Old Certificates into New)
Sec. 9. That within three months after the passage of this act the preparation, issue, and paying out by the United States of present gold coin certificates shall cease. For all gold coin certificates then owned or thereafter acquired by the United States there shall be substituted, dollar for dollar, gold bullion dollar certificates certifying that
"the United States of America will pay the bearer on demand $100 in standard gold bars of not less than 5 ounces each and any smaller balance in any lawful money."
Upon such substitution such gold coin certificates shall be destroyed.
(To accelerate said correction at the start especially through the Federal Reserve and National Banks)
Provided: That the United States, under rules and regulations to be prescribed by the Secretary of the Treasury, shall receive all gold coin certificates offered to it and pay for the same in lawful money at the rate of ten dollars and one cent of lawful money for every ten dollars of gold certificates so offered from the date when their issue ceases to December 31, 1920, inclusive, and at the rate of one dollar of lawful money for every dollar of such certificates so offered after December 31, 1920.
(Government Gold "Reserve" and "Surplus")
Sec. 10. That the Secretary of the Treasury shall divide all the gold against which gold coin certificates and gold bullion dollar certificates are outstanding at 3 A.M. January 1, 1921, into two parts, one part to be known as the "reserve" against outstanding gold bullion dollar certificates and equal to 50% of the value of the gold certificates then outstanding and the remaining part to be known as the "surplus," in excess of said reserve.
This remainder or "surplus" shall be forthwith transferred to the general fund of the Treasury as the initial profits of the new system.
The "reserve" shall be maintained daily, as nearly as possible at 50% of the gold bullion dollar certificates outstanding from time to time.
If, on any date, the reserve falls short of 50% it is to be restored by withdrawing from circulation and cancelling gold bullion dollar certificates.
If, on any date, the reserve exceeds said 50% it is to be restored by issuing, and putting into circulation, the requisite number of new gold bullion dollar certificates.
The Secretary of the Treasury is authorized to make said withdrawals of certificates from circulation by withdrawing from the Government deposits in National Banks, and to issue certificates and place them in circulation by adding to those deposits.
(Certificates Available for Bank Reserves)
Sec. 11. That all provisions of existing banking laws of the United States regulating the holding of gold reserves, including reserves of any Federal Reserve Bank, National Bank, or other bank, shall be deemed to be satisfied by such holding of gold bullion dollar certificates.
(Legal Tender)
Sec. 12. (a) That gold coin of the United States shall not be a legal tender in payment of debts falling due after December 31, 1920.
(b) That all debts, public and private, falling due after December 31, 1920, including debts theretofore created and expressed in dollars of "gold coin of the present standard of weight and fineness," or expressed in words of like import, shall be payable in standard gold bars at the rate in grains per dollar fixed by or under this Act for the time when each debt falls due, and the balance, if any, less than five ounces, in lawful money. Such standard bars shall be lawful money and a legal tender for this purpose.
(Publicity)
Sec. 13. The Computing Bureau shall, as promptly as possible, make public in suitable public documents all the pertinent facts and figures concerning the calculation of the index number and its percentage deviation from par, including the market quotations for the constituent commodities. The Mint Bureau shall likewise make public its findings as to the adjustment of the dollar's weight.
(Financing the Administration of This Act)
Sec. 14. That a sum equal to the initial profit as defined in Sec. 10, or so much thereof as may be necessary, is hereby appropriated and is made available until expended as the Secretary of the Treasury shall direct for all expenses necessary for the administration of this Act; and the Secretary of the Treasury is authorized to use the receipts from time to time from the "brassage charge" as defined in Sec. 6, for the same purpose.
(Future Revisions of Index Numbers)
Sec. 15. That immediately after the data of the census of 1920, and other subsequent censuses respectively, are available, the Computing Bureau, from such data and the best other available data, shall revise the list of selected commodities and designate a revised composite "goods-dollar" by the same method as hereinbefore described and such that, at the moment of revision, the value of the new or revised goods-dollar shall be equal to that of the old.
(Penal Code Amendment)
Sec. 16. That Section 147 of the Penal Code approved Mar. 4, 1909, defining "obligation or other security of the United States" is hereby amended to include the gold bullion dollar certificates hereby authorized.
(Repeal of Former Acts)
Sec. 17. That all Acts and parts of Acts inconsistent with this Act are hereby repealed. ·······
The above Act assumes that a reasonable banking system, such as our Federal Reserve System, already exists under which deposits subject to check will be kept in some reasonable relation to bank reserves.
The Federal Reserve Board could assist in the prompt and efficient operation of the new system by having due regard to the rise and fall of the Index Number, as suggested by Mr. Paul Warburg. This would help its adjustment of the rate of discount and its general loan policy to be such as to keep the volume of individual deposits subject to check approximately proportional both to bank reserves and to the Government gold reserve against gold bullion dollar certificates.
- ↑ It will be noted that, if gold is depreciating, the value of the gold reserve diminishes and taxation (or other financing) is required to keep it up to 100%. Under such circumstances the Government is in the position of the holder of a perishable commodity. Its gold is like ripe fruit spoiling on its hands and the Treasury suffers a loss accordingly. It taxes the public to provide for the depreciation.
The loss from gold depreciation is not, however, due to stabilizing the dollar and maintaining the reserve. The same loss, in some form, occurs whenever gold is depreciating and whether or not the dollar is stabilized. Under our present system the loss falls on the individual holder of gold certificates instead of on the Government Treasury. Every dollar of these certificates now in our pockets shrinks in purchasing power whenever gold depreciates. To stabilize the dollar simply affords a specific measure of this loss, and the operation of maintaining the reserve translates that loss into taxes.
The same principle applies to the opposite case. Under our present system, when gold appreciates every individual holder of gold certificates receives an increment of value. The gold certificates grow in value in our pockets. Under the system of a stabilized dollar, and a constant 100% reserve, the Government Treasury would reap this advantage and bestow it back on the public by lightening, by that much, the tax burden.
Thus, maintaining the reserve constant at 100% merely changes the form of the gain or loss always involved when the gold in existence varies in value. Any gain or loss, under the stabilization plan, would simply be more conspicuous than at present, entering as it would into Government accounts.
Such gain or loss must, of course, not be confused with the gains and losses of contracting parties which would be annihilated altogether by stabilization. (See Appendix II, §I, J.)
- ↑ The weight of the gold bullion dollar, at any time, may be called the redemption-weight of gold, i.e. the weight of gold in which a gold bullion dollar certificate can be redeemed. The amount of gold which must be deposited at any time for a gold bullion dollar certificate may be called the deposit-weight of gold (corresponding to the present mint-weight). This exceeds the redemption-weight (i.e. the "dollar") by the brassage fee. If this fee be 1%, reckoned on the dollar's weight, and the dollar's weight (in pure gold) were 23.22 grains as at present, this fee would be .2322 grains. Hence the depositor of gold, in order to receive a dollar of certificates, would have to deposit not only a dollar of gold bullion (23.22 grains) but a "brassage" fee of .2322 grains besides, or 23.4522 grains in all. In other words, while the redemption-price would be $20.67 an ounce (i.e. 480 grains in an ounce ÷ 23.22 grains in a dollar) the depositor of gold would receive a deposit-price (corresponding to the present mint-price) not of $20.67 an ounce, but of 480 ÷ 23.4322 or $20.47. Under this system of terminology the dollar and the official price of gold are defined in terms of redemption, not of deposit, which latter involves the brassage fee as well.
- ↑ This is shown in figures in § 9 below.
- ↑ If the view which has been given (that such bull speculation would be too trifling to require any special provisions against it) were incorrect,—if, after all, the Government might be seriously embarrassed,—such a raid on the Treasury could be altogether avoided by a special proviso: the price of gold could be further restricted, so far as any upward change is concerned, so as not to be raised more than, say, one half of one per cent a month, i.e. at so small a rate, at most, as to be more than offset by the interest, etc., which the bull speculator would have to carry.
This restriction would only slightly hamper the stabilizing process; for it is only seldom, and never for long periods, that gold has appreciated relatively to goods more than one half of one per cent a month. This safeguard is mentioned, however, merely to meet completely all possible objections, however far-fetched or imaginary. Such a proviso would, I believe, be as superfluous as it would be innocuous.
If, as I suggested in "A" above, the brassage were 1% and the adjustment period two months the terms of the restriction just mentioned would be met anyway.
- ↑ See, e.g., tables of silver and rupees in relation to gold in Financial and Commercial Statistics for British India, Calcutta, 1895, p. 353. After the closure of the mints in June, 1893, the first figures available, which were dated about a month and a half after that event, show a marked appreciation of the rupee.
- ↑ The average index number at the five dates was 195 (on the basis of 100 for 1913), calculated as follows:
1st Liberty Loan June 1917 index number 184×2.0 billion = 368 2d Liberty Loan Nov. 1917 index number 182×3.6 billion = 655 3d Liberty Loan May 1918 index number 191×4.1 billion = 783 4th Liberty Loan Oct. 1918 index number 204×7.0 billion = 1428 5th Liberty Loan May 1919 index number 200×5.3 billion = 1060 weighted average index number 195×22 billion = 4294 - ↑ This will be found in the Quarterly Journal of Economics, February, 1913, pp. 213-235. It is interesting to observe that Simon Newcomb, one of the earliest writers who anticipated me in formulating the plan, also suggested this feature by which gold coin could be retained. See North American Review, September, 1879.
- ↑ This power is, I understand, well recognized in a general way although no case precisely like that here considered seems to be on record. The nearest cases were, apparently, the famous legal tender cases in reference to which the Supreme Court certified the right of Congress to make United States notes legal tender for the payment of debts contracted prior to the legislation. The legal tender act, it is true, related only to contracts to pay money generally and not to contracts to pay a specific kind of money such as "gold coin of the present weight and fineness." But Justice Bradley (12 Wall. 457, 566, 567) said: "I do not understand the majority of the Court to decide that an Act so drawn as to embrace in terms contracts payable in specie would not be constitutional. Such a decision would completely nullify the power claimed for the Government."
- ↑ This assumes the existence of the "indefinite" reserve system (see Appendix I, § 1, G).
If the "definite" reserve system (see Appendix I, § 1, B and F) is maintained, inflation of certificates would be impossible; for as fast as the issue of certificates went on, their redundancy and backflow would require their cancellation. Other forms of inflation, such as deposit inflation, would, however, still be possible. These would have the effect (through raising prices and weighting the dollar) of decreasing both the gold reserve and the certificates in unison. The result would be not to weaken the Government gold reserve as against certificates, but to weaken all other, e.g. bank, reserves held in these certificates. The ultimate disaster, which would still overtake continued inflation, would then consist in a cleavage, not between gold and certificates, but between these two and the other forms of money and credit based on them. - ↑ Federal Reserve Bulletin, December, 1918, p. 1178.
- ↑ For further discussion see Appendix I, § 4.
- ↑ The common crude idea that a mere difference in the purchasing power of monetary units of two countries will help exporters in the country with the "cheaper" money and hurt importers is, of course, absurd. If this idea were correct, there would be an enormous stimulus to the flow of goods from Mexico to the United States and check to the flow from the United States to Mexico because the Mexican dollar is only half our dollar. Naturally that difference between the dollars is fully taken into account. It is only when the relation between the two is disturbed and before the new relation has been fully taken into account that exporters and importers are affected, even in a slight degree.
- ↑ We may, to fix our ideas, consider this interval between successive adjustment dates to be two months. But its absolute length affects neither the argument nor the calculations.
- ↑ In all the calculations of this section it is assumed that either the mint price rules the market all the time or the redemption price rules it all the time. If, or when, the market price shifts between the two, in the manner discussed in Appendix I, § 2, the results would be slightly different, as can readily be calculated.
- ↑ This column also shows (by subtracting 100) the deviation from par and the adjustment of the dollar's weight, which is equal thereto.
- ↑ Since the lag is beyond our regulation, while the adjustment interval is what we make it, the lengthening of the lag in terms of adjustment intervals really means, in practice, the shortening of the adjustment interval in terms of the lag.
- ↑ This column also shows (by subtracting 100) the deviation from par and the adjustment of the dollar's weight (except as this is limited to 1% by the brassage).
- ↑ This column also shows (by subtracting 100) the deviation from par and (by subtracting 100 and dividing by 2) the adjustment of the dollar's weight. The latter is also always equal, numerically, to its influence, given in the second column.
- ↑ This column also shows (by subtracting 100) the deviation from par and the adjustment of the dollar's weight, which is equal thereto.
- ↑ This column also shows (by subtracting 100) the deviation and (by subtracting 100 and multiplying by 2) the adjustment.
- ↑ A study of this sort has been made by Professor J. M. Clark in his able paper "Possible Complications of the Compensated Dollar," American Economic Review, September, 1913, pp. 576-588.
- ↑ Except that, beginning with January, 1913, I have substituted the special index number of responsive commodities described in Appendix I, § 3. The difference in results between the two index numbers is not great.
- ↑ This column also shows (by subtracting 100) the deviation from par, and the adjustment (except that this is limited to 1% by the brassage).
- ↑ It should be remembered that this stabilization of wholesale prices would carry with it the stabilization of retail prices as explained in Appendix I, § 3. In fact, as retail prices change sluggishly, their index number would doubtless keep even closer to par than that of wholesale prices.
- ↑ This close conformity to par is maintained in spite of the fact that, as already noted, the "lag" assumed is much greater than we may reasonably believe is the truth. In fact the conformity would be close even if the lag were much longer. Assuming that the influence of each adjustment came even a full year later, the index number would (up to the close of 1915 when the great influence from the war began) keep within 3% of par half of the time, within 5% two thirds of the time, and within 10% nineteen twentieths of the time.