The story [of electronic money as a way to end recessions quickly and end inflation forever] really goes back to 1932, in the middle of the Great Depression. A guy named Robert Eisler realized that they could get out of the great depression if they distinguished between bank money and paper money.

That is how I began my remarks at the Cryptocurrency Conference I participated in this past Tuesday. I learned about Robert Eisler from posts by Willem Buiter, in which Willem Buiter gives the modern theory of eliminating the zero lower bound.

One of the audience mentioned Silvio Gesell, who is much better known because John Maynard Keynes talked about him in his book The General Theory of Employment, Interest and Money. (You can see the passage in my post on Silvio Gesell here.) But I see my proposal for eliminating the zero lower bound as closer in spirit to Robert Eisler’s proposal, first because he emphasizes an exchange rate between bank money (=electronic money now) and paper currency, and second, because he envisioned the bank money as having a stable value in relation to the goods and services that people buy. By contrast, instead of being comfortable with an exchange rate between bank money and paper money, Silvio Gesell proposed a system that encouraged people to buy stamps to put on the depreciated paper money to bring it up to par, and Silvio put less stress on having at least one type of money that maintained a stable value in relation to goods and services.  

Needless to say, Robert Eisler’s proposal is far from identical to mine. The biggest difference is that Robert Eisler imagined international fixed exchange rates in bank money coexisting with flexible exchange rates between bank money and paper currency. Also, Robert Eisler is far from clear about the necessity to keep the level of economic stimulus just right if the bank money is to keep a stable value. He writes as if using paper money price indices to set the exchange rate would be enough to yield a stable value for the bank money. Finally, he does not emphasize as I would the great importance of encouraging everyone to use the bank money as the unit of account. Indeed, in what he writes, he does not himself always use the bank money as the numeraire, and so sets a bad example in relation to what I consider the crucial unit of account function of the stable-value bank money. Nevertheless, in the most important respects, Robert Eisler anticipated the kind of monetary system I advocate.  

In a tweet, I called Robert Eisler the “grandfather of emoney monetary policy.” The father of emoney monetary policy is Willem Buiter, who first explained the principles clearly. (Also see the children’s fairy tale about Willem Buiter’s role.) 

Robert Eisler’s book Stable Money is not available on Amazon, and even the University of Michigan library system had to ask for an interlibrary loan from the University of Nebraska library in order to put a copy in my hands. I consider what he has to say important enough that I scanned the key chapter, available at this link. For those who prefer a modern font, I also typed out the key passage, comprising most of that chapter, which lays out Robert Eisler’s proposal for a new paper currency policy. (I put footnotes in square brackets at the appropriate points.) 

Here is Robert Eisler, Stable Money: The Remedy for the Economic World Crisis. A Programme of Financial Reconstruction for the International Conference 1933. Chapter XVII: Compensating the Effects of Inflation on the Cost of Living, pp. 232—247:

The present writer himself would condemn any scheme of monetary reorganisation which began by raising the purchasing power of one group at the expense of another. [Cp. Sir William Beveridge, Unemployment, p. 416: ‘The absorption of the unemployed in a temporary boom could probably be achieved very rapidly by a government prepared for inflation, but the inevitable after-effects of such a policy rule it out.’]

Fortunately, the proper method of for avoiding this and for counteracting automatically and continually the above enumerated evils of inflation is quite well known ever since, in 1747, Massachusetts Bay Colony introduced a simple ‘tabular standard’ based on the prices of wheat, meat, leather and wool in order to compensate for the effects of the inflation of the ‘Colonial Notes’ circulation on the revenues of its creditors, officers and soldiers. [Cp. Willard C. Fisher, ‘Tabular Standard in Massachusetts History,’ Quarterly Journal of Economic, 1913, pp. 415-417.]

Nothing is necessary but the introduction of a new, very simple monetary law in all countries willing to participate in the proposed currency expansion scheme—a law extending the well-known principle of ‘index-wages’ to all existing contracts in terms of money.

The legislation would provide that every monetary obligation in force at a given date—all contracts concerning wages, salaries and appointments, all laws fixing taxes, rates and duties, all stipulations concerning insurances, bank deposits or credits, loans or mortgages, rent or interest, all deeds of sale, all commercial bills, etc.—should be carried out with due regard to the actual purchasing power of the national money at the time when the payment stipulated is effected, the temporary purchasing power of the money being ascertained by means of index-numbers representing the average retail price [Only retail prices affect the owners of contractual income whom we want to protect (cp. On the whole index-problem, below, pp. 248-271).] of a fair number of consumable goods, the cost of rents, [Rent-restriction, originally intended to compensate the effects of war-time inflation and labour shortage in the building trade, which still exists in most countries, becomes superfluous as soon as proper weight is given everywhere in the calculation of the cost-of-living index to the cost of adequate housing. The importance of this as an element in the solution of the housing-problem is obvious.] rates, taxes, local transport, and elementary education at that particular moment in terms of the appropriate monetary unit. Such legislation would be tantamount to redefining the monetary unit of the U.K. or of the U.S.A. in case of their joining the proposed monetary federation. The following is a tentative draft of the proposed law:

The monetary unit of the currency of the United Kingdom/U.S.A. is the pound sterling/U.S.A. dollar.

The pound sterling/U.S.A. dollar is the Bank of England/Federal Reserve Bank note which was equivalent to 122.24719 grains/23.22 grains of fine gold at the London gold price of the 18th of September 1931.

The pound sterling/U.S.A. dollar note is legal tender to the extent of its actual purchasing-power.

The term ‘actual purchasing-power’ is defined as the reciprocal value of the cost-of-living index obtaining on the day of payment.

The cost-of-living index is calculated on the basis of the average costs of the following commodities: … of local transport, of rents, taxes and rates, and of the expenses for elementary education.

Through such a law the two functions of money—its use as a medium of exchange (legal tender) and its use as a means of accumulating capital (deferred purchasing power) or contractual income for the future—would henceforward be separated more completely than now. Even to-day bank-money and securities are generally used for the purpose of accumulation, and the keeping of considerable amounts of ready cash (hoarding of gold as well as of paper) is penalized by the loss of interest on that sum of money and is therefore avoided as much as possible by every intelligent steward of moneys—except in periods when a progressive slump in prices promises to compensate the hoarder for the small interest he might get on the over-full market for short loans. During periods of credit and currency expansion the penalty for letting money lie idle is aggravated by the certainty of seeing it constantly depreciation.

Under the new system there would be two sorts of money: (1) legal tender, called a pound or a U.S. dollar of ‘current money’ [The term ‘current money’ is well known to legal practice and theory as the equivalent of ‘legal tender’ as opposed to divisionary money which need not be accepted in unlimited quantities.] or money proper (£ cr. or $ cr.) and (2) bank or contract money of account, called a pound or dollar banco [The history of the term is told in W. F. Spalding’s article ‘Pound Banco’ concerning the author’s address to the Parliamentary Finance Committee in The Times Trade and Engineering Supplement of February 27th, 1932. The expression goes back to the twelfth century: the idea of a stable money is an ancient Roman invention (cp. Eisler, Das Geld, Munich, 1924, pp. 201 and 211f.)] (£ bo. or $ bo.). Money banco would be obtained by concluding a contract about a future payment of money proper, or by depositing ‘current money’ with a bank or similar institution. Current money would be exclusively used for small transactions between persons not well know to each other or not in possession of a bank account, especially for the payment of wages, transport fares and occasional retail purchases. All other payments would be effected by means of bank-money, that is by cheques or transfers of money banco. All prices in catalogues of shops selling goods of which the price does not vary much (e.g. books, clothes, jewellery, cigars, tobacco, wines) would be marketed in money banco. The index multiplicator of the week would be affixed to the desk of the cashier, who would calculate by means of simple multiplication or conversion tables published in the Sunday papers the sums due in ‘current money’. This is how retail business, inn and hotel-keeping was done in Germany at the height of the inflation in 1923, when the ‘stable mark’ was introduced as a money of account alongside the paper mark used as ‘current money’.[This simple expedient was criticized as ‘a nightmare’ by Prof. Edwin Cannan in his review of the author’s book, This Money Maze in the Economic Journal of 1932. We should like very much to put it to the vote of all shop assistants, however, whether they would object to this little extra work if they could thus be permanently freed from the menace of unemployment.] All bank and business accounts would, however, be calculated without any additional complication in money banco, current money being exchanged against bank-money by special tellers[Keeping special ledgers for ‘cash bought and sold’ as the money changing cashier of an English Cook’s office or of a continental bank.] only in so far and as soon as a client wanted to pay in or draw out current money. In each country the external exchanges between the national and foreign bank-moneys would be pegged,[See above, Chap. xiv.] but a variable internal exchange rate between current and bank money would be determined by the cost-of-living index.

The distinction between current money and ‘pounds banco’ and the use of an Italian word seem to arouse the distrust of the present inhabitants of Lombard Street, although they have no objection to the Bank of England being owned by a merchants’ Compagnia.[Abbreviated, Compa. on the notes of the Bank of England.] Mr. Alec Wilson, the well-known lecturer and propagandist of the League of Nations Union, has therefore suggested a simple and perhaps, to Englishmen, more acceptable way of distinguishing the unit of current money from the unit of ‘money-of-account’. The British public is quite accustomed, ever since the memorable 21st of September, 1931, to seeing in the popular papers ‘how many shillings the £ is worth’ on a given day. In these most misleading statements,[Cp. Above, p. 152] which are intended to keep alive the ‘gold mentality of the ignorant masses’, the equation between a £ and so many shillings is meant to tell the reader how many hypothetical ‘gold shillings’—if such a gold coin of 122.245/20 = 6.11235 grains did exist—an inconvertible paper pound would buy in London, Paris or New York.  In an analogous, but more realistic way we might henceforward decide to call the units of current money—the medium of exchange for petty trade and for paying out wages—‘shillings’ and ‘pence’ as usual, or ‘shillings’ and ‘pence cash’, reserving the term ‘pound’ or ‘pound cheque’ for the unit of bank-money used for all the more important transactions. The present £ notes would have to be withdrawn and exchanged against 5s., 10s., 20s., 100s., 200s. and 1,000s. notes corresponding to the continental type of bank-notes. Fractions of the unit of the bank-money of account would not be expressed in shillings but written out as decimal fractions of the £–e.g. £4.2 (meaning £4 4s. at the index 100).

TWENTY SHILLING

This twenty shilling Note of the Bank of England equivalent to one pound sterling, that is 122,247 grains of fine gold at the legal price of the 18th of September 1931, is legal tender to the extent of its actual purchase-power. The figure determining the purchase-power of this note on the day of payment, being the reciprocal value of the cost-of-living index in the United Kingdom of Great Britain at that date, will be published at least each Sunday by the British Board of Trade. The Bank of England promises to exchange its notes at sight against gold ingots of 400 oz. standard fineness at the gold-market price of the day and to buy gold at any time and in any quantity at the legal price of the 1st of January 1931, that is £1,500 13s. 4d. bank-money[It is equally possible to stabilise gold at the lowest market price of the day of the ratification of the international agreement printed below, App. V, and that is, at a figure round about £ 6 bank-money for the fine ounce (see above, p. 165, note 1). In this case speculation would be diverted from the commodity market to the gold-market, which would be desirable from the consumer’s point of view.] for the standard ingot of 400 ozs. Troy,

Act of Parliament of the __th of ______1933.

The pound would thus become again what it was when it meant a pound weight of silver coins, that is, a purely theoretical unit of account.[Curiously, Mr. Alec Wilson has hit, without knowing it, upon a system of currency expansion which is almost as old as the conception of the money banco. Cp. The author’s history of money Das Geld, p. 181, on the sixteenth century (unadulterated) Rhenish goulden, equivalent to more and more (debased) divisionary money (kreutzer): this was the oldest attempt to have a full-weight legal tender ‘principal currency’ alongside of a slowly depreciating expansible token currency for everyday trading.] This would be desirable because the proposed scheme has often been erroneously criticized as an attempt to introduce a double currency, which would bring ‘Gresham’s law’ into play. If Mr. Wilson’s suggestion is adopted, it will be seen that, as before, only one currency, consisting of shilling notes and shilling and pence token-coins, exists and circulates, the £ being merely the theoretical stable unit measuring credit, i.e. bank-money. Gresham’s law would immediately become operative if £ notes (£ bo.  notes of ‘stable’ value) were issued alongside of 10s., 20s., 100s. (or s. cr.) notes. In this case the two kinds of currency would not circulate together, the £ bo. notes being hoarded and the £ cr. Or shilling notes being eventually repudiated. Nothing of this kind can happen as long as the privileged banks do not issue anything but ‘current’ money proper, i.e. shilling notes. 

If the use of the decimal fractions were found too radical an innovation, the graphical expression £4,4/20 could be adopted, an odd penny being written as 1/240, a halfpenny as 1/480, a farthing as 1/960.

The proposal is, then, that we should have cheques, bills, promissory notes, IOUs, receipts, and any other documents of a similar nature expressed in pounds, one £ being equivalent to a number of shillings and pence of actual currency directly indicated by a cost-of-living number. If the use of decimal fractions were rejected, the basis of the British cost-of-living index would no longer be called 100—the index figure rising per cents to 101, 102, 103 …, but the scale would start at 240, the figure rising by 1/240 (corresponding to a penny in the £), and if necessary by 1/480 or 1/960, a halfpenny or a farthing in the £.[Englishmen are already used to reckoning rates and taxes by so many shillings and pence in the pound.]

A shopkeeper would keep over his cash registering machine a notice containing a space for inserting a statement of the currency standard at the time. It would be something like this:

For the convenience of our regular customers our prices are marked in £ and fractions of £. These figures should be used when writing out cheques and are put down by us in all our accounts and bills. Clients paying in cash will kindly note that to-day the value of the £=21s. 6¾d.

Beside this placard a copy of the corresponding printed conversion table[A specimen is given in App. IV.  App III shows a specimen butcher’s bill for two weeks.] might be affixed for the public to check the calculation of the cashier or vendor.

The price of single tram-, bus- and underground railway-tickets could not of course be adjusted to slight percentage variations of the purchase-power of current money. They would be sold—as they are to-day in Paris—in little folders, each containing a dozen penny ride tickets for 6 francs = 1/20 £, single tickets being sold at a considerably higher price (e.g. 1½d. for a penny ticket). Since local transport costs are included in the cost-of-living index,[Above, p. 233, §5] it does not matter if the single ticket price has to be raised more than absolutely necessary in terms of farthings.[During the German and French inflation all forms of transport were run at a loss, because fares were always adjusted too late to the rate of monetary depreciation. Under the proposed system the profits of the local transport services would be somewhat higher than normal, this super-profit compensating the outlay on new ticket-selling and money-changing machinery. At first a larger issue of farthings coinage would probably be convenient for purely psychological reasons, in order to avoid such small prices as penny fares being raised immediately by halfpennies, i.e. by 50%, because no smaller coins are available. ]

If groceries, butchers’ meat and similar goods are sold on credit, the shillings and pence figures may provisionally be put down in the ledger with the date and the index (of the week or of the day), to be added and converted into pounds at a convenient time by the book-keeper who writes out the customers’ bills. 

Current money will, of course, slowly but constantly[Not, of course, for ever—but until the optimal price-level is reached. See below, pp. 280f.] depreciate through the expansion of credit and currency necessary to meet the requirements of increasing production and consumption, the rate of increase being no longer limited except by the necessity of expanding the expenses of the various countries in due proportion to their actual spending power as measured by the volume of their present gold-standard currencies and by the physical limits of production. This will naturally force commercial and clearing banks to diminish as far as possible their cash reserves by increasing their balances at the National Banks,[These balances would, of course, be expressed in terms of bank-money (£ bo.) and convertible into actual cash at the index-rate of the day. The Macmillan Report (p. 146, § 343) has already pointed out that by establishing depots of its notes in various parts of the country the Bank of England could considerably increase the balances which the joint-stock and clearing banks keep with its Banking Department and diminish the need for till-money kept by the banks in their own vaults. The arrangement suggested below (p. 240, n. 3) and the increased circuit-velocity would probably enable the bank to provide the necessary amount of current money without raising the maximum limit of note issue above the figure of £400,000,000 recommended by the Macmillan Report (l.c.). On the other hand, the joint-stock banks certainly would not care to keep to the old ratio of 1:11 between deposits and ‘cash in hand’, but would substitute ‘balances with the Bank of England’ for most of their cash.] and all private people to spend or to deposit current money as quickly as possible. In this way the circuit velocity of currency, which is now one of the most intractable actors of monetary instability, will always be maintained at its maximum value, i.e. stabilized as far as possible, and the maximum of available credit will be automatically placed at the disposition of traders and producers. 

Currency reformer like Sylvio Gesell, the ancestor of the ‘Free Credit’ school, have tried to obtain this desirable result by the imposition of a weekly stamp-duty on bank notes. Leaving aside the foolish idea that such a duty could annihilate the receipt of interest on loans, and disregarding the intolerable inconvenience of having constantly to stick stamps on every bank-note, we may observe that a tax on currency simply produces a constantly depreciating money, which is the characteristic feature of inflation. The idea of a ‘dwindling’ monetary unit was first conceived in Brandenburg in the fourteenth century.[See Eisler, Das Geld, p. 117, on the ‘shophel’-pennies, of which a growing quantity (12 in the 1st, 13 in the 2nd, 14 in the 3rd, 15 in the 4th quarter and 16 in the new year) were equivalent to a shilling silver. About ordinary inflation being a crude form of taxation, see above p. 193, note 1, and J. M. Keynes, Monetary Reform, pp. 42.f] In 1927 the ‘Reichsbund Deutscher Technik’—the German political organization of civil engineers—proposed[Absatzstockung, Arbeitslosigkeit und ihre Beseitigung, Schriften d. R. D. T. Heft II, p. 14.] to introduce a ‘stable mark’ (Festmark) and a constantly (slightly) depreciating ‘currency mark’ (Zahlmark), the growing rate of exchange between the two being fixed by the government with regard to the price-index and with the purpose of flattening out the trade-cycle.[L.c., p. 22.] The author of this idea thought that you could have deposits of ‘current money’ on running account with the banks, failing to see that this contradicts the idea of a stable ‘contract money’ of account, since bank deposits constitute a contract. But otherwise the idea of preventing hoarding by ‘mild’ compensated inflation is absolutely sound and can be realized in the simplest way by the measures suggested above.

The slight automatic profits which the Central Banks will make at the expense of all those obliged through their particular trades to keep a certain amount of ready cash in their tills, but prevented by all kinds of accidental circumstances from immediately exchanging their surplus cash for bank-money,[If indices were published each Sunday, all loss could be avoided by depositing surplus cash at the banks every Saturday, as soon as shops were shut. Commercial banks would deposit their surplus cash at the next Central Bank branch office. In each town one big bank or post office could function as such a branch, the cashier being preferably a Bank of England official, paid by the joint-stock bank or the post office.] will serve to cover the cost of printing the bank-notes, the remaining surplus being heavily taxed or entirely appropriated by the state, which is—though [sic] conferring on the Central Bank the privilege of issuing notes—the ultimate source of this profit. 

The power of bank-money to purchase consumable goods would be absolutely constant, even in periods of credit or currency expansion.

Wages and salaries, interest and rent being fixed in bank-money, i.e. subject by statute to regular increase corresponding to an eventual rise in the index figures, the (nominally) rising price-level will in no way diminish the real wages of the manual and intellectual worker or the income of the rentier, nor will it increase the wages-bill of the employer.[It does not matter in the least to the employer how much current money the wage-earner will cash at the bank for his weekly wage-cheque of, say, £3 or £4. The employer buys his materials, etc., sells his product, and pays his wages with cheques—i.e. bank-money; the cost-of-living index does not interest him one way or another, except in his capacity as a householder.] Conversely, the real cheapening of production through the increase of output and the proportionate diminution of overhead charges made possible by increased consumption would accrue in the shape of higher profits to the entrepreneur, so that he[On this question cp. Also above, p. 29, last §.] will be able to spend or to reinvest on a larger scale and thus create further opportunities for employing labour in the production of new capital goods and in the luxury trades. 

In terms of this bank-money the price of each individual commodity or service will vary as before in comparison with the price of any other or in relation to the average price of all commodities. But this average—the so-called retail price-level of consumable goods—is practically stabilized within very narrow limits, that is to say, it can only vary during the interval between the publication of two successive cost-of-living indices. Since the cost-of-living index is a retail price index based on figures which move much more slowly than the ‘reagible’ wholesale prices, the movements which are possible between successive index-calculations based on the market prices of each Saturday—the day following the Friday evening wage-payments in current money or in wage-cheques—are practically negligible. 

Readers who remember the vertiginous price-movement on German, Austrian and Polish retail markets in the heyday of inflation may be in inclined to imagine that equally rapid price-movements would take place under the influence of a compensated currency-expansion. They forget that these price-movements were caused by a demand for commodities which was artificially stimulated because people were prevented from converting current into stable money, that is, into claims on stable purchasing power. Under the new system price-rings of traders will still be able to fleece the customer to a certain extent, but the practice will be limited by the competition of outsiders and of the co-operative stores.

If the reader will take the trouble to look at the curves describing the typical effects of a classic currency inflation,[Eisler, Das Geld, Munich, 1924, fig. 128, p. 252 (our pl. 8). Only the later part of the curves shows the hyperbolic shape determined by the ‘quantity law’, according to which a rapid initial depreciation should gradually slow down. At the beginning, the movement follows, on the contrary, Fechner’s law of psychological effect: beginning slowly, it gains gradually in speed and momentum.] he will find that prices are rising in a rather sluggish way at the beginning and that a rapid depreciation of the currency does not set in until the late phase of the movement.

If the slow and unimportant initial depreciation of current money is compensated from the start by the facility of converting current money either into commodities (which are depreciating not only because they are perishable and cannot be kept free of ‘carrying costs’, but also because of the rapidly increasing production) or into stable bank-money, it is highly probably that the movement of nominal prices will be slow and rather regular. If the contrary should happen through the panicky reaction of people who remember the days of former inflations, it would not matter, because the only untoward consequence of a rapid depreciation of current money would be the need to publish bi-weekly or even daily index figures for a short time until the public got accustomed to the new mechanism. 

On first consideration, the expert reader will be inclined to object that the use of a money of account alongside current money is not a new invention, and that all previous attempts to introduce such a double standard have been failures, as is proved by the fact that it has always disappeared after a short time. At the height of inflation in 1923, Germany tried to introduce a money-of-account, variously called Festmark or ‘gold-mark’, and attached to the U.S.A. dollar. In a similar way Poland had in 1922 its zloty, which was at first nothing but a money-of-account equivalent to the Swiss franc.

Quite recently (14th August 1931) Hungary has introduced, in view of the beginning of the depreciation of its ‘pengo’ notes a theoretical ‘gold pengo’ to serve as standard of all private and public monetary contracts and obligations: in this way it was hoped to maintain Hungarian gold prices and contractual obligations on a par with those in gold-standard countries.[The law has proved mere ‘eye-wash’ because the Hungarian National Bank maintains a pretence of being ‘on par’ by means of exchange restrictions.] All these various moneys-of-account were and are, however, entirely independent of the commodity price-level of the country.

If gold appreciates internationally, Hungarian prices must all fall further in gold-pengos and the total burden of contractual claims established in the past on a gold-standard basis will continue to grow more and more unbearable. If gold were to depreciate internationally, Hungarian gold-pengo prices would rise. If the national gold claims are not constantly convertible by the National Bank into foreign gold-exchange, such claims will be sold abroad at a discount. When this discount becomes known within the country, the legal gold claims de facto depreciate and prices begin to rise, even in terms of national gold claims, in the same proportion as they would rise in paper money according to the quantity law. As soon as this happens, gold-pengos become merely larger denominations of ordinary pengo notes, which means that they become quite superfluous. If, However, and as long as they remain convertible into gold-exchange, the elasticity of the total circulation—bold notes plus ordinary notes—is limited by the gold-exchange reserves of the country, just as if only a single currency existed.

The maintenance of legal index-wages and salaries in one country is of course impossible beyond the limits set by the competition of other countries. Index-wages cannot be maintained nationally with the exception of those in the so-called sheltered industries. In the other industries the basic ‘gold’-wages must be reduced under the pressure of foreign competition.

It follows (a) that national compensated currency systems cannot be permanent; (b) that a national money-of-account legally defined on the basis of the gold parity with money of a gold-standard country cannot be maintained at par without sufficient gold-exchange reserves; © that therefore the stability of the general price-level, the gold value of rents, taxes, wages, etc., defined in terms of such gold notes, is no more stable than if they were defined in terms of ordinary inconvertible paper. In other words, there is no way of stabilizing nationally the retail price-level or reciprocally the purchase-power of a national money. 

The bank-moneys (banco talers, ducati di Banco,etc.), of the Venetian, the Hambourg and other early clearing banks, known to historians of the monetary system,[Eisler, Das Geld, p. 213] were nothing but bank transfers of fixed weights of metallic gold or silver deposited in the bank vaults; the banker had a private money distinct from the continually depreciated money of the various states, or rather their rulers, and comparable to the modern gold and silver certificates of the U.S.A. Treasury, which are covered by gold, dollar for dollar. These forms of bank-money were used exclusively between wholesale merchants and never used as contract-money for stabilising wages or retail prices. The purchase-power of such bullion transfers varied, of course, with the quantity of gold or silver entering into world circulation in a given period and with the circuit-velocity of the transfers. As soon as these banks began to give credits and to depart from the 100% proportion of metallic reserves, the purchase-power of their notes became subject to the quantity law and began to show all the symptoms of the ‘inherent instability of credit’.

While it is obvious that a national, as well as an international, money-of-account based on a gold parity relation is incapable of stabilising the general price-level nationally or internationally, this is not at first sight so clear in the case of a national money of account based on a tabular commodity standard of the type of the primitive system applied in eighteenth-century Massachusetts.[Above, p. 232, n. 2.] This system was tried in Russia in 1922 in the form of the so-called ‘goods rouble’,[See G. R. Hawtrey, Currency and Credit, London, 1930, p. 423; Eisler, Das Geld, p. 32.] only to be abandoned very soon in favour of a gold claim currency (the tchervoncy or ‘goulden’ notes). The explanation is very simple in this case too. In order to finance a country’s imports, accumulated claims in national money must be exchangeable without loss into monetary claims in foreign countries. Since the fixed quantity of commodities and services to which a national claim based on a tabular standard at a constant retail price[In the system proposed above, pp. 232f, the unit of bank-money is not equivalent to a quantity of goods bought and sold internationally at fixed wholesale prices, but to a ‘basketful’ of consumable goods at a fixed retail price. The transfer of such money is useful only to a tourist, a consumer wanting to acquire a fixed claim to the means for maintaining a certain standard of life in a given country, not to a trader, wanting to export or import a quantity of goods wholesale.] cannot itself be freely transferred like gold into another country such a claim must be convertible into a freely and cheaply transportable quantity of gold or gold-exchange or into a foreign warehousing warrant for which an equal quantity of goods and services can be obtained at a fixed wholesale price in other countries. If this is not the case, such a claim will be offered abroad for what it will fetch. The price in terms of gold or gold-exchange or foreign goods warrants which the seller will be able to obtain for such a claim, will fall with the growing value of the imports needed by a country and with the diminishing quantity of its exports. As soon as the foreign exchange value of such a claim has fallen below par, it will not be exchanged against gold claims, but at a discount, even within the country itself. In other words, increased exports will diminish the quantity of goods offered in the home markets and thus raise the real banco prices of all commodities. This will of course destroy the inherent essential stability of the purchase-power of a monetary claim based on the nation cost-of-living index. The internal purchase-power of such a claim will in the last instance depend on the quantity of purchase-power in gold or gold-exchange standard countries, against which it can be exchanged at home or abroad at a given moment. That is to say, the country’s money-of-account will lose all connection with the cost of living and the internal price-level and attach itself automatically to the purchase-power of foreign gold-exchange. Hence the money-of-account of a country depending on the trade with gold-standard countries cannot be anything but a given quantity of gold or gold-exchange, and nothing can be gained by the national introduction of a commodity-standard except a slight, temporary slowing down of the inevitable transfer of purchase-power from the rentier and the wage-earner to the entrepreneur, which takes place whenever credit and currency are expanded. 

In other words: there is no way of stabilising nationally the general price-level or reciprocally the purchase-power of a national money.

The first condition for the stabilization of the general price-level and of the national purchase-power of a currency is the international stabilization of the exchanges through the parallel, proportionate and synchronic development of production and consumption in the various countries by means of proportional supplementary budgets and the preliminary exchange of mutual stabilization credits.[Above, pp. 194ff.] If these indispensable conditions are realised by international treaty, the simultaneous application of the tabular standard in each country is the necessary complement of the measures stated. Since all previous experiments with index-currencies cannot be used as an argument against the possibility of organizing an expandable currency of monetary units with stable purchase-power by means of a logical co-ordination of the three above-explained measures. 

Another prima facie objection to the proposed system is based on the fact that the cost of living is naturally very different in the various countries which might wish to participate in such a monetary reorganisation.[The so-called purchase-power equation applies only to the wholesale prices of goods bought and sold internationally, not to the retail prices of local products or services. Unilever Ltd. Have had very interesting figures compiled illustrating the comparative cost of maintaining in various European countries a ‘typical’ English standard of life corresponding to English ‘earned’ incomes ranging from £500 to £3000 (‘International Middle Class Living Costs’, The Economist, November 8th, 1930).] The obvious answer is that, equally so, within each country the cost of living differs in various cities, towns and villages. In Germany, officials are paid varying ‘local supplements’ to their wages, according to whether they have to live in the capital or in a cheap provincial town. Nobody would expect the difference in the local purchase-power of money having any other effect, but that impoverished rentiers occasionally migrate from Berlin to Pasewalk. 

The cost-of-living variation cannot cause any exchange fluctuations[See J. M. Keynes, A Treatise on Money, vol. I, p. 73: ‘There is no justification even in the long run for any precise necessary or immediate relationship between changes in the rate of exchange of currencies of two countries and the changes in their Consumption Standards relatively to one another.’] in the proposed system, since international trade is carried on exclusively by means of bank-money (bills of exchange, cheques and bank transfers). It does not matter in the least if, beside the pegged exchanges for travellers’ cheques and letters of credit, a variable rate exists for the negligible business of exchanging national against foreign small bank-notes and coins. The continental use of large dollar-notes for hoarding purposes would, of course, disappear very quickly, nobody being the worse for it.

Experience has shown, moreover, that in countries with pegged exchanges, like Egypt and the U.K., the national price-levels will rise and fall together.

Leaving aside such big divergences as are bound to occur in times of war under any system and which would disturb the functioning of any international monetary standard, the differences in the parallel evolution of the various national costs of living in peace-time will not upset the stability of the exchanges any more than they do now.

As long as the exchanges are maintained at par by the above-explained procedure, it does not matter in the least whether the different countries are using exactly the same index figures for exchanging national bank-money into national current money, since nothing but bank-money will be sold and bought internationally. Nobody would think of buying large quantities of another country’s continually depreciating current money if he can always get the foreign stable bank-money by means of a cable transfer or a traveller’s cheque, or by having a bill of exchange discounted by a banker.