A MEASURE INTRODUCED in Finland in May 1955 appears to be one of the first steps taken by a banking system to introduce, in its pure form, a device that has long been advocated by theoretical economists. Most savings banks in Finland introduced at that time a new form of time deposit for their customers. In future, money could be deposited for one year, and at the end of the year the bank would undertake to credit the account with the amount that the cost of living index indicated was necessary to restore the account's original purchasing power.1 The account would bear 4¾ per cent interest, 1½ per cent below the prevailing rate on deposits without the purchasing power guarantee. Since this step was taken at a time when there was, in other countries, a rising practical interest in the device, it seems useful to re-examine closely the advantages claimed for such an arrangement and to study the practical problems and fears that have made such experiments rare.
The theoretical support among economists for a purchasing power guarantee for deferred payments is most impressive. Marshall in 18862 strongly advocated the use of a standard unit of purchasing power in contracts for deferred payments, and reaffirmed this view in 1911.3 Jevons4 even earlier thought so much of the proposal that he suggested that it be made compulsory, after a short transition period, for “every money debt of, say, more than three months’ standing [to] be varied according to the tabular standard”5 of purchasing power. Keynes6 advocated the issue by the U.K. Treasury of “bonds of which the capital and interest would be paid not in a fixed amount of sterling but in such amount of sterling as has a fixed commodity value as indicated by an index number.” More recently, the proposal that the U.S. Government should issue a bond whose value would be guaranteed in terms of purchasing power was enthusiastically supported by such eminent economists as Milton Friedman,7 Fritz Machlup,8 and Sumner Slichter.9 Economists in some non-English-speaking countries have also supported the concept. In France, Courtin,10 Gael Fain,11 and Sauvy12 have proposed purchasing power clauses of various forms to meet current French problems. In Latin America, Javier Márquez has urged the widespread use of guaranteed purchasing power bonds to secure additional resources for development.13 In China, S.C. Tsiang14 supported the use of an indexed savings certificate to overcome some of the disadvantages of the inflation.
In financial circles, on the other hand, the weight of opinion has been overwhelmingly against this “artificial” device. Jevons’ arguments were rejected by Bagehot.15 Keynes’ proposal was dismissed by Sir Otto Niemeyer16 of the Bank of England with the assertion that no one would be interested. More recently, the case made in the United States by some economists before the Patman Subcommittee was attacked vigorously by business economists and in the replies of the Treasury17 and the Council of Economic Advisers.18
The weight of financial opinion has been shown as impressively in the lack of practical application—at least until recently. There are still only a few scattered examples of the use of the device in its “pure form,” i.e., using a price index to measure the adjustment needed to assure a contract with stable purchasing power.19 Even before the use of the clause by banks, Finland provided the broadest application. Government securities with a partial purchasing power guarantee based on price indices have been issued there since 1944. During succeeding years, a range of private securities with similar clauses, which have also been used in some private mortgages, has been issued. In Sweden in 1952 a large cooperative, the Kooperativa Forbundet, issued a bond with a partial guarantee. In the United States there was an early official application of the index principle in Massachusetts in the “equity bills” of 1742 and the “Massachusetts depreciation notes” of 1780, both based on the aggregate price of a few items.20 An interesting modern application using a full index number was made in 1925 when the Rand Kardex Company, at the instance of Professor Irving Fisher, issued a 30-year bond with interest and principal linked to the wholesale price index. The U.S. Government in 1946 accepted a contract in which payments were determined by the movements of a cost of living index.21 In Mainland China, time deposits with a guarantee related to an index based on the prices of a few major commodities were allowed during 1950–52. Official proposals for security issues with an index guarantee have been made in Chile and Israel.
The device in a form less precise than that which involves the use of a price index has been much more widespread. In many Latin American countries, deposits in domestic banks may be denominated in U.S. dollars, a device intended to have the effect of stabilizing their purchasing power. In France the Pinay Loan of 1952, providing repayment guaranteed at a price related to the free market rate of the napoleon, was another variant. This loan was followed in France by issues of semi-governmental corporation bonds which had their purchasing power guaranteed by reference to the average price received for the corporation’s product. Thus the Coal Corporation issued a bond repayable at a price related to the wholesale price of coal; railway bonds were related to the price of a third-class railroad ticket for one kilometer; and those of the Electricity Board were tied to the price of a kilowatt hour. In Austria there was a similar issue by the electricity authority.
In many countries and over periods stretching far back into history, there have been many arrangements tying payments to specific prices. Jevons22 recalled the way in which the colleges of Oxford and Cambridge were obliged by law to lease their lands for corn units, which had the effect over the long period of maintaining the purchasing power of their endowments. Marshall23 referred to the payment of tithes being related to average prices of wheat, barley, and oats. In the German Republic of 1919–23, the price of rye was widely used as a deflator for contracts of all types. Gold clauses have been common in many countries, and they show a remarkable persistence in the face of many legal obstacles. In Greece, particularly, there is widespread use of the gold sovereign as a unit of account in spite of a law making the gold clause unenforceable.
Price indices have been widely used in the related sphere of wage contracts. Both in private arrangements made between a trade union and individual employers and in public arbitration, the insertion of a cost of living clause has become quite normal. Even when no such clause is formally included, the relationship between wages and cost of living is normally in the forefront of any negotiation to such an extent that, in times of rising prices, it almost invariably is the basis on which negotiations begin.
In the following discussion, attention will be centered on the use of purchasing power guarantees in a rather narrow sense. There will be no discussion of their application to wage contracts or pension arrangements. Attention will be focused on the denominating of debts, either private or governmental, in a form which provides that final payment be varied in accordance with an index. Only indirect reference will be made to guarantees under which the return is related to gold, dollars, or the price of the debtor’s product.24
I. Details of Some Recent Applications
II. Selection of the Price Index
If commodity price movements were uniform, there would be no problem in the selection of an appropriate price index to be used as a basis for a purchasing power guarantee, for all indices would give the same result. But, of course, price movements are far from uniform, and there is a bewildering array of possible prices and price indices that could be used and would give varying results. In selecting from these possibilities, many considerations should be borne in mind. Although some of these considerations will be particular to the case in question, many of the more important factors can be analyzed in advance.
Especially for countries where inflation has already been severe, and which are therefore the most likely to introduce a purchasing power guarantee, there is, however, one point of primary importance. In such countries the use of any price index would create an obligation that was more stable in real terms than obligations denominated in terms of money. Any differences between the various measures of price movements will therefore be much less significant than the difference between any one of them and a money standard to which no purchasing power guarantee has been attached.
The simplest device that has been suggested consists of denominating the loan in terms of the selling price of the borrower. This is the procedure followed by the French public utilities. The advantage of this price is, of course, a reduction of the risk for the borrower that his debts will grow in money terms without a commensurate increase in his ability to service them. On the other hand, there is the disadvantage to the lender of a variable purchasing power of the loan in terms of his expenditures. If the borrower’s selling price rises less than the general price level, the lender bears the loss. Abstractly, it would seem that the risk involved in sectional price movements might be more properly borne by the borrower simply because he should be more expert in determining the future comparative prosperity of his industry. Moreover, some of the sectional price movements will be related to differential rates of technological advance, in which case price declines may indicate improved, rather than lessened, ability to service the debt. The significance of these arguments would seem to be sharply reduced if the sector price were defined on a broad basis; for example, a sector as wide as agriculture.
In the particular case of public utilities, there are further considerations. The movements of public utility charges are notoriously a poor indicator of the degree of inflation. In almost every country they move much less and much more slowly than the general price level. Consequently, a purchasing power clause expressed in terms of public utility charges would still leave the lender with a reduced real return in the event of inflation. Moreover, when the utility’s debts are tied to its own selling price, it seems probable that there would be an even greater reluctance to vary the selling price. This would not only reduce the inflation hedge to the lender; it would also have the unfortunate effect of underpricing the utility, with the consequent problems of perpetual overdemand or overinvestment supported by subsidies.
The use of individual selling prices introduces into the bond market a bewildering array of standards. Instead of the relatively simple double standard—money for current payments and an index for deferred payments—there is a multiple standard with increasing complexity for the investor and one that would probably inhibit attempts to obtain more liberal standards for trustee investments. If there is an intention to encourage the widespread use of a purchasing power clause, it would seem desirable to standardize its form as much as possible.
Another single price that has been widely used is that of gold or of foreign exchange. Gold (or dollars) has the advantage of being precisely defined and a sensitive index of inflation. Furthermore, the historical use of gold in payments has made its use for deferred payments more easily comprehended than the use of abstract indices. On the other hand, this history has made both courts and governments sensitive to the implicit challenge to the sovereign right of a country to determine the legal tender. Consequently, such clauses are illegal in many countries, and it has proved easy to pass retroactive legislation making any present legality of doubtful permanence.34 Also, the precision of the price of gold has suffered in recent years. Governments have so many regulations affecting the prices—both official and free—of gold and foreign exchange that any price chosen for a purchasing power clause is subject to the danger of arbitrary variation, arbitrary exclusion from variation, or even disappearance.
Moreover, the use of gold or foreign exchange as a standard of reference has further disadvantages. Whereas the selling price of public utilities is likely to underestimate the degree of price inflation, the price of gold or foreign exchange is likely to overestimate it. Through speculation, the depreciation of the local currency in terms of gold is likely to be much more rapid than in terms of local goods. Although this unrepresentativeness might be somewhat reduced if most deferred payments were on a purchasing power basis, thereby eliminating some of the desire to speculate, it is certain in any significant inflation to cause borrowers to lose unduly as they overcompensate lenders. It therefore partly fails in its purpose to remove risk in loan contracts.35
For government borrowing, the use of a gold clause in favorable circumstances is rather advantageous. In particular, if the government is attempting to raise a loan in a stabilization program after a period of inflation, the free gold price is likely to be particularly favorable to it. The gold price will have overstressed the inflation and is likely to fall significantly if the stabilization is successful. In such circumstances, the government has much to gain and little to lose from a gold clause.
Particular prices—whether of gold or of the goods sold by the borrower—are necessarily inferior for the basic purpose of a purchasing power clause. The arguments for the clause are based directly on the need to avoid the distortion arising from movements of the general price level. This purpose can be achieved only if a direct attempt is made to obtain a representative price index in which the special factors affecting individual prices are submerged in an average that measures only the general pervading influence of inflation or deflation.
But, although the object is to measure general prices, there is one basic question on which a decision is required. Which is more appropriate, the measurement of prices paid to producers of goods and services or the measurement of prices paid by consumers? If, as is ideally desirable, the measurements covered all producers’ prices and all consumers’ prices, the two would be identical in a closed economy. But, in an open economy, the two may differ significantly, for producers’ prices include the prices of exports, while consumers’ prices exclude the prices of exports but include the prices of imports.
The two price indices differ only when the terms of trade vary. If there is an adverse movement in the terms of trade, the use of a producers’ price index will cause the lender to lose in terms of consumers’ prices while the use of consumers’ prices would cause a loss for the producers. The choice between them thus becomes a question of which side should bear the uncertainty of loss (or of gain if the terms of trade improve). Since it seems that the uncertainty would weigh more heavily with the borrower, his debt payments being more significant to his net income than the interest receipts would be to most lenders, the decision would seem to be in favor of using producers’ prices. This, of course, is not directly relevant when the government is borrowing, insofar as the intention is to maximize the savings offered to it. However, even a government may wish to provide for a lightening of the real burden of its debt when there is a severe loss in terms of trade.
Theoretically, it would seem that the index used should be as inclusive as possible, covering as many prices as is feasible. In practice, of course, the index will have to be compiled from a limited sample of data, weighted to represent classes of goods. The extent of the coverage might depend primarily on the number of items for which standardized prices are available and the size of the statistical staff available to compile the index. There is, however, a possibility that in some countries the sample might be deliberately kept small to increase the likelihood that people would understand it. An index based on prices of five or ten major commodities might be sufficiently simple to be understood widely, and at the same time accuracy of measurement would not be sacrificed.
Mr. Finch, economist in the Finance Division, is a graduate of the University of Melbourne, Australia, and the London School of Economics. He was formerly a lecturer in the University of Tasmania.
In his Reply to the Royal Commission on the Depression of Trade and Industry, 1886, reprinted in Official Papers by Alfred Marshall (London, 1926), pp. 9–12. The proposal is also set forth in an article in the Contemporary Review, March 1887, reproduced in Memorials of Alfred Marshall (London, 1925), pp. 188–211.
In a letter to Irving Fisher, reproduced in Memorials of Alfred Marshall (London, 1925), p. 476.
W. Stanley Jevons, Money and the Mechanism of Exchange (New York, 1898), Chap. XXV, “A Tabular Standard of Value,” pp. 318–26. (First published in 1875; page references are to the New York edition published in 1898.)
Ibid., p. 324.
In evidence before the Committee on National Debt and Taxation (Colwyn Committee), Minutes of Evidence (London, 1927), Vol. I, pp. 278 and 287.
Joint Committee on the Economic Report, Monetary Policy and the Management of the Public Debt (82nd Congress, 2nd Session, Washington, 1952), Replies to Questions … for the Use of the Subcommittee … (Wright Patman, Chairman), p. 1105.
Ibid, p. 1106.
In “We Can Win the Economic ‘Cold War,’ Too,” New York Times Magazine (New York), August 13, 1950, pp. 7, 22–26, and in later letters and addresses. His proposal began the discussion leading to the examination by the Patman Subcommittee. A thorough theoretical analysis of the proposal with a favorable conclusion is given by Richard Goode in “A Constant-Purchasing-Power Savings Bond,” National Tax Journal (Lancaster, Pa.), December 1951, pp. 332–40.
According to Le Monde (Paris), May 23, 1954, he proposed the use of the cost of an hour’s unskilled labor as an index for loans.
La Lutte contre inflation et la stabilisation monétaire (Paris, 1947). A price index deflation is suggested.
L’Observation Economique (Paris), October 1949, p. 7. The issue of bonds by public utilities denominated in terms of the selling price of their product is advocated.
Javier Márquez, “Bonos de Poder Adquisitivo Constante,” El Trimestre Económico (Mexico, D.F.), January-March 1954, pp. 6–43.
Economic Review (Shanghai), April 6 and May 10, 1947.
Walter Bagehot, “A New Standard of Value,” The Economist (London), November 20, 1875. Reprinted in The Economic Journal (London), Vol. II (1892), pp. 472–77.
Committee on National Debt and Taxation, op. cit., p. 633.
Joint Committee on the Economic Report, op. cit., pp. 142–45.
Ibid., pp. 888–89.
W. C. Fisher, “The Tabular Standard in Massachusetts History,” Quarterly Journal of Economics (Cambridge, Mass.), Vol. 27 (1912–13), pp. 417–61.
The U.S. -Cuba Sugar Agreement of 1946 provided that the price paid by the Commodity Credit Corporation for Cuban sugar should be linked with the monthly indices of the U.S. Bureau of Labor Statistics. The text is quoted in Arthur Nussbaum, Money in the Law, National and International (Brooklyn, 2nd ed., 1950), p. 305, footnote 32.
W. Stanley Jevons, op. cit., p. 319.
Memorials of Alfred Marshall (London, 1925), p. 197.
In Appendix II, the form of the guarantee is discussed in detail and the merits of various indices are analyzed.
W. Stanley Jevons, op. cit., p. 324.
Memorials of Alfred Marshall (London, 1925), p. 193.
There would possibly be substituted another prestige problem, i.e., the implication that, in resorting to loans with a purchasing power guarantee, the government’s credit is weak. However, this would not seem a valid implication in view of the use of such clauses by the Finnish Government and French governmental institutions, and of the proposals for its use by the U.S. and U.K. Governments.
The power would be reduced but not eliminated because the repayment of a purchasing power bond must be linked to a past price index. Repayment of a debt adjusted in terms of a price index calculated for some earlier date would still permit some reduction in the real value of the settlement.
Robert Giffen made this point explicitly in his article, “Fancy Monetary Standards,” The Economic Journal (London), Vol. II (1892), pp. 463–71.
E.g., Walter Bagehot, op. cit.
Arthur Nussbaum, op. cit., p. 306.
The law in many countries is set forth clearly by F. A. Mann in The Legal Aspect of Money (London, 2nd ed., 1953), Chap. IV, “Methods of Negativing the Effects of Nominalism,” pp. 103–34. The strongest action against such clauses appears to be the decision of the Allied Control Commission in Germany in 1947 which declared all such attempts invalid. In Revue Economique (Paris), March 1955, which deals with the experience with index clauses in France, Joseph Hamel expresses somewhat greater doubts about their validity in France than that shown by Mann.
Proposed for the United Kingdom by the author of a series of articles entitled “Agenda for the Age of Inflation,” The Economist (London), August 25, 1951, pp. 435–37.
Arthur Nussbaum, op. cit., p. 280, lists over 30 countries in which legislation affecting the validity of gold clauses has been passed.
In some circumstances it fails completely. For example, in the United States the devaluation of 1933–34 caused a 70 per cent rise in the price of gold while other prices were approximately stable. This caused the burden of loans with a gold clause to increase sharply—extremely sharply when account is taken of the fall in commodity prices after many of them were contracted. Consequently, there was strong agitation by debtors against the enforcement of the gold clause, which led to the passing of abrogatory legislation.