The Theory of Money and Credit - Ludwig von Mises [82]
Thus if the direction and extent of variations in the exchange value of money could be foreseen, they would not be able to affect the relations between debtor and creditor; the coming alterations in purchasing power could be sufficiently allowed for in the original terms of the credit transaction. [5] But since this assumption, even so far as fluctuations in credit money or fiat money relatively to gold money are concerned, never holds good except in a most imperfect manner, the allowance made in debt contracts for future variations in the value of money is necessarily inadequate; while even nowadays, after the big and rapid fluctuations in the value of gold that have occurred since the outbreak of the world war, the great majority of those concerned in economic life (one might, in fact, say all of them, apart from the few who are acquainted with theoretical economics) are completely ignorant of the fact that the value of gold is variable. The value of gold currencies is still regarded as stable.
Those economists who have recognized that the value of even the best money is variable have recommended that in settling the terms of credit transactions, that is to say, the terms on which present goods are exchanged for future goods, the medium of exchange should not be one good alone, as is usual nowadays, but a "bundle" of goods; it is possible in theory if not in practice to include all economic goods in such a "bundle". If this proposal were adopted, money would still be used as a medium for the exchange of present goods; but in credit transactions the outstanding obligation would be discharged, not by payment of the nominal sum of money specified in the contract, but by payment of a sum of money with the purchasing power that the original sum had at the time when the contract was made. Thus, if the objective exchange value of money rises during the period of the contract, a correspondingly smaller sum of money will be payable; if it falls, a correspondingly larger sum.
The arguments devoted above to the problem of measuring variations in the value of money show the fundamental inadequacy of these recommendations. If the prices of the various economic goods are given equal weight in the determination of the parity coefficients without consideration of their relative quantities, then the evils for which a remedy is sought may merely be aggravated. If variations in the prices of such commodities as wheat, rye, cotton, coal, and iron are given the same significance as variations in the prices of such commodities as pepper, opium, diamonds, or nickel, then the establishment of the tabular standard would have the effect of making the content of long-term contracts even more uncertain than at present. If what is called a weighted average is used, in which individual commodities have an effect proportioned to their significance,[6] then the same consequences will still follow as soon as the conditions of production and consumption alter. For the subjective values attached by human beings to different economic goods are just as liable to constant fluctuation as are the conditions of production; but it is impossible to take account of this fact in determining the parity