A Tract on Monetary Reform

by John Maynard Keynes

[55] Hawtrey, Monetary Reconstruction, p. 105.

As regards the criteria, other than the actual trend of prices, which should determine the action of the controlling authority, it is beyond the scope of this volume to deal adequately with the diagnosis and analysis of the credit cycle. The more deeply that our researches penetrate into this subject, the more accurately shall we understand the right time and method for controlling credit-expansion by bank-rate or otherwise. But in the meantime we have a considerable and growing body of general experience upon which those in authority can base their judgements. Actual price-movements must of course provide the most important datum; but the state of employment, the volume of production, the effective demand for credit as felt by the banks, the rate of interest on investments of various types, the volume of new issues, the flow of cash into circulation, the statistics of foreign trade and the level of the exchanges must all be taken into account. The main point is that the objective of the authorities, pursued with such means as are at their command, should be the stability of prices.

It would at least be possible to avoid, for example, such action as has been taken lately (in Great Britain) whereby the supply of “cash” has been deflated at a time when real balances were becoming inflated,—action which has materially aggravated the severity of the late depression. We might be able to moderate very greatly the amplitude of the fluctuations if it was understood that the time to deflate the supply of cash is when real balances are falling, i.e. when prices are rising out of proportion to the increase, if any, in the volume of cash, and that the time to inflate the supply of cash is when real balances are rising, and not, as seems to be our present practice, the other way round.

II. How can we best combine this primary object with a maximum stability of the exchanges? Can we get the best of both worlds—stability of prices over long periods and stability of exchanges over short periods? It is the great advantage of the gold standard that it overcomes the excessive sensitiveness of the exchanges to temporary influences, which we analysed in Chapter III. Our object must be to secure this advantage, if we can, without committing ourselves to follow big movements in the value of gold itself.

I believe that we can go a long way in this direction if the Bank of England will take over the duty of regulating the price of gold, just as it already regulates the rate of discount. “Regulate,” but not “peg.” The Bank of England should have a buying and a selling price for gold, just as it did before the war, and this price might remain unchanged for considerable periods, just as bank-rate does. But it would not be fixed or “pegged” once and for all, any more than bank-rate is fixed. The Bank’s rate for gold would be announced every Thursday morning at the same time as its rate for discounting bills, with a difference between its buying and selling rates corresponding to the pre-war margin between £3 : 17 : 10½ per oz. and £3 : 17 : 9 per oz.; except that, in order to obviate too frequent changes in the rate, the difference might be wider than 1½d. per oz.—say, ½ to 1 per cent. A willingness on the part of the Bank both to buy and to sell gold at rates fixed for the time being would keep the dollar-sterling exchange steady within corresponding limits, so that the exchange rate would not move with every breath of wind but only when the Bank had come to a considered judgement that a change was required for the sake of the stability of sterling prices.

If the bank rate and the gold rate in conjunction were leading to an excessive influx or an excessive efflux of gold, the Bank of England would have to decide whether the flow was due to an internal or to an external movement away from stability. To fix our ideas, let us suppose that gold is flowing outwards. If this seemed to be due to a tendency of sterling to depreciate in terms of commodities, the correct remedy would be to raise the bank rate. If, on the other hand, it was due to a tendency of gold to appreciate in terms of commodities, the correct remedy would be to raise the gold rate (i.e. the buying price for gold). If, however, the flow could be explained by seasonal, or other passing influences, then it should be allowed to continue (assuming, of course, that the Bank’s gold reserves were equal to any probable calls on them) unchecked, to be redressed later on by the corresponding reaction.

Two subsidiary suggestions may be made for strengthening the Bank’s control:

(1) The service of the American debt will make it necessary for the British Treasury to buy nearly $500,000 every working day. It is clear that the particular method adopted for purchasing these huge sums will greatly affect the short-period fluctuations of the exchange. I suggest that this duty should be entrusted to the Bank of England to be carried out by them with the express object of minimising those fluctuations in the exchange which are due to the daily and seasonal ebb and flow of the ordinary trade demand. In particular the proper distribution of these purchases through the year might be so arranged as greatly to mitigate the normal seasonal fluctuation discussed in Chapter III. If the trade demand is concentrated in one half of the year the Treasury demand should be concentrated in the other half.

(2) It would effect an improvement in the technique of the system here proposed, without altering its fundamental characteristics, if the Bank of England were to quote a daily price, not only for the purchase and sale of gold for immediate delivery, but also for delivery three months forward. The difference, if any, between the cash and forward quotations might represent either a discount or a premium of the latter on the former, according as the bank desired money rates in London to stand below or above those in New York. The existence of the forward quotation of the Bank of England would afford a firm foundation for a free market in forward exchange, and would facilitate the movement of funds between London and New York for short periods, in much the same way as before the war, whilst at the same time keeping down to a minimum the actual movement of gold bullion backwards and forwards. I need not develop this point further, because it is only an application of the argument of Section III. of Chapter III. which will be most readily intelligible to the reader, if he will refer back to the previous argument.

There remains the question of the regulation of the Note Issue. My proposal here may appear shocking until the reader realises that, apart from its disregarding the conventions, it does not differ in substance from the existing state of affairs. The object of fixing the amount of gold to be held against a note issue is to set up a danger signal which cannot be easily disregarded, when a curtailment of credit and purchasing power is urgently required to maintain the legal tender money at its lawful parity. But this system, whilst far better than no system at all, is primitive in its ideas and is, in fact, a survival of an earlier evolutionary stage in the development of credit and currency. For it has two great disadvantages. In so far as we fix a minimum gold reserve against the note issue, the effect is to immobilise this quantity of gold and thus to reduce the amount actually available for use as a store of value to meet temporary or sudden deficits in the country’s international balance of payments. And in so far as we regard an approach towards the prescribed minimum or a departure upwards from it as a barometer warning us to curtail credit or encouraging us to expand it, we are using a criterion which most people would now agree in considering second-rate for the purpose, because it cannot give the necessary warning soon enough. If gold movements are actually taking place, this means that the disequilibrium has proceeded a very long way; and whilst this criterion may pull us up in time to preserve convertibility on the one hand or to prevent an excessive flood of gold on the other, it will not do so in time to avoid an injurious oscillation of prices. This method belongs indeed to a period when the preservation of convertibility was all that any one thought about (all indeed that there was to think about so long as we were confined to an unregulated gold standard), and before the idea of utilising bank-rate as a means of keeping prices and employment steady had become practical politics.