Lecture I NATIONAL MONETARY SYSTEMS 1

When I was honoured with the invitation to deliver at the Institut five lectures “on some subject of distinctly international interest”, I could have little doubt what that subject should be. In a field in which I am particularly interested I had been watching for years with increasing apprehension the steady growth of a doctrine which, if it becomes dominant, is likely to deal a fatal blow to the hopes of a revival of international economic relations. This doctrine, which in the title of these lectures I have described as Monetary Nationalism, is held by some of the most brilliant and influential economists of our time. It has been practised in recent years to an ever increasing extent, and in my opinion it is largely responsible for the particular intensification of the last depression which was brought about by the successive breakdown of the different currency systems. It will almost certainly continue to gain influence for some time to come, and it will probably indefinitely postpone the restoration of a truly international currency system. Even if it does not prevent the restoration of an international gold standard, it will almost inevitably bring about its renewed breakdown soon after it has been re-established.

When I say this I do not mean to suggest that a restoration of the gold standard of the type we have known is necessarily desirable, nor that much of the criticism directed against it may not be justified. My complaint is rather that most of this criticism is not concerned with the true reasons why the gold standard, in the form in which we knew it, did not fulfill the functions for which it was designed; and further that the only alternatives which are seriously considered and discussed, completely abandon what seems to me the essentially sound principle—that of an international currency system—which that standard is supposed to embody.

But let me say at once that when I describe the doctrines I am going to criticize as Monetary Nationalism I do not mean to suggest that those who hold them are actuated by any sort of narrow nationalism. The very name of their leading exponent, Mr. J. M. Keynes, testifies that this is not the case. It is not the motives which inspire those who advocate such plans, but the consequences which I believe would follow from their realization, which I have in mind when I use this term. I have no doubt that the advocates of these doctrines sincerely believe that the system of independent national currencies will reduce rather than increase the causes of international economic friction; and that not merely one country but all will in the long run be better off if there is established that freedom in national monetary policies which is incompatible with a single international monetary system.

The difference then is not one about the ultimate ends to be achieved. Indeed, if it were, it would be useless to try to solve it by rational discussion. The fact is rather that there are genuine differences of opinion among economists about the consequences of the different types of monetary arrangements we shall have to consider, differences which prove that there must be inherent in the problem serious intellectual difficulties which have not yet been fully overcome. This means that any discussion of the issues involved will have to grapple with considerable technical difficulties, and that it will have to grapple with wide problems of general theory if it is to contribute anything to their solution. My aim throughout will be to throw some light on a very practical and topical problem. But I am afraid my way will have to lead for a considerable distance through the arid regions of abstract theory.

There is indeed another way in which I might have dealt with my subject. And when I realized how much purely theoretical argument the other involved I was strongly tempted to take it. It would have been to avoid any discussion of the underlying ideas and simply to take one of the many concrete proposals for independent national currency systems now prevalent and to consider its various probable effects. I have no doubt that in this form I could give my lectures a much more realistic appearance and could prove to the satisfaction of all who have already an unfavourable opinion of Monetary Nationalism that its effects are pernicious. But I am afraid I would have had little chance of convincing anyone who has already been attracted by the other side of the case. He might even admit all the disadvantages of the proposal which I could enumerate, and yet believe that its advantages outweigh the defects. Unless I can show that these supposed advantages are largely illusory, I shall not have got very far. But this involves an examination of the argument of the other side. So I have come rather reluctantly to the conclusion that I cannot shirk the much more laborious task of trying to go to the root of the theoretical differences.

2

But it is time for me to define more exactly what I mean by Monetary Nationalism and its opposite, an International Monetary System. By Monetary Nationalism I mean the doctrine that a country’s share in the world’s supply of money should not be left to be determined by the same principles and the same mechanism as those which determine the relative amounts of money in its different regions or localities. A truly International Monetary System would be one where the whole world possessed a homogeneous currency such as obtains within separate countries and where its flow between regions was left to be determined by the results of the action of all individuals. I shall have to define later what exactly I mean by a homogeneous currency. But I should like to make it clear at the outset that I do not believe that the gold standard as we knew it conformed to that ideal and that I regard this as its main defect.

Now from this conception of Monetary Nationalism there at once arises a question. The monetary relations between small adjoining areas are alleged to differ from those between larger regions or countries; and this difference is supposed to justify or demand different monetary arrangements. We are at once led to ask what is the nature of this alleged difference? This question is somewhat connected but not identical with the question what constitutes a national monetary system, in what sense we can speak of different monetary systems. But, as we shall see, it is very necessary to keep these questions apart. For if we do not we shall be confused between differences which are inherent in the underlying situation and which may make different monetary arrangements desirable, and differences which are the consequence of the particular monetary arrangements which are actually in existence.

For reasons which I shall presently explain this distinction has not always been observed. This has led to much argument at cross purposes, and it is therefore necessary to be rather pedantic about it.

3

I shall begin by considering a situation where there is as little difference as is conceivable between the money of different countries, a case indeed where there is so little difference that it becomes doubtful whether we can speak of different “systems”. I shall assume two countries only and I shall assume that in each of the two countries of which our world is assumed to consist, there is only one sort of widely used medium of exchange, namely coins consisting of the same metal. It is irrelevant for our purpose whether the denomination of these coins in the two countries is the same, so long as we assume, as we shall, that the two sorts of coins are freely and without cost interchangeable at the mints. It is clear that the mere difference in denomination, although it may mean an inconvenience, does not constitute a relevant difference in the currency systems of the two countries.[1]

In starting from this case we follow a long established precedent. A great part of the argument of the classical writers on money proceeded on this assumption of a “purely metallic currency”. I wholly agree with these writers that for certain purposes it is a very useful assumption to make. I shall however not follow them in their practice of assuming that the conclusions arrived from these assumptions can be applied immediately to the monetary systems actually in existence. This belief was due to their conviction that the existing mixed currency systems not only could and should be made to behave in every respect in the same way as a purely metallic currency, but that—at any rate in England since the Bank Act of 1844—the total quantity of money was actually made to behave in this way. I shall argue later that this erroneous belief is responsible for much confusion about the mechanism of the gold standard as it existed; that it has prevented us from achieving a satisfactory theory of the working of the modern mixed system, since the explanation of the rôle of the banking system was only imperfectly grafted upon, and never really integrated with, the theory of the purely metallic currency; and that in consequence the gold standard or the existence of an international system was blamed for much which in fact was really due to the mixed character of the system and not to its “internationalism” at all.

For my present purpose, however, namely to find whether and in what sense the monetary mechanism of one country can or must be regarded as a unit or a separate system, even when there is a minimum of difference between the kind of money used there and elsewhere, the case of the “purely metallic currency” serves extraordinarily well. If there are differences in the working of the national monetary systems which are not merely an effect of the differences in the monetary arrangements of different countries, but which make it desirable that there should be separate arrangements for different regions, they must manifest themselves even in this simplest case.

It is clear that in this case the argument for a national monetary system cannot rest on any peculiarities of the national money. It must rest, and indeed it does rest, on the assumption that there is a particularly close connection between the prices—and particularly the wages—within the country which causes them to move to a considerable degree up and down together compared with prices outside the country. This is frequently regarded as sufficient reason why, in order to avoid the necessity that the “country as a whole” should have to raise or lower its prices, the quantity of money in the country should be so adjusted as to keep the “general price level” within the country stable. I do not want to consider this argument yet. I shall later argue that it rests largely on an illusion, based on the accident that the statistical measures of prices movements are usually constructed for countries as such; and that in so far as there are genuine difficulties connected with general downward adjustments of many prices, and particularly wages, the proposed remedy would be worse than the disease. But I think I ought to say here and now that I regard it as the only argument on which the case for monetary nationalism can be rationally based. All the other arguments have really nothing to do with the existence of an international monetary system as such, but apply only to the particular sorts of international systems with which we are familiar. But since these arguments are so inextricably mixed up in current discussion with those of a more fundamental character it becomes necessary, before we can consider the main arguments on its merits, to consider them first.

4

The homogeneous international monetary system which we have just considered was characterised by the fact that each unit of the circulating medium of each country could equally be used for payments in the other country and for this purpose could be bodily transferred into that other country and be bodily transformed into the currency of that country. Among the systems which need to be considered only an international gold standard with exclusive gold circulation in all countries would conform to this picture. This has never existed in its pure form and the type of gold standard which existed until fairly recently was even further removed from this picture than was generally realized. It was never fully appreciated how much the operation of the system which actually existed diverged from the ideal pure gold standard. For the points of divergence were so familiar that they were usually taken for granted. It was the design of the Bank Act of 1844 to make the mixed system of gold and other money behave in such a way that the quantity of money would change exactly as if only gold were in circulation; and for a long time argument proceeded as if this intention had actually been realized. And even when it was gradually realized that deposits subject to cheque were no less money than bank notes, and that since they were left out of the regulation, the purpose of the Act had really been defeated, only a few modifications of the argument were thought necessary. Indeed in general this argument is still presented as it was originally constructed, on the assumption of a purely metallic currency.

In fact however with the coming of modern banks a complete change had occurred. There was no longer one homogeneous sort of money in each country, the different units of which could be regarded as equivalent for all relevant purposes. There had arisen a hierarchy of different kinds of money within each country, a complex organisation which possessed a definite structure, and which is what we really mean when we speak of the circulating medium of a country as a “system”. It is probably much truer to say that it is the difference between the different kinds of money which are used in any one country, rather than the differences between the moneys used in different countries, which constitutes the real difference between different monetary systems.

We can see this if we examine matters a little more closely. The gradual growth of banking habits, that is the practice of keeping liquid assets in the form of bank balances subject to cheque, meant that increasing numbers of people were satisfied to hold a form of the circulating medium which could be used directly only for payments to people who banked with the same institution. For all payments beyond this circle they relied on the ability of the bank to convert the deposits on demand into another sort of money which was acceptable in wider circles; and for this purpose the banks had to keep a “reserve” of this more widely acceptable or more liquid medium.

But this distinction between bank deposits and “cash” in the narrower sense of the term does not yet exhaust the classification of different sorts of money, possessing different degrees of liquidity, which are actually used in a modern community. Indeed, this development would have made little difference if the banks themselves had not developed in a way which led to their organisation into banking “systems” on national lines. Whether there existed only a system of comparatively small local unit banks, or whether there were numerous systems of branch banks which covered different areas freely overlapping and without respect to national boundaries, there would be no reason why all the monetary transactions within a country should be more closely knit together than those in different countries. For any excess payments outside their circle the customers of any single bank, it is true, would be dependent on the reserve kept for this purpose for them by their bank, and might therefore find that their individual position might be affected by what other members of this circle did. But at most the inhabitants of some small town would in this way become dependent on the same reserves and thereby on one another’s action,[2] never all the inhabitants of a big area or a country.

It was only with the growth of centralized national banking systems that all the inhabitants of a country came in this sense to be dependent on the same amount of more liquid assets held for them collectively as a national reserve. But the concept of centralisation in this connection must not be interpreted too narrowly as referring only to systems crowned by a central bank of the familiar type, nor even as confined to branch banking systems where each district of a country is served by the branches of the same few banks. The forms in which centralisation, in the sense of a system of national reserves which is significant here, may develop, are more varied than this and they are only partly due to deliberate legislative interference. They are partly due to less obvious institutional factors.

For even in the absence of a central bank and of branch banking the fact that a country usually has one financial centre where the stock exchange is located and through which a great proportion of its foreign trade passes or is financed tends to have the effect that the banks in that centre become the holders of a large part of the reserve of all the other banks in the country. The proximity of the stock exchange puts them in a position to invest such reserves profitably in what, at any rate for any single bank, appears to be a highly liquid form. And the greater volume of transactions in foreign exchange in such a centre makes it natural that the banks outside will rely on their town correspondents to provide them with whatever foreign money they may need in the course of their business. It was in this way that long before the creation of the Federal Reserve System in 1913 and in spite of the absence of branch banking there developed in the United States a system of national reserves under which in effect all the banks throughout their territory relied largely on the same ultimate reserves. And a somewhat similar situation existed in Great Britain before the growth of joint stock banking.

But this tendency is considerably strengthened if instead of a system of small unit banks there are a few large joint stock banks with many branches; still more if the whole system is crowned by a single central bank, the holder of the ultimate cash reserve. This system, which to-day is universal, means in effect that additional distinctions of acceptability or liquidity have been artificially created between three main types of money, and that the task of keeping a sufficient part of the total assets in liquid form for different purposes has been divided between different subjects. The ordinary individual will hold only a sort of money which can be used directly only for payments to clients of the same bank; he relies upon the assumption that his bank will hold for all its clients a reserve which can be used for other payments. The commercial banks in turn will only hold reserves of such more liquid or more widely acceptable sort of money as can be used for inter-bank payments within the country. But for the holding of reserves of the kind which can be used for payments abroad, or even those which are required if the public should want to convert a considerable part of its deposits into cash, the banks rely largely on the central bank.

This complex structure, which is often described as the one-reserve system, but which I should prefer to call the system of national reserves, is now taken so much for granted that we have almost forgotten to think about its consequences. Its effects on the mechanism of international flows of money will be one of the main subjects of my next lecture. To-day I only want to stress two aspects which are often overlooked. In the first place I would emphasize that bank deposits could never have assumed their present predominant rôle among the different media of circulation, that the balances held on current account by banks could never have grown to ten times and more of their cash reserves, unless some organ, be it a privileged central bank or be it a number of or all the banks, had been put in a position, to create in case of need a sufficient number of additional bank notes to satisfy any desire on the part of the public to convert a considerable part of their balances into hand-to-hand money. It is in this sense and in this sense only that the existence of a national reserve system involves the question of the regulation of the note issue alone.

The second point is that nearly all the practical problems of banking policy, nearly all the questions with which a central banker is daily concerned, arise out of the co-existence of these different sorts of money within the national monetary system. Theoretical economists frequently argue as if the quantity of money in the country were a perfectly homogeneous magnitude and entirely subject to deliberate control by the central monetary authority. This assumption has been the source of much mutual misunderstanding on both sides. And it has had the effect that the fundamental dilemma of all central banking policy has hardly ever been really faced: the only effective means by which a central bank can control an expansion of the generally used media of circulation is by making it clear in advance that it will not provide the cash (in the narrower sense) which will be required in consequence of such expansion, but at the same time it is recognised as the paramount duty of a central bank to provide that cash once the expansion of bank deposits has actually occurred and the public begins to demand that they should be converted into notes or gold.

I shall be returning to this problem later. But in the next two lectures my main concern will be another set of problems. I shall argue that the existence of national reserve systems is the real source of most of the difficulties which are usually attributed to the existence of an international standard. I shall argue that these difficulties are really due to the fact that the mixed national currencies are not sufficiently international, and that most of the criticism directed against the gold standard qua international standard is misdirected. I shall try to show that the existence of national reserve systems alters the mechanism of the international money flows from what it would be with a homogeneous international currency to a much greater degree than is commonly realized.

5

But before I can proceed to this major task I must shortly consider the third and most efficient cause which may differentiate the circulating media of different countries and constitute separate monetary systems. Up to this point I have only mentioned cases where the ratio between the monetary units used in the different countries was given and constant. In the first case this was secured by the fact that the money circulating in the different countries was assumed to be homogeneous in all essential respects, while in the second and more realistic case it was assumed that, although different kinds of money were used in the different countries, there was yet in operation an effective if somewhat complicated mechanism which made it always possible to convert at a constant rate money of the one country into money of the other. To complete the list there must be added the case where these ratios are variable: that is, where the rate of exchange between the two currencies is subject to fluctuations.

With monetary systems of this kind we have of course to deal with differences between the various sorts of money which are much bigger than any we have yet encountered. The possession of a quantity of money current in one country no longer gives command over a definite quantity of money which can be used in another country. There is no longer a mechanism which secures that an attempt to transfer money from country to country will lead to a decrease in the quantity of money in one country and a corresponding increase in the other. In fact an actual transfer of money from country to country becomes useless because what is money in the one country is not money in the other. We have here to deal with things which possess different degrees of usefulness for different purposes and the quantities of which are fixed independently.

Now I think it should be sufficiently clear that any differences between merely interlocal and international movements of money which only arise as a consequence of the variability of exchange rates cannot themselves be regarded as a justification for the existence of separate monetary systems. That would be to confuse effect and cause—to make the occasion of difference the justification of its perpetuation. But since the adoption of such a system of “flexible parities” is strongly advocated as a remedy for the difficulties which arise out of other differences which we have already considered, it will be expedient if in the following lectures I consider side by side all three types of conditions under which differences between the national monetary systems may arise. We shall be concerned with the way in which in each case redistributions of the relative amounts of money in the different countries are effected. I shall begin with the only case which can truly be described as an international monetary standard, that of a homogeneous international currency. Consideration of this case will help me to show what functions changes in the relative quantities of money in different regions and countries may be conceived to serve; and how such changes are spontaneously brought about. I shall then proceed to the hybrid “mixed” system which until recently was the system generally in vogue and which is meant when, in current discussion, the traditional gold standard is referred to. As I said at the beginning, I shall not deny that this system has serious defects. But while the Monetary Nationalists believe that these defects are due to the fact that it is still an international system and propose to remove them by substituting the third or purely national type of monetary system for it, I shall on the contrary attempt to show that its defects lie in the impediments which it presents to the free international flow of funds. This will then lead me first to an examination of the peculiar theory of inflation and deflation on which Monetary Nationalism is based: then to an investigation of the consequences which we should have to expect if its proposals were acted upon; and finally to a consideration of the methods by which a more truly international system could be achieved.

---

[1] Since these lines were written a newly published book has come to my hand in which almost the whole argument in favour of Monetary Nationalism is based on the assumption that different national currencies are different commodities and that consequently there ought to be variable prices of them in terms of each other. (C. R. Whittlesey, International Monetary Issues, New York, 1937.) No attempt is made to explain why or under what conditions and in what sense the different national moneys ought to be regarded as different commodities, and one can hardly avoid the impression that the author has uncritically accepted the difference of denomination as proof of the existence of a difference in kind. The case illustrates beautifully the prevalent confusion between differences between the currency systems which can be made an argument for national differentiations and those which are a consequence of such differentiations. That it is “only a difference in nomenclature” (as Professor Gregory has well put it) whether we express a given quantity of gold as Pounds, Dollars or Marks, and that this no more constitutes different commodities than the same quantity of cloth becomes a different commodity when it is expressed in meters instead of in yards, ought to be obvious. Whether different national currencies are in any sense different commodities depends on what we make them, and the real problem is whether we should create differentiations between the national currencies by using in each national territory a kind of money which will be generally acceptable only within that territory, or whether the same money should be used in the different national territories.

[2] Compare on this and the following L. Robbins, Economic Planning and International Order, 1937, pp. 274 et seq.