IV. International versus National Policies 10. A COMMODITY RESERVE CURRENCY

In these two brief extracts Professor Hayek points to the advantages of an international economic system, and emphasises the necessity of an international monetary system for its proper functioning.

The gold standard as we knew it undoubtedly had some grave defects. But there is some danger that the sweeping condemnation of it which is now the fashion may obscure the fact that it also had some important virtues which most of the alternatives lack. A wisely and impartially controlled system of managed currency for the whole world might, indeed, be superior to it in all respects. But this is not a practical proposition for a long while yet. Compared, however, with the various schemes for monetary management on a national scale, the gold standard had three very important advantages: it created in effect an international currency without submitting national monetary policy to the decisions of an international authority; it made monetary policy in a great measure automatic and thereby predictable; and the changes in the supply of basic money which its mechanism secured were on the whole in the right direction.

The importance of these advantages should not be lightly underestimated. The difficulties of a deliberate coordination of national policies are enormous, because our present knowledge gives us unambiguous guidance in only a few situations, and decisions in which nearly always some interests must be sacrificed to others will have to rest on subjective judgements. Uncoordinated national policies, however, directed solely by the immediate interests of the individual countries, may in their aggregate effect on every country well be worse than the most imperfect international standard. Similarly, though the automatic operation of the gold standard is far from perfect, the mere fact that under the gold standard policy is guided by known rules, and that, in consequence, the action of the authorities can be foreseen, may well make the imperfect gold standard less disturbing than a more rational but less comprehensible policy. The general principle that the production of gold is stimulated when its value begins to rise and discouraged when its value falls is right at least in the direction, if not in the way, in which it operates in practice.

An Irrational but Real Prestige

It will be noticed that none of these points claimed in favour of the gold standard is directly connected with any property inherent to gold. Any internationally accepted standard based on a commodity whose value is regulated by its cost of production would possess essentially the same advantages. What in the past made gold the only substance on which in practice an international standard could be based was mainly the irrational, but no less real factor of its prestige—or, if you will, of the ruling superstitious prejudice in favour of gold, which made it universally more acceptable than anything else. So long as this belief prevailed it was possible to maintain an international currency based on gold without much design or deliberate organisation to support it. But if it was prejudice which made the international gold standard possible, the existence of such a prejudice at least made an international money possible at a time when any international system based on explicit agreement and systematic cooperation was out of the question. It is probably true to say that all the rational arguments which can be advanced in favour of the gold standard apply even more strongly to this proposal, which is at the same time free from most of the defects of the former. In judging the feasibility of the plan, it must, however, not be regarded solely as a scheme for currency reform. It must be borne in mind that the accumulation of commodity reserves is certain to remain part of national policy and that political considerations render it unlikely that the markets for raw commodities will in any future for which we can now plan be left entirely to themselves. All plans aiming at the direct control of the prices of particular commodities are, however, open to the most serious objections and certain to cause grave economic and political difficulties. Even apart from monetary consideration, the great need is for a system under which these controls are taken from the separate bodies which can but act in what is essentially an arbitrary and unpredictable manner and to make the controls instead subject to a mechanical and predictable rule. If this can be combined with the reconstruction of an international monetary system which would once more secure to the world stable international currency relations and a greater freedom in the movement of raw commodities, a great step would have been taken in the direction toward a more prosperous and stable world economy.

(‘A Commodity Reserve Currency’, pp. 176–77, 184)

11. KEYNES’S COMMENT ON HAYEK

In a comment on Professor Hayek’s article (No. 10) Keynes argues that an international monetary system is incompatible with ‘nationallydetermined’ wage policies: i.e., the domestic restraint imposed by such an international system would be incompatible with the freedom of the organised trade union movement to determine wage-rates.

There are two complaints which it has been usual to lodge against a rigid gold standard as an instrument to secure stable prices. The first is that it does not provide the appropriate quantity of money. This is the familiar, old-fashioned criticism naturally put forward by adherents of the Quantity Theory. The way to meet it is, obviously, to devise a plan for varying appropriately the quantity of gold or its equivalent—for example, the tabular standard of Marshall sixty years ago, the compensated dollar of Irving Fisher forty years ago, or the commodity standard of Professor Hayek expounded in the article printed above.

The peculiar merit of the Clearing Union as a means of remedying a chronic shortage of international money is that it operates through the velocity, rather than through the volume, of circulation. A volume of money is only required to satisfy hoarding, to provide reserves against contingencies, and to cover inevitable time-lags between buying and spending. If hoarding is discouraged and if reserves against contingencies are provided by facultative overdrafts, a very small amount of actually outstanding credit might be sufficient for clearing between well-organised Central Banks. The CU, if it were fully successful, would deal with the quantity of international money by making any significant quantity unnecessary. The system might be improved, of course, by further increasing the discouragements to hoarding.

Conditions for National Price Stability

On another view, however, each national price-level is primarily determined by the relation of the national wage-level to the national efficiency; or, more generally, by the relation of money-costs to efficiency in terms of the national unit of currency. And if price-levels are determined by money-costs, it follows that whilst an appropriate quantity of money is a necessary condition of stable prices, it is not a sufficient condition. For prices can only be stabilised by first stabilising the relation of money-wages (and other costs) to efficiency.

The second (and more modern) complaint against the gold standard is, therefore, that it attempts to confine the natural tendency of wages to rise beyond the limits set by the volume of money, but can only do so by the weapon of deliberately creating unemployment. This weapon the world, after a good try, has decided to discard. And this complaint may be just as valid against a new standard which aims at providing the quantity of money appropriate to stable prices, as it is against the old gold standard.

In the field of price stabilisation international currency projects have, therefore, as I conceive it, only a limited objective. They do not aim at stable prices as such. For international prices which are stable in terms of unitas or bancor cannot be translated into stable national price-levels except by the old gold standard methods of influencing the level of domestic money-costs. And, failing this, there is not much point in an international price-level providing stability in terms of an international unit which is not reflected in a corresponding stability of the actual price-levels of member countries.

Different National Policies Needed

The primary aim of an international currency scheme should be, therefore, to prevent not only those evils which result from a chronic shortage of international money due to the draining of gold into creditor countries but also those which follow from countries failing to maintain stability of domestic efficiency-costs and moving out of step with one another in their national wage-policies without having at their disposal any means of orderly adjustment. And if orderly adjustment is allowed, that is another way of saying that countries may be allowed by the scheme, which is not the case with the gold standard, to pursue, if they choose, different wage policies and, therefore, different price policies.

Thus the more difficult task of an international currency scheme, which will only be fully solved with the aid of experience, is to deal with the problem of members getting out of step in their domestic wage and credit policies. To meet this it can be provided that countries seriously out of step (whether too fast or too slow) may be asked in the first instance to reconsider their policies. But, if necessary (and it will be necessary, if efficiency wage-rates move at materially different rates), exchange rates will have to be altered so as to reconcile a particular national policy to the average pace. If the initial exchange-rates are fixed correctly, this is likely to be the only important disequilibrium for which a change in exchange rates is the appropriate remedy.

It follows that an international currency scheme can work to perfection within the field of maintaining exchange stability, and yet prices may move substantially. If wages and prices double everywhere alike, international exchange equilibrium is undisturbed. If efficiency wage-rates in a particular country rise ten percent more than the norm, then it is that there is trouble which needs attention.

The fundamental reason for thus limiting the objectives of an international currency scheme is the impossibility, or at any rate, the undesirability, of imposing stable price-levels from without. The error of the gold standard lay in submitting national wage-policies to outside dictation. It is wiser to regard stability (or otherwise) of internal prices as a matter of internal policy and politics. Commodity standards which try to impose this from without will break down just as surely as the rigid gold-standard.

Some countries are likely to be more successful than others in preserving stability of internal prices and efficiency wages—and it is the offsetting of that inequality of success which will provide an international organisation with its worst headaches. A communist country is in a position to be very successful. Some people argue that a capitalist country is doomed to failure because it will be found impossible in conditions of full employment to prevent a progressive increase of wages. According to this view severe slumps and recurrent periods of unemployment have been hitherto the only effective means of holding efficiency wages within a reasonably stable range. Whether this is so remains to be seen. The more conscious we are of this problem, the likelier shall we be to surmount it.

(‘The Objective of International Price Stability’, pp. 185–87)

12. F.D. GRAHAM’S CRITICISM OF KEYNES

Professor F.D. Graham, in a comment on Keynes’s criticism, indicated the full inflationary implications of Keynes’s ideas; in the economic system Keynes envisaged, monetary policy would be subordinated to the wage-rate policies followed by the unions, so that the politically powerful were enabled to exploit the politically weak (since those who were unable to raise their incomes along with the trade unionists would have to face ever-increasing prices on static or more slowly rising incomes). Professor Graham’s essay is notable for its early exposition of this important consideration.

The issues raised in Lord Keynes’s reply to Professor F.A. Hayek’s article on a commodity reserve currency, in a recent issue of Economic Journal, seem worthy of more extended discussion.¹

It will perhaps do no great injustice to Professor Hayek’s views to assert that he brought the great weight of his authority to an all but unqualified support of the proposal to give free coinage to warehouse receipts covering representative bales of the standard storable raw materials of industry and trade.²

Professor Hayek believes that the defects of the gold standard lay not in conception, but in adequacy to its task. The gold standard always operated in the right direction, but not with sufficient power or speed. Whenever the public showed an increasing preference for liquidity—with a consequent fall in the price level—the mining of gold was stimulated in compensation of the unemployment with which other industries were then afflicted. But the relative unimportance of gold mining as an employer of labour, or its complete absence from many economies, reduced this compensation to negligible importance everywhere but in South Africa. The gold standard also operated, at long last, to check any secular trend in the price level through the increase in the rate of gold supply which attended a rise in the real value of gold, and the reduction in the rate of gold supply which occurred when the real value of gold fell off. But, if the nineteenth century may be taken as a criterion, the attendant ‘cycle’ takes something like a quarter of a century to run its course.

Though the gold standard thus tended towards the maintenance of full employment, and to the preservation of a stable price level, the tendency in both cases was so faint as to be of no practical importance. Professor Hayek and other advocates of a commodity reserve standard assert that it would greatly ameliorate, if not completely cure, these defects of adequacy in its gold counterpart.

The ‘Natural Tendency of Wages’

Lord Keynes, I take it, is not concerned to deny these asserted virtues of a commodity reserve standard, but says that it is open, along with the gold standard, to another, more modern and, one gathers, more important, objection, in that it would attempt ‘to confine the natural tendency of wages to rise beyond the limits set by the volume of money’, and that it could do so only by deliberately creating unemployment.

I do not know what Lord Keynes means by the ‘natural’ tendency of wages to rise beyond the limits set by the volume of money, unless it is that the wage-earner would always like to have higher money wages than he can currently earn on the basis of a stable price level, and that no one is in a position to prevent his getting them. This would certainly be news to Karl Marx, and if both Marx and Keynes were right in their day and generation, the proletariat has surely come into its own, and more, in a way that Marx never envisaged. The degree in which it is true that there is any ‘natural’ tendency towards an increase of money wages per unit of output is, of course, a matter of time, place and circumstance, and what should be done about it, in any given case, is a political rather than an economic problem. The problem, that is, does not at all touch the question as to whether the commodity reserve standard is an economically good monetary standard, but is solely concerned with its reception by a politically potent group.

That Lord Keynes is under no illusions about the dangers of appeasement of such a group is shown by the fact that, in his concluding paragraph, he says that ‘some people argue that a capitalist country is doomed to failure because it will be found impossible in conditions of full employment to prevent a progressive increase in wages’ (beyond the point which can be sustained without a persistent rise in the price level). Disregarding the query as to whether a country so situated could be called ‘capitalist’, rather than under the domination of a not very enlightened proletariat, it may, perhaps, be at once conceded that there is much to be said for Lord Keynes’s contention that, in dealing with the problem, it is essential that any given country have sovereignty over its own monetary arrangements. His opposition to an international stabilisation of prices, imposed from outside on all participating countries, applies, of course, to an international commodity reserve standard with fixed exchange rates. (It applies still more strongly to an international system of unstable prices, with fixed or viscous exchange-rate relationships, since the international price level might then fall rather than rise and would, in any case, inevitably fail to correspond with the varying shifts in independently determined efficiency-wage rates in the several countries.) Since Lord Keynes justifiably believes that the difficulty of securing the allegiance of the wage-earning group to a policy of stable national price levels would be greatly enhanced if it could be made to appear that such a policy was the result of an international convention, rather than of purely national interest, he looks askance at what he believes to be a proposal, through international action, to fasten stable price levels on all participating countries.

Gold Standard ‘Dictation’

Lord Keynes, however, is, I think, not right in saying that ‘the error of the gold standard lay in submitting national wage-policies to outside dictation’. The original gold standard did not submit wage-policies to dictation, by governing authority anywhere, but made them the resultant of impersonal forces issuing out of the disposition, and potentiality, of individuals to follow what they conceived to be their own interest. This system, as Professor Hayek points out, had many virtues, and we should be badly advised if we should throw away its virtues along with its imperfections. The automaticity of the gold standard was, per se, all to the good, and what we need is a similarly automatic system which will be free of the vices of the traditional gold standard. We should not forget that the once well-nigh universal adhesion to the gold standard was spontaneous rather than imposed, and that it was only after the gold standard had been subjected to varying national management, in an attempt to overcome the original objections against it, that it was abandoned by those countries that could not make their ideas on its management effective, that is, after (unstable) price levels had been imposed from without.

Lord Keynes’s assertion that a commodity reserve standard imposed from without (such as he supposes Professor Hayek to endorse) would break down just as surely as the rigid gold standard, is not obviously true, but I am not concerned to dispute it.³ Professor Hayek, in his article, fails to state explicitly whether or not he posits fixed exchange rates of all national currencies against the international commodity standard and, therefore, against each other. But if, in line with Professor Hayek’s suggestion, some international organisation, such as, e.g., the new ‘Fund’ or the Bank for International Settlements, should offer freely to exchange, both ways, an international currency unit against warehouse receipts covering a designated composite of raw materials, no monetary policy would thereby be imposed on any country. So far as any country chose to keep the exchange value of its own currency fixed, against the international unit and other currencies tied to it, it would automatically have a substantially stable price level. So far, however, as, for one reason or another, it preferred an unstabilised price level, the exchange value of its currency, vis-à-vis the international unit and the currencies of countries with stable exchange rates against that unit, would, as a result of commodity arbitrage, automatically shift in strictly appropriate correspondence with its shifting domestic purchasing power. It seems to me, therefore, that Lord Keynes’s argument that an international commodity reserve currency would impose, from without, a price-level policy on any country, or would break down, is quite untenable.

There would seem to be no reason why an international monetary unit of this sort should not be the international currency around which the operations of a Clearing Union, on Lord Keynes’s lines, or any international fund, could be centred. Through the concurrent free purchase and sale of gold, at a fixed price in the international currency, the gold value of the international unit could also be fixed, or, what is the same thing, the commodity value of gold would be stabilised. This would avoid all the controversy which would be involved in any proposal to deprive gold of its present, or traditional, functions.

Such a standard would represent a great advance over anything we have had in the past. Not only would it be of great value in connection with international investment, but it would furnish a point d’appui to which any country desirous of stable price levels, and of fixed exchange rates with other like-minded countries, could, by linking its own currency to the international unit through purchase and sale at fixed prices, repair.

Unanchored Medium of Exchange

When once a tie with any and every asset, or group of assets, is abandoned, and resort is had to a pure debt currency, one has, in my judgement, no standard at all, but merely a wholly unanchored medium of exchange and unit of account. Though I have the fullest sympathy with the wage-earner’s getting the highest (real) wages possible, in the current state of the industrial arts, it seems to me that any monetary policy which does not confine such tendency as (money) wages may have to rise beyond the limits within which it is possible to preserve a stable price level, provides a very vicious ‘standard’. If Lord Keynes takes the contrary view, he seems to me, in effect, to be plumping for a progressive inflation, wholly indefinite as to time and amount. Against any argument for such a currency I would assert that movements in the price level have no functional significance, or that, if they have, we cannot hope to run a satisfactory economic system with price as the regulating mechanism. In that case, the more quickly we go to some not very limited form of responsible totalitarianism the better it will be for all concerned. If we cannot have a distributorily neutral money, any group that can get control of the monetary system will have totalitarian power over the lives and fortunes of their fellows, without any clear recognition of responsibility.

In a perfectly free monetary system, there is, of course, no rate of money wages which would ever, of itself, bring unemployment, since there is nothing to prevent commodity prices from rising (under the stimulus of new issues of money) to whatever level is necessary to cover the stepped-up money cost of the labour factor of production. All of us, moreover, are impatient with the senseless unemployment with which we have so long been afflicted. But, if we refuse even to accept the threat of unemployment under any conditions whatever, we shall, under any ‘natural’ tendency of money wages to rise faster than efficiency, be forced to pay whatever money wages labourers may be pleased to demand and to jack up the price level unendingly to take care of the situation. The knowledge of what unlimited inflation can mean would seem to preclude the prevention, in this way, of a mote of unemployment.

A commodity reserve currency would operate to provide unlimited employment, through the unlimited demand for the commodities in the reserve, provided the workers did not seek to drive money wages above the figure which, at a stable price level, is warranted by their real productivity. They are entitled to so much—not less and not more—and, if we shrink from saying ‘No’ when they press their demands beyond this point, we shall no longer have an economic system, but merely a racket. One may contend, if he will, that to say ‘No’ is a deliberate induction of unemployment, but the answer is that employment will be available if the workers refrain from pushing what, in the circumstances, are quite impossible demands for higher (real) wages. Higher money wages, if granted in the circumstances, would do the workers no good as consumers, since such wages must be compensated by a higher price level, and, even if some slight unemployment were thus prevented, it would be at devastating cost in social freedom.

The Real Problem of Unemployment

Our real problem of unemployment is not that people are denied the opportunity of work at whatever fancy wage they may desire, but that they are denied that opportunity at wages that they could readily earn under conditions of normal liquidity preference. It is the merit of commodity reserves that they would operate to keep the preference for liquidity from rising, or would sate the appetite for it, by offering it freely in such a way as not to interfere with production.

It is true, as Lord Keynes says, that ‘prices can only be stabilised by first stabilising the relation of money wages (and other costs) to efficiency’. This is precisely the purpose of commodity reserve money, and I can see no reason for not pursuing it. As the efficiency of labour rises, money wages would tend to rise in correspondence—no more and no less—and there would be a steady tendency towards full employment without a trace of inflation.

It is also true that ‘international prices which are stable in terms of unitas or bancor [the international unit] cannot be translated into stable national price-levels except by . . . influencing the level of domestic money-costs’. Domestic money-costs would be so influenced, under fixed rates of exchange of a national currency against the international commodity unit, and I see no strong reason for objecting to this consequence. Whatever the objections, or lack of them, the influence would, in any case, not be present if, as pointed out above, the exchange rate of the national against the international unit were left free to move in correspondence with variations in the local currency price of the commodity composite relative to the fixed price in the international currency.⁴ If one insists upon an unstabilised price level at home, there is nothing in a stabilised international unit to prevent it, or nothing to prevent other countries having stable price levels if they so desire. No country, therefore, would be any more inhibited in the presence of an international monetary unit of stable value than in the presence of an international unit without anchor, and a stable-value international unit would not interfere in any way with anything that Lord Keynes has proposed in his Clearing Union.

Professor Hayek’s ‘Intransigence’

It is the intransigence of the attitude taken here, and by Professor Hayek, which is, I think, troubling Lord Keynes. To him it seems ruthless to accept, or provoke, unemployment as a means of enforcing adherence to pecuniary purity. How much otherwise avoidable unemployment, he asks, would you be willing to bring about for this purpose?⁵ The query reflects not only Lord Keynes’s humanitarian concern, but also his doubts as to political possibilities. He thinks that other, less punitive, means must be found if the desired end is to be attained. It would be churlish, and foolish, to deny the cogency of his objections on this point, and the answer, I think, lies in the adoption of a minimum wage policy with a normal yearly increase in the minimum equal to a generously computed expectation of enhancement in the general level of efficiency. Experience goes to show that wages above the minimum will respond, at least proportionately, to any increase in the lowest group, and if, at any time, the expectation of improvement in general efficiency were shown to be over-computed (by the fact or immediate threat of unemployment in the industries producing the goods in the commodity unit), the stated increase in the minimum wage should be temporarily suspended in accordance with appropriate provisions in the legislation. Some such measures as this would reduce the acerbity of disputes over the distribution of income and would promote adjustments in an orderly rather than chaotic manner.

So long as our economic system deviates widely, and in certain respects progressively, from ideally free competition, there is bound to be some friction in the determination of who gets what, and why. So long, moreover, as we preserve anything whatever of the spirit of free contract, the enterpriser must be as free to reject the demands of workers as are the latter to reject the terms that the enterpriser may offer. Any unemployment that may result from this cause is an inevitable phase of freedom. It would be as fatal to freedom to insist that, to avoid any unemployment whatever, the enterpriser must pay whatever monetary wages organised workers may demand, and that the State must so shape its monetary policy as to make this possible, as it would be to insist, to the same end, that workers must accept whatever monetary wage a fascist group of employers might see fit to impose.

(‘Keynes vs. Hayek on a Commodity Reserve Currency’, pp. 422–28)

13. KEYNES’S REPLY TO GRAHAM

In his reply to F.D. Graham, Keynes failed to take up the major issue that had been raised: that of allowing monetary policy to be used as a supplement to union wage-rate fixing. Indeed, Keynes continued to assume that prices, as a normal course, would be adjusted to whatever wage-rates unions succeeded in obtaining, i.e., he assumed a continuous inflation.

Professor Graham’s statement of my point of view is a very fair one. But in the note on which he comments I expressed myself much more briefly than the nature of the subject matter really allowed. So, to diminish the chances of misunderstandings, there are one or two points I should like to restate and emphasise.

I have no quarrel with a tabular standard as being intrinsically more sensible than gold. My own sympathies have always fallen that way. I hope the world will come to some version of it some time. But the opinion I was expressing was on the level of contemporary practical policy; and on that level I do not feel that this is the next urgent thing or that other measures should be risked or postponed for the sake of it. These are some of my reasons:

1. The immediate task is to discover some orderly, yet elastic, method of linking national currencies to an international currency, whatever the type of international currency may be. So long as national currencies change their values out of step with one another, I doubt if this task is made easier by substituting a tabular standard for gold. Indeed the task of getting an elastic procedure may be made more difficult, since a tabular standard might make rigidity seem more plausible. Perhaps unjustly, I was suspecting Professor Hayek of seeking a new way to satisfy a propensity towards a rigid system.

2. In particular, I doubt the political wisdom of appearing, more than is inevitable in any orderly system, to impose an external pressure on national standards and therefore on wage levels. Of course, I do not want to see money wages forever soaring upwards to a level to which real wages cannot follow. It is one of the chief tasks ahead of our statesmanship to find a way to prevent this. But we must solve it in our own domestic way, feeling that we are free men, free to be wise or foolish. The suggestion of external pressure will make the difficult psychological and political problem of making good sense prevail still more difficult.

3. This does not strike me as an opportune moment to attack the vested interests of gold holders and gold producers. Why waste one’s breath on what the Governments of the United States, Russia, Western Europe and the British Commonwealth are bound to reject?

4. The right way to approach the tabular standard is to evolve a technique and to accustom men’s minds to the idea through international buffer stocks. When we have thoroughly mastered the technique of these, which is sufficiently difficult without the further complications of the tabular standard and the oppositions and prejudices which this must overcome, it will be time enough to think again. On buffer stocks I can enthusiastically join forces with Professor Frank Graham and Mr. Benjamin Graham. Though even here I am beginning to feel a slight reserve about whether just this moment, when many materials are scarce, is the right moment to start; they can so easily be turned into producers’ ramps, and if they start that way the prospect of a brilliant improvement will have been prejudiced.

All this, I agree, is very low-level talk; for which I apologise. But it was in fact from a low level that I was, in the first instance, addressing Professor Hayek on his dolomite.

(‘Note by Lord Keynes’, pp. 429–30)

---

¹Professor Hayek’s article, ‘A Commodity Reserve Currency’, was followed by Lord Keynes’s reply, ‘The Objective of International Price Stability’, Economic Journal LIII, Nos. 210–11 (June–September, 1943): 176–87.

²The proposal is elaborated by its initiator, Mr. Benjamin Graham, in his book Storage and Stability (New York: McGraw-Hill, 1937), and is now so well known as not to require further exposition here.

³The reason that it is not obviously true is that a policy of stable price levels would, I believe, prove to be much more generally acceptable than the caprices of the unmanaged gold standard or the arbitrariness of that standard in its managed form.

⁴This whole matter is treated in detail in my brochure Fundamentals of International Monetary Policy, International Finance Section, Department of Economics and Social Institutions, Princeton University, No. 2.

⁵The question was raised in private correspondence with the author.