The Monetary Approach to the Balance of Payments
Essential Concepts and Historical Origins
JACOB A. FRENKEL AND HARRY G. JOHNSON
1 ESSENTIAL CONCEPTS
The main characteristic of the monetary approach to the balance of payments can be summarised in the proposition that the balance of payments is essentially a monetary phenomenon. The term ‘the balance of payments’ refers to items that are ‘below the line’ in the over-all balance of payments (which must balance exactly, by the principles of double-entry accounting); the items in question constitute the ‘money account’. In general, the approach emphasises the budget constraint imposed on the country’s international spending and views the various accounts of the balance of payments as the ‘windows’ to the outside world, through which the excesses of domestic flow demands over domestic flow supplies, and of excess domestic flow supplies over domestic flow demands, are cleared. Accordingly, surpluses in the trade account and the capital account respectively represent excess flow supplies of goods and of securities, and a surplus in the money account reflects an excess domestic flow demand for money. Consequently, in analysing the money account, or more familiarly the rate of increase or decrease in the country’s international reserves, the monetary approach focuses on the determinants of the excess domestic flow demand for or supply of money.
Clearly, a consistent use of the budget constraint implies that the money account—the current rate of change of reserves—can be analysed in terms of the determinants of all the other accounts—at the simplest level of aggregation, the goods account and the capital account. The monetary approach, however, recommends an analysis in terms of the behavioural relationship directly relevant to the money account, rather than an analysis in terms of the behavioural relationships directly relevant to the other accounts and only indirectly to the money account via the budget constraint. Since the money account is determined by the excess flow demand for money, it is clear why the balance of payments is regarded as a monetary phenomenon and this approach is referred to as ‘the monetary approach’. To repeat, the monetary approach should in principle give an answer no different from that provided by a correct analysis in terms of the other accounts. The main reason for preferring the monetary approach is that less direct alternative approaches have almost invariably attempted to explain the behaviour of the markets they concern themselves with, by analytical constructs in which the role of money in influencing behaviour, and the connection between these other markets and the money markets, are neglected as being ‘of the second order of smalls’, which may be a legitimate procedure for many economic problems, but cannot be so for an analysis which aims to explain or predict behaviour in the money market.
The surplus or deficit in the goods account (more generally the current account) measures the extent to which the economy’s income is greater than consumption (‘absorption’) and the economy is therefore accumulating claims on future income (assets) from abroad or vice versa. By virtue of the budget constraint, the sum of the deficit on the capital account (net purchase of foreign securities) and the surplus on the money account equally represents the accumulation of foreign assets (decumulation if negative). The so-called ‘absorption approach’ to the balance of payments, associated with Sidney Alexander [1], emphasises the rate of accumulation or decumulation of foreign assets (securities plus money). In so doing, it constitutes an improvement over at least the cruder versions of the ‘elasticity approach’, which emphasises relative-price-induced substitution of domestic demand away from or towards imports and of foreign demand towards or away from exports, on the implicit assumption that such substitutions are matched by equal increases or decreases in absorption. The monetary approach selects for emphasis a subset of the spectrum of foreign assets whose accumulation or decumulation is emphasised by the absorption approach. The main reasons for this are, firstly, that the accumulation of foreign assets does not necessarily imply the accumulation of money through the balance of payments—it may mean the opposite, as for example when a monetary policy of lowering interest rates leads domestic asset-holders to move their funds from domestic to foreign securities. Secondly, the monetary authorities in their role as stabilisers of the exchange rate in a fixed rate system are concerned with what causes the stock of international reserves to change and how to prevent such changes. Thirdly, the monetary authority, as the ultimate source of domestic money, controls the rate of change of the domestic credit component of the monetary base—the other component being international reserves—and thereby the flow supply of money. It should be noted that control of the flow supply of money does not mean control of the stock supply of money, or of ‘the quantity of money’, since in an open economy the public can and does determine the total stock of money through its ability to convert domestic money into goods and securities in the international markets. The assumption that the residents of the country have a demand for money which depends on variables at least in part different from those that determine the quantity of domestic credit extended by the banking system, or alternatively that the rate of change of money demanded (the rate of hoarding) is independent of the rate of change of the domestic credit source component of the monetary base, implies that the money account of the balance of payments is influenced directly by monetary policy.
The accumulation or decumulation of assets depends on the aggregate relationship between domestic expenditure and income and does not depend on the composition of expenditure between exportables and importables, or between goods that, given the price structure, are classifiable into tradeable and non-tradeable goods. Consequently, though relative prices do influence the composition of expenditures, they play a secondary or negligible role in the monetary approach (as in the absorption approach). On the other hand the general price level does play a central role, since it determines the real value of nominal assets—money, and possibly fixed-interest debt traded internationally—and this role is emphasised along with the roles of other macro-economic aggregate magnitudes.
This brief account of the monetary approach to balance of payments theory, and comparison of it with the alternative elasticity and absorption approaches, implicitly disposes of various, and often ill-informed, criticisms that have tended to be made of it, and which amount essentially to red herrings across the trail of scientific study and understanding.
To begin with, the approach is described as ‘monetary’, and not ‘monetarist’, precisely to avoid confusion with recent domestic policy debates in which the term ‘monetarist’ has been used by the debaters to represent alternatively attaching ‘appropriate’ and ‘too much’ importance to money, and specifically to the use of monetary as contrasted with fiscal policy in economic stabilisation. The monetary approach to the balance of payments asserts neither that monetary mismanagement is the only cause, nor that monetary policy change is the only possible cure, for balance of payments problems; it does suggest, however, that monetary processes will bring about a cure of some kind—not necessarily very attractive—unless frustrated by deliberate monetary policy action, and that policies that neglect or aggravate the monetary implications of deficits and surpluses will not be successful in their declared objectives.
Secondly, the essential assumption of the monetary approach, like the restated quantity theory of money according to Friedman, is that there exists an aggregate demand function for money that is a function of a relatively small number of aggregate economic variables. In this respect it makes exactly the same assumptions as the Keynesian theory, except for the extreme versions of the latter theory, not to be found in the writings of Keynes himself, in which the demand for money as a function of interest rates is subject to a ‘liquidity trap’, or the supply of (domestic) money is made completely elastic by monetary policy itself; and these extreme cases cannot be made plausible for a single member of an international monetary system of fixed exchange rates. One need not even make the standard assumption of monetary theory, that the demand for money is homogeneous of degree one in all prices and nominal money assets (absence of money illusion), though this assumption is overwhelmingly supported by the empirical evidence and has the convenient result of ensuring the classical property of ‘the neutrality of money’. The monetary approach to the balance of payments, like the classical quantity theory of money, can be readily applied to conditions of price and wage rigidity and consequent response of quantities—employment, output, consumption—rather than money wages and prices to monetary changes. That the monetary approach largely assumes a fully employed economy is partly the result of the fact that in the context of a growing world economy in the long run the assumption of wage rigidity and variable employment becomes uninteresting; either employment expands into the full employment range and quantity adjustments yield to money price and wage adjustments, or it contracts and people either starve to death and go back to full employment numbers, or there is a revolution on Marxist lines, or more likely the public simply votes for the other political party than the one in power, since all of them promise to maintain full employment and the public expects them to do it. More fundamentally, the assumption of normally full employment reflects the passage of time and the accumulation of experience of reasonably full employment as the historical norm rather than the historical rarity that Keynes’s theory and left-wing Keynesian mythology made it out to be.
The fact that the essential foundation of the monetary approach is the assumption that the demand for money is a stable function of a few macro-economic variables, incidentally, disposes of a criticism that has been brought against it, as it has been against the restated quantity theory of money before it, and also the old quantity theory of money, namely that it is not a theory but merely a tautology. Every theory starts from a tautology of some kind, and has to do so to define the variables it seeks to explain. Thus for example demand theory starts from the tautology that income must be spent on a particular commodity or something else, and Keynesian theory from the tautology that expenditure must be classed as either consumption or investment expenditure. What converts the first tautology into demand theory is the assumption that expenditure choices are rationally dependent on income and prices, the second, that consumption is rationally dependent on aggregate income received. What converts the quantity equation into the quantity theory is the assumption that velocity is some kind of stable function (primitively, an institutionally-determined constant); and what underlies the monetary approach is the same assumption that the demand for money is a stable function. In each case, the tautology–converted–into–theory is useful and interesting only if the arguments of the posited stable function can be treated as independently given prices by the market, factor prices, or technology and tastes, investment by rational entrepreneurial calculation, money supply by policy, or by fractional reserve banking based on an ultimate reserve.
Thirdly, a number of criticisms that appear to be cogent from the standpoint of the received alternative theory that suggests them turn out not to be so, either intrinsically or because they can be and have been easily incorporated in the monetary approach. To be specific, criticism has been directed against ‘the small-country assumption’, that is, the assumption that the country under analysis can be regarded as faced with a parametric set of world prices and interest rates. But there is in principle no theoretical difference, or increased theoretical difficulty, in regarding demand and supply functions dependent on prices, rather than the prices themselves, as parametric. More fundamentally, size becomes relevant to the analysis only if the process under analysis involves changes in relative size; and this problem will arise only if the analysis must compare a surplus-deficit situation in which one side absorbs more and the other less of a total of current production than it would in the case of international equilibrium with the equilibrium situation. Apart from the question of the meaning and relevance of such a comparison, the difference between the two situations will depend, as transfer theory has long understood, not on the relative sizes of the national units but on the marginal effects on demand for goods and assets of a shift in expenditure from one nation’s control to the other’s—and there is no presumption of correlation between national sizes on the one hand and direction of expenditure-shift and effects on the terms of trade (or, by extension, relative interest rate levels) between the two countries on the other hand. Where the small-country assumption does become relevant is on the monetary side of the analysis; concretely, a large country—the United States, and to a lesser extent other international financial centres—may be able to operate its domestic policies on the assumption that its national money is internationally acceptable so that, say, an expansion of its domestic credit through a ‘cheap money’ policy will lead to an accumulation of its money in the hands of foreign holders—and eventually to world inflation—rather than to a loss of international reserves.
By the same logic, the changes in the terms of trade—relative prices of import and export goods—that are the centre-piece of the elasticity approach, are by the monetary approach either a secondary consequence of the movement from payments disequilibrium (more accurately, balance-of-payments imbalance, since such an imbalance must always in some sense represent a temporarily chosen equilibrium position for the nation whose aggregate behaviour it represents) to equilibrium (balance) in the balance of payments—which, it should be repeated, may go either way, according to standard transfer theory—or else a transient feature of adjustment from an equilibrium prevailing before a disturbance to the same equilibrium after the disturbance.
The standard elasticity model, like standard ‘real’ trade theory, assumes that all goods are traded, or in other words can be divided into export and import goods. Criticism of this assumption in the pure theory of international trade, stressing the presence and role in ‘real’ adjustment to ‘real’ disturbance of non-traded goods and the fact that in their presence the relative price of imports in terms of exports cannot be simply related to the nature of the disturbance, has been accompanied by the development of elasticity models of monetary and exchange rate adjustment aggregating goods into traded (including both exportable and importable) goods, and non-traded goods, and focusing upon relative price-change—induced substitutions between tradeables and non-tradeables. This alternative to the standard ‘elasticity’ approach, with its presumption of a predictable direction of terms-of-trade changes in adjustment to a specified disturbance, instead emphasises predictable effects on the relative prices of non-traded as compared with traded goods, and has naturally led to criticism of the commonly used formulation of the monetary approach on the basis of the small-country assumption, taken to imply that all goods have common world-market prices. To this there are two conclusive counter-arguments. Firstly, even if goods cannot be traded, the factors used in producing them generally can be, in the sense that in the relevant run of time a barber has the alternative of being a machine tool operator producing machinery or consumers’ durables for export, or instead of imports, and the price of haircuts must be such as to give the barber labour earnings comparable to the wages paid in exporting (and import competing) industries. Secondly, even where there are no comprehensive direct links between costs of production of tradeable and non-tradeable goods, the obvious case being that in which the non-tradeable good employs an internationally immobile specific factor such as climate or land of a specific quality—note that cases are harder to construct than might seem true at first sight, since though coal mines are specific, coal can move, though at a cost—the prices of non-traded goods will be linked to the prices of tradeables through tastes, supply conditions, and the over-all budget restraint, and fixed given the other factors and the relation between domestic expenditure and income. In other words, the influence of non-traded goods must either depend on a situation involving comparison of an imbalanced with a balanced balance of payments, or be merely a transient feature of the adjustment process. On this basis, the abstraction of the main core of the analysis—though not of all analyses based on this approach—from the existence of non-traded goods is a secondary matter. The existence of non-traded goods does, however, become relevant in the empirical application of the theory, in both the static case when price indexes may differ from the prediction of simple purchasing power parity theory owing to differences in the money prices and expenditure weights of such goods between countries, and the dynamic case of growth of productivity at different rates in the traded and non-traded goods sectors, which implies different price trends in the two sectors when factor mobility equalises factor prices between them.
A somewhat different species of red herring is the objection that the monetary approach puts all the emphasis on adjustment of the trade or current account and not on the capital account, which the critics—especially the ‘practical-minded’ central bankers—regard as the more important arena for international adjustments. This criticism is far more valid as applied to the pre-Keynesian version of the monetary approach, which developed well before the emergence into visibility of the institutional machinery of international capital movements, and the Keynesian approach, with its special concern with the effects of international adjustment on domestic employment, than as directed against the modern monetary approach (classical theorists such as Mill were, however, concerned with capital movements and adjustment in the capital account). For concentration on the balance of payments as mirroring the excess demand or supply of money leaves open the question of which of the other two markets bears the main burden of adjustment, whereas an approach centring on either of the other two markets inevitably tends to prejudge the issue.
A final red herring is the contention that the monetary approach is inapplicable to the current regime of generally floating exchange rates. This is a red herring in the empirical sense that the numerical majority of countries still maintains parities explicitly or in practice with one of a few major currencies, and in the historical sense that much of the development of the monetary approach has been occasioned by concern with the analysis of currency devaluation in a theoretical world of low trade barriers and high international mobility of capital. It is a redder herring, that is, a misunderstanding in a more fundamental sense, in terms of basic theory. In a fixed exchange rate world, a country controls neither its price level nor its quantity of domestic money in anything but the short term; its money supply is endogenous, and what it controls by credit policy is simply the international reserve portion of the monetary base. In a floating rate world, the theory commonly misapplied to a fixed rate world—that the monetary authority controls the money supply and the price level—again becomes valid; but for the monetary approach this merely shifts the focus of analysis from the determination of the balance of payments to the determination of the exchange rate. This is an easy switch of emphasis; it was certainly clear enough to Gustav Cassel [6], who saw purchasing power parity as determining either a nation’s price level via its exchange rate under a fixed rate system, or its exchange rate via its domestic money supply under a floating rate system.
The Contemporary Revival of the Monetary Approach
As documented in the next section, the monetary approach to balance-of-payments theory has a long, solid, and academically overwhelmingly reputable history. The continuity of its development, however, was reversed and the approach suppressed in international economic theory for upwards of a quarter of a century by the events of the 1930s. These included the international monetary collapse of 1931 and after, which produced a situation in which tolerable price and wage level adjustments could not possibly have operated as rapidly to restore monetary equilibrium as classical monetary and balance-of-payments theory assumed, adjustments occurred through quantity changes (mass unemployment), and it was natural, though fundamentally wrong, to regard monetary behaviour as irrelevant to real problems; and the ‘Keynesian revolution’ in monetary theory, which had the effect of making quantity rather than price adjustments the focus of theorising and money relevant, if at all, only as an influence through the effects on investment flows of interest rates, as influenced by the real quantity of money. (In fairness, however, it must be emphasised that Keynes did not deal with an open economy or a closed but international monetary system, and that his closed economy analysis incorporated demand for and supply of money and the necessity of equilibrium between them as part of his general equilibrium framework; also that the better trained and more perceptive of the economists who applied the multiplier analysis to an open economy were quite aware of what monetary influences and equilibrating forces they were excluding from consideration.)
The modern revival of the monetary approach may be said to have begun, in an important but indirect sense, with James Meade’s The Balance of Payments [35]. The sense is in part an unintended one—Meade’s taxonomic compulsion led him to close the monetary side of his general model with a formal analysis of the gold standard—and in part a Pickwickian one, since Meade, in contrast to many of his contemporaries, spelt out the formal structure of his model and ways in which it could be modified, at great length, and it was easy for successors schooled in Meade’s analytical structure to extend the model to tackle new problems and in so doing to become aware of serious faults requiring repair or reconstruction of the model. This was particularly true of Johnson’s work on the generalisation of the theory of the balance of payments, and of Mundell’s concern with the theory of the policy mix on the assumption of international capital mobility, the first of which emphasised the fundamentally monetary nature of balance-of-payments disequilibria, and the second of which led gradually to the recognition that in a capital-mobile world the central bank controls not the money supply and employment, but domestic credit and the balance of payments. From the formal point of view, the main defect of Meade’s work was the confusion of the marginal propensity to save with the marginal propensity to hoard, and the treatment of hoarding as a flow related to the income flow rather than as a transient stock-adjustment or portfolio-balancing phenomenon. Other writers of the same period, especially A. C. Harberger [14], had the insight to recognise that the concept involved was more properly described as a marginal propensity to hoard than as a marginal propensity to save; but this terminology if anything diverted attention from the inappropriateness of treating hoarding as a flow concept.
Significant, but at the time unappreciated, contributions to the revival and formal refinement of the monetary approach, were a short-lived burst of theoretical interest at the International Monetary Fund in the analytical foundations of the Fund’s practices in prescribing for its balance-of-payments-crisis clients, typically small, inflation-ridden economies in Latin America—most notably a theoretical contribution by S. J. Prais [44]; an isolated treatment by F. H. Hahn [13] of the effects of devaluation in the framework of monetary general equilibrium analysis propagated by Don Patinkin’s closed-economy analysis in Money, Interest, and Prices [42], which besides being starkly mathematical confined itself to the short-run impact effect of devaluation; an important contribution by I. F. Pearce [43] and a series of articles by Murray C. Kemp [21, 22] unfortunately rather aside from the main stream of international monetary theory, which worked the analysis through from the concept of a marginal propensity to hoard to a proper monetary analysis on contemporary lines. There were also some significant papers by R. I. McKinnon [33, 34], who emphasised the endogenity of the money supply and the role of public finance and capital mobility; some important contributions by R. Komiya [27, 28], and a useful analysis by A. Collery [7]. The core of the development of the monetary approach as commonly understood, and credited to the author, is to be found in the work of R. A. Mundell which, as already noted, began with a thoroughly Meade–Keynesian development of Meade’s analysis to take account of two developments in the international monetary system that occurred subsequently to Meade’s major work. One was the growing reluctance of countries to resort to either depreciation or appreciation of their currencies to correct balance-of-payments imbalances, together with the imposition of restraints on the ability of countries to use exchange controls and trade interventions—respectively through the restoration of European currency convertibility in 1958, and through successive GATT agreements on the lowering of tariff barriers and the elimination of quotas. This ruled out the practical application of the general structure of Meade’s analysis, which assumed the use of two policy instruments—control over aggregate expenditure via fiscal or monetary policy, and control over the allocation of total expenditures between domestic and foreign goods, by exchange rate adjustment or variation of trade and exchange controls—to serve two policy objectives. On the other hand, currency convertibility and capital mobility introduced a differentiation between expansionary fiscal and expansionary monetary policy, the former improving the capital account by raising interest rates and the latter worsening the capital account by lowering interest rates. With two independent policy instruments again available, it was theoretically possible to serve the two policy objectives of full employment and balance-of-payments balance (internal and external balance) at fixed exchange rates by an appropriate ‘mix’ of fiscal and monetary policies.
The theory of fiscal-monetary mix, though important for policy theorising in the early 1960s, was essentially a by-product of Mundell’s gradual realisation that with capital mobility the money supply ceased to be exogenously determined by the monetary authority, and monetary policy’s prime role was to influence the international flow of reserves. This realisation was accompanied by a switch of analytical method from the Keynesian multiplier analysis to analysis in terms of the Walras’ Law relation between excess demands and supplies in various markets, the need for equilibrium simultaneously in two out of three markets in a three-market system, and the dynamics of reaction of a system out of equilibrium, a method of analysis popularised by Patinkin, but going back to Hicks’ Value and Capital. This development, embodied in a long series of scattered articles subsequently collected in Mundell’s two books on International Economics [38] and Monetary Theory [39], and taught to his students in his courses at Chicago and in the International Trade Workshop, has been the main stimulus to the rapid development of theoretical and subsequently also of empirical work on the monetary approach.
2 SOME DOCTRINAL ASPECTS OF THE MONETARY APPROACH1
As was indicated earlier, the monetary approach to the balance of payments has a long tradition originating with the writings of the classical school in economics. It is for students of the history of economic thought to study in full scholarly detail the exact origins and evolution of the various ideas that are embodied in the monetary approach. In the present section we make no claim to provide such a comprehensive analysis. Rather, our purpose is to cite the writings of some of the eminent classical and neo-classical economists to document our assertion that throughout the last two centuries the monetary approach to the balance of payments has been the dominant intellectual approach to that collection of economic problems, even though, as always in intellectual history, progenitors for other approaches can be found.
A. The Integrated World
One of the major cornerstones of the classical and the neo-classical schools that is adopted by the monetary approach is the conception of a system of world markets. Profit maximisation implies that when there is free and frictionless trade in both securities and goods, rates of return on identical domestic and foreign securities must be equalised, as must the money prices of goods in terms of either currency. These equalities are brought about through international arbitrage. Consequently Mill’s description of the determination of the terms of trade takes as axiomatic the fact that ‘By the fall, however, of cloth in England, cloth will fall in Germany also…. By the rise of linen in Germany, linen must rise in England also.’ (Mill, 1829, Essay I in [36].) This was later formulated, in 1856, by Whewell [51] as the ‘principle of international prices’, and was still later given a policy-oriented expression in Cassel’s Purchasing Power Parity [6]. Of course, when goods are not identical and when transportation costs exist, the simple version of the purchasing power parity will not hold exactly; but for practical purposes the deviation from parity would be confined between very narrow limits and in any case would be logically secondary to the main proposition. These practical considerations led Wicksell to conclude that ‘there could not possibly exist different prices of the same commodity on both sides of the frontier …, difference of prices in the two countries [w]ould be theoretically impossible and practically confined between very narrow limits’. (Wicksell, 1918 [52], p. 405.) Therefore, aside from the cost of carriage, arbitrage will assure that the ‘law of one price’ holds:
‘[t]he action of the international markets, with telegraphic quotations from every part of the world precludes the supposition that gold prices would in general remain on a higher level in one country than another (cost of carriage apart) even for a brief time, because in order to gain the profit, merchants would seize the opportunity to send goods to the markets where prices are high.’ (Laughlin, 1903 [30], p. 369.)
The relevant market in which prices are determined, is the world market: ‘in truth, the value of gold is an affair of a world demand and a world supply’ (Laughlin, 1903 [30], p. 370). Similarly Hawtrey, in his description of the organisation of the market, concludes that:
‘[the] revolutionary changes in the means of communication … have unified markets to such a degree that … there is practically a single world market and a single world price…. It was fallacious to explain the adjustment wholly in terms of the price level. There was even at that time [Ricardo’s time], an approximation to a world price.’ (Hawtrey, 1932 [18], p. 144.)
The proposition that markets are unified due to arbitrage activities applies to a wide spectrum of transactions. This is the basis for Ricardo’s Reply to Mr. Bosanquet, which contains a description of the mechanism of adjustment which assumes consistency of cross exchange rates:
‘[T]heory takes for granted, that whenever enormous profits can be made in any particular trade, a sufficient number of capitalists will be induced to engage in it, who will, by their competition, reduce the profits to the general rate of mercantile gains. It assumes that in the trade of exchange does this principle more especially operates; it not being confined to English merchants alone, but being perfectly understood, and profitably followed, by the exchange and bullion merchants of Holland, France and Hamburgh; and competition in this trade being well known to be carried to its greatest height.’ (Ricardo, 1811 [46], pp. 9–10.)
Similarly, the concept of a unified world capital market and an understanding of the role of capital mobility were an integral part of the analysis of the classical school. Such statements as ‘capital is becoming more and more cosmopolitan’ (Mill, 1893 [37], book III, ch. XVII), and ‘A cosmopolitan loan fund exists which runs everywhere as it is wanted, and as the rate of interest tempts it’ (Bagehot, 1880 [2]) were frequently made. In fact, securities are in a ‘continual course of transition from places where the rate of interest is high to places where it is low’. (Fullerton, 1845 [9], p. 149.)
B. The Natural Distribution of Money Among Countries
The classical assumption that the various countries in the world comprise a well ordered system which is interdependent led to a concept of a natural equilibrium which is similar to the equilibrium concepts of the biological and physical sciences. One of the implications is the proposition that there is a ‘natural distribution of specie’ among the countries of the system.
The natural distribution of species hypothesis appeared in print more than 250 years ago in an impressive pamphlet by Isaac Gervaise: ‘A Nation cannot retain more than its natural Proportion of what is in the world and the Balance of Trade must run against it.’ (Gervaise, 1720 [10], p. 12.) The natural distribution theorem was regarded as almost a law of nature: ‘All water, wherever it communicates, remains always at a level. Ask naturalists the reason; they tell you, that were it to be raised in any one place, the superior gravity of that part not balanced must depress it.’ (Hume, 1752 [19], pp. 62–4.) The mechanism which assures this unique distribution is competition and arbitrage, as stated by Ricardo: ‘Gold and Silver having been chosen for the general medium of circulation, they are, by the competition of commerce, distributed in such proportions amongst the different countries of the world as to accommodate themselves to the natural traffic.’ (Ricardo, 1821 [47], p. 123.) What is the ‘natural traffic’ is left ambiguous—an ambiguity which is typical of many of the classical writings on the determinants of the demand for money. There are, however, quite a few classical economists who clarified understanding of the determinants of the demand for money, as documented by Patinkin in his historical appendix to Money Interest and Prices. With respect to the international distribution, Gervaise left very little ambiguity:
‘Supposing two equal Nations, and that one hath a Power or Right over the other; as, for example, one quarter of the yearly Produce of its Labour be expended in the other: in that case the imperial creditor Nation will support a Denominator one quarter above its natural Proportion …, and the subjected debtor country will support but three quarters of its natural Denominator.’ (Gervaise, 1720 [10], pp. 23–4.)
Thus, the demand for cash balances depends on income (or assets), rather than having merely a mechanical dependence on output. The explicit recognition of an important corollary of the ‘natural distribution’ theorem, the neutrality of money from the world standpoint, was provided by Mill: ‘A newly acquired stock of money would diffuse itself over all countries until money has diffused itself so equally that prices had risen in the same ratio in all countries, so that the alteration of price would be for all practical purposes ineffective.’ (Mill, 1893 [37], Book III, pp. 194–5.)
The important implication of the natural distribution hypothesis for economic policy is that under a fixed exchange rate the supply of money in any country is endogenous. The structural relationship through which this endogeneity is effected is the balance of payments. Consequently, the credit policy of the banking system is linked directly to the balance of payments, and the adjustment mechanism cannot be properly analysed without focusing on monetary policy and the resultant excess flow demand or supply of money. In fact, since the balance of payments reflects an excess flow demand for, or supply of, money, it must—in the absence of continuous domestic credit creation or economic growth—be a transitory phenomenon. This was the major criticism that Hume advanced against mercantilism: ‘Suppose twenty million was brought into Scotland …, how much would remain in the quarter of a century? Not a shilling more than we have at present.’ (Hume, 1752 [19], pp. 197–8.) Thus, the gradual process of adjustment will continue as long as there is an excess stock supply of money (which induces an excess flow supply).
This leads to one of the basic principles of the classics according to which there can be no deficit unless there is an excess supply of money: ‘The temptation to export money in exchange for goods, or what is termed an unfavourable balance of trade, never arises but from a redundant currency.’ (Ricardo, 1809 [45], p. 267.) Although some of the classical economists used the term ‘balance of trade’ instead of ‘balance of payments’, it is clear that some of them recognised the consequences of international mobility of capital, and incorporated these consequences into their analysis. According to Mill, a once-and-for-all rise in the money stock ‘would create a sudden fall in the rate of interest, which would probably send a great part of the twenty millions of gold out of the country as capital…. All prices would rise greatly.’ (Mill, 1893 [37], Book III, pp. 195–6.) How much of the adjustment will be reflected in the trade account and how much in the capital account was not analysed explicitly. As an empirical matter, however, Mill recognised that the role played by the capital account was larger than allowed for by previous writers: ‘It is a fact now beginning to be recognised that the passage of the precious metals from country to country is determined much more than was formerly supposed, by the state of the loan market in different countries, and much less by the state of prices.’ (Mill, 1893 [37], ch. VIII, p. 4.) Since the various behavioural relationships are interrelated by the budget constraints, an excess supply of money should be reflected in all of the other accounts of the balance of payments—a proposition which is clearly emphasised by Hawtrey and is an integral part of the monetary approach: ‘An expansion of credit causes an unfavourable balance of payments … the change in the balance is no more than a symptom. And of course the balance that is in question is always the balance of payments, not the balance of trade in the narrower sense of the difference between exports and imports of material commodities.’ (Hawtrey, 1927 [17], p. 75, italics in original.)
C. Absorption versus Relative Prices
As was argued in the previous section, the absorption and the monetary approaches to the balance of payments emphasise the fact that accumulation or decumulation of assets depends on the relationship between expenditures and income and does not depend on the composition of expenditures. Although the typical textbook version of the classical adjustment mechanism attaches much importance to variations in relative prices, there is much evidence that an adjustment theory which is based on the relationship between expenditures and income was widely accepted by the classic and, even more, the neo-classic writers.
One of the very first writers to emphasise the role of divergences between expenditures and income without mentioning relative prices is Gervaise (1720). A rise in the monetary stock raises expenditures and induces a deficit. The process continues until the excess stock is diffused, and simultaneously, once the stock is reduced to its equilibrium level, expenditures and income coincide:
‘If a Nation adds to its Denominator … [t]he rich … consume more Labour [goods] than before, … so that there enters in that Nation, more Labour than goes out of it…. The deficit is paid in Gold and Silver until the Denominator be lessen’d, in proportion to other Nations; which also, and at the same time, proportions the number of the Poor [consumption] to that of Rich [production].’ (Gervaise, 1720 [10], p. 7.)
If the balance of payments is viewed as a monetary phenomenon it is reasonable to apply the general principles of adjustment to monetary changes. This approach was adopted in the early 1730s by one of the originators of the theory of adjustment to monetary changes, Richard Cantillon: ‘The increase in money will bring about an increase in expenditure … states which have acquired a considerable abundance of money ordinarily import many things from neighboring countries where money is scarce.’ (Cantillon, 1735 [5], pp. 263–7.) Even Hume and Mill, who are usually quoted as the main proponents of the relative price theory, have made statements implicitly endorsing the contrary view of the adjustment mechanism. Hume in ‘Of the Jealousy of Trade’ (1758) writes: ‘The inhabitants, having become opulent and skillful, desire to have every commodity in the utmost perception; … they make large importations from every foreign country.’ (Hume [19], pp. 79–80.)
Mill’s statement—on the basis of which, and other similar statements, he has been interpreted as supporting the relative price theory of adjustment—is as follows: ‘The natural effect of a rise in the money stock would be a rise in prices. This would check exports and encourage imports.’ (Mill, 1893 [37], Book III, p. 194.) Such statements were later criticised by Bastable who argued that people ‘having larger money incomes will purchase more at the same prices and thus bring about the necessary excess of imports over exports’. (Bastable, 1889 [3].) However, it is quite possible that Mill’s concepts of ‘cheapness’ and ‘dearness’ do not necessarily refer to actual price changes since he argues explicitly that: ‘I say this being well aware that the article would be actually at the very same price … in England and in other countries. The cheapness however of the article is not measured solely by the money price, but by the price compared with money incomes of the consumers.’ (Mill, 1893 [37], Book III, p. 187.) Similarly, ‘the English public, having more money, will have a greater power of purchasing foreign commodities … there will be an increase of imports; and by this, and the check to exportation, the equilibrium of imports and exports will be restored’. (Mill, 1893 [37], ibid.)
We do not intend, however, to convey the impression that changes in relative prices did not play a role in the classical adjustment mechanism, but rather, that such changes were not treated as the essential factor in the process of adjustment.
In fact, the question of whether the adjustment process operates mainly through changes in relative prices, or through divergences between expenditures and income, was the subject of a series of debates that were mainly associated with the transfer problem. Among the key figures in these debates, Taussig, on the one hand, believed that changes in relative prices are an essential part of the process of adjustment: ‘The Gold moves, not “automatically”, but as a result of changes of prices, or (for short periods) of changed rates of discount.’ (Taussig, 1918 [49], p. 411.) On the other hand, Wicksell argued forcefully that changes in relative prices are of secondary importance: ‘The stimulus to these altered conditions of trade is not to be found in a difference of prices in the two countries … the increased demand for commodities in one country, the diminished demand in the other, would in the main be sufficient to call forth the change alluded to.’ (Wicksell, 1918 [52], p. 405.)
A similar debate culminated in the discussion between Keynes [23], Ohlin [40] and Rueff [48] in a series of articles in the Economic Journal (1929), and the major issues were summarised by Haberler [12], Iversen [20], Kindleberger [26], Ohlin [41] and Machlup [32, pt. 5]. By the 1930s it had become generally accepted that divergences between expenditures and income are important in the process of adjustment, as reflected by Kindleberger’s statement:
‘Present-day writers generally agree that the price-specie-flow mechanism, with its emphasis on price levels, does not embrace the whole process of adjustment. Changes in demand schedules which have their origin in alterations in national income, which again can be related to changes in money or its velocity, perform a large part of the task of adjusting the balance of payments directly, without the necessity of price movements.’ (Kindleberger, 1937 [26], p. 19.)
Although the analytical issues were clarified by that time, much was left to be done in assessing the empirical magnitudes of the alternative mechanisms. An early observation was provided by Bresciani-Turroni in his analysis of the German situation: ‘I have the impression that the increase in imports was due much more to an increase in German demand (in the schedule sense) for foreign goods than to an expansion in demand provoked by a diminution of prices in lending countries.’ (Bresciani-Turroni, 1932 [4], p. 93.)
D. Banking Policies
The central implication for economic policy of viewing the balance of payments as a monetary phenomenon is that as long as the exchange rate is fixed, monetary policy in the form of control over credit creation has a direct effect on the balance of payments. This strong link has been recognised throughout the development of balance of payments theory. Again Gervaise is among the first to emphasise this link: ‘Increase in Credit will act on that Nation as if it had drawn an equal sum from a Gold or Silver Mine, and will preserve but its Proportion of that Increase.’ (Gervaise, 1720 [10], p. 9.) Thus, the money supply is completely endogenous and credit policy is eventually reflected only in changes in the composition of the assets of the central bank. Similar statements can be found in Hume, 1752 [19], p. 70, and Mill, 1893 [37], Bооk III, pp. 195–6.
Since the balance of payments is a monetary phenomenon, the proper way to correct it is through an appropriate monetary policy:
‘… a disturbance of the balance of payments from a change in the demand or supply of commodities, whether seasonal or other, is properly met by a suitable adjustment of credit. The transmission of gold is only necessary insofar as the adjustment of credit does not exactly keep pace with disturbance. The loss or gain of gold as the case may be tends to bring about the contraction or expansion of credit required. That, however, is pure mismanagement. It has no real connection with the Balance of Payments, and is merely a special case of monetary disturbance.’ (Hawtrey, 1926 [16], p. 56.)
The policy implications are clear and in anticipation of Mundell’s celebrated theory of the policy mix, Hawtrey concludes that ‘It is a mistake to lay too much stress on the actual means of discharging the trade balance. An adverse trade balance may be created by a redundant currency and can be corrected by a contraction of the currency.’ (Hawtrey, 1927 [17], p. 290.)
Since the world as a whole is a closed system, the balance of payments reflects monetary policies in the various countries. Consequently it is not determined solely by the domestic rate of credit creation but rather by the domestic rate relatively to that in the rest of the world: ‘One lets credit expand a little faster than the others and loses gold; another lags behind and receives gold.’ (Hawtrey, 1927 [17], p. 10.)
The basic truism that the world is a closed system that is composed of interdependent countries implies that an individual central bank in the world system has no more power than a single member bank of a national system. This simple fact and its policy implications were widely recognised and accepted. Keynes himself, in his pre- General Theory [25] writing, forcefully emphasised the mutual interdependence of the various central banks:
‘One or two Central Banks acting alone would not, unless they were very preponderant in size, be able to change the weather or direct the storm—any more than a single member bank can control the behaviour of a national system. If a single Central Bank lags behind in a boom it will be overwhelmed by excess reserve-resources, and if it steps in in a depression it will have its reserve-resources quickly drawn from it…. each Bank is necessarily governed by the average policy of the Banks as a whole.’ (Keynes, 1930 [24], pp. 281–6, italics in original.)
It is quite clear from these statements why the monetary approach to the balance of payments views the balance of payments as essentially a monetary phenomenon.
As was already emphasised, a theory of the balance of payments has very little to do with the composition of expenditures as between various aggregates of commodities, and therefore the effects of a devaluation on the terms of trade have little to do with their effects on the flow of reserves. Historically, however, much of the analysis of the effects of changes in the exchange rate was carried out by using the elasticity approach. Using this framework, the typical disaggregation of commodities has been between exportables and importables with much emphasis on the terms of trade. The popular version of that approach ignores considerations of general equilibrium and is marred by assuming that the behavioural equations depend on money prices rather than on relative prices and real income or ‘expenditure’ as well as by ignoring the consequences of the budget constraint. More sophisticated treatments have assumed that there are some things ‘in the background’ which assure consistency, though the nature of these is rarely, if ever, investigated in detail.
The monetary approach rejects this emphasis given to the role of relative prices in the analysis of devaluation. This rejection and the reasons for it are noted in the classical approach and are foreshadowed by analytical statements by some of the writers in that tradition. For example, Graham forcefully stated that ‘so far as balance is concerned, elasticity is never of any relevance, either in the short run or the long run…. the whole discussion of elasticities in international trade has been of negligible or even negative value’. (Graham, 1949 [11], p. 249.)
The basic claim that is made by the proponents of the monetary approach is that the balance of payments effects of any policy measure cannot be properly analysed without specifying the monetary consequence of the policy itself. This principle implies that in contrast with the traditional textbook analysis of the effects of a tariff, which assumes without further discussion that a diversion of demand away from imports must improve the trade and payments balance, a proper analysis should investigate the effects of the tariff on the excess demand for money. Consequently, a tariff will improve the balance of payments only if it induces an excess demand for money. This rather simple condition is very different from the typical textbook analysis which emphasises the effect of the tariff on the relative price of goods.
Even this insight originates with the writings of earlier economists, such as Hawtrey: ‘… protective tariff does raise the price level … it accordingly requires an increased monetary circulation, and if the monetary system is such that that cannot be provided without the importation of gold, gold will be imported.’ (Hawtrey, 1932 [18], p. 244.)
We conclude this section by repeating that there is plenty of evidence to support the claim that throughout the last two centuries the monetary approach to the balance of payments was widespread.2
REFERENCES
[1] Alexander, Sidney, ‘The Effects of Devaluation: A Simplified Synthesis of Elasticities and Absorption Approaches’, American Economic Review, xlix, no. 2 (March 1959), 22–42.
[2] Bagehot, W., Economic Studies, ed by R. H. Hutton (London, 1880).
[3] Bastable, S. F., ‘On Some Applications of the Theory of International Trade’, Quarterly Journal of Economics, iv (1889), 1–17.
[4] Bresciani-Turroni, C., Inductive Verification of the Theory of International Payments (Cairo, Noury and Son, 1932).
[5] Cantillon, Richard, ‘Essai sur la nature du commerce en general (1755)’ in Arthur E. Monroe (ed.), Early Economic Thought (Cambridge, Harvard University Press, 1927).
[6] Cassel, Gustav, Post-War Monetary Stabilization (New York, 1928).
[7] Collery, Arnold, ‘International Adjustment, Open Economies, and the Quantity Theory of Money’, Princeton Studies in International Finance, no. 28 (June 1971).
[8] Frenkel, Jacob A., ‘Adjustment Mechanism and the Monetary Approach to the Balance of Payments: A Doctrinal Perspective’, presented at the Third Paris-Dauphine Conference on Recent Issues in International Monetary Economics, March 1974; forthcoming in the Proceedings, edited by E. Claasen and P. Salin.
[9] Fullerton, John, On the Regulation of Currencies, 2nd edn, 1845.
[10] Gervaise, Isaac, ‘The System or Theory of the Trade of the World, 1720’; reproduced as a reprint in Economic Tracts (Johns Hopkins University Press, 1956).
[11] Graham, Frank D., ‘A Letter from Graham to R. Hinshaw’, Journal of International Economics, i, no. 2 (May 1971), 249.
[12] Haberler, Gotfried, The Theory of International Trade (London, Hodge, 1936).
[13] Hahn, F. H., ‘The Balance of Payments in a Monetary Economy’, Review of Economic Studies, xxvi (2), no. 70 (February 1959), 110–25.
[14] Harberger, A. C., ‘Currency Depreciation, Income and the Balance of Trade’, Journal of Political Economy, lviii, no. 1 (February 1950), 47–60.
[15] Harrod, Roy F., International Economics, 1st edn, 1933, 2nd edn, 1939, 4th edn, 1957.
[16] Hawtrey, R. G., ‘The Gold Standard and the Balance of Payments’, Economic Journal, xxxvi (March 1936), 50–68.
[17] Hawtrey, R. G., Currency and Credit (London, Longmans, Green, 3rd edn, 1927).
[18] Hawtrey, R. G., The Art of Central Banking (London, Longmans, Green, 1932).
[19] Hume, David, ‘Political Discourses, 1752’ in E. Rotwein, David Hume; Writings on Economics (London, Nelson, 1955).
[20] Iversen, Carl, Aspects of the Theory of International Capital Movements (London, Oxford University Press, 1935).
[21] Kemp, Murray C., ‘The Rate of Exchange, The Terms of Trade and the Balance of Payments in Fully Employed Economics’, International Economic Review, iii, no. 3 (September 1962), 314–27.
[22] Kemp, Murray C., ‘The Balance of Payments and the Terms of Trade in Relation to Financial Controls’, Review of Economic Studies, xxxviii (January 1970), 25–31.
[23] Keynes, John M., ‘The German Transfer Problem’, Economic Journal, xxxix (1929), 1–7.
[24] Keynes, John M., A Treatise on Money, vol. II (New York, Harcourt, Brace, 1930).
[25] Keynes, John M., The General Theory of Employment Interest and Income (London, Macmillan, 1936).
[26] Kindleberger, Charles P., International Short-Term Capital Movements (New York, Columbia University Press, 1937).
[27] Komiya, Ryutaro, ‘Monetary Assumptions, Currency Depreciation and the Balance of Trade’, The Economic Studies Quarterly, xvii, no. 2 (December 1966), 9–23.
[28] Komiya, Ryutaro, ‘Economic Growth and the Balance of Payments: A Monetary Approach’, Journal of Political Economy, 77, no. 1 (January/February 1969), 35–48.
[29] Kouri, P. J. K., and Porter, M. G., ‘Portfolio Equilibrium and International Capital Flows’, Journal of Political Economy, 82, no. 3 (May/June 1974), 443–68.
[30] Laughlin, J. Lawrence, The Principles of Money (New York, Scribner, 1903).
[31] Machlup, Fritz, International Trade and the National Income Multiplier (Philadelphia, Blakiston, 1943).
[32] Machlup, Fritz, International Payments, Debts and Gold (New York, Scribner, 1964).
[33] McKinnon, Ronald I., ‘Portfolio Balance and International Payments Adjustment’ in Mundell, R. A., and Swoboda, A. K. (eds), Monetary Problems of the International Economy (Chicago, University of Chicago Press, 1968), 199–234.
[34] McKinnon, Ronald I., and Oates, W. E., ‘The Implications of International Economic Integration for Monetary, Fiscal and Exchange Rate Policy’, Princeton Studies in International Finance, no. 16 (1966).
[35] Meade, James E., The Balance of Payments (London, Oxford University Press, 1951).
[36] Mill, John Stuart, Essay on Some Unsettled Questions of Political Economy (London, 2nd edn, 1874).
[37] Mill, John Stuart, Principles of Political Economy (New York, Appleton, 5th edn, 1893).
[38] Mundell, Robert A., International Economics (New York, Macmillan, 1968).
[39] Mundell, Robert A., Monetary Theory (Pacific Palisades, Goodyear, 1971).
[40] Ȯhlin, Bertil, ‘The Reparation Problem: A Discussion’, Economic Journal, xxix (June 1929), 172–8.
[41] Ohlin, Bertil, Interregional and International Trade (Cambridge, Mass., Harvard University Press, 1933).
[42] Patinkin, Don, Money Interest and Prices, 2nd edn (New York, Harper & Row, 1965).
[43] Pearce, I. F., ‘The Problem of the Balance of Payments’, International Economic Review, 11, no. 1 (January 1961), 1–28.
[44] Prais, S. J., ‘Some Mathematical Notes on the Quantity Theory of Money in an Open Economy’, IMF Staff Papers, viii, no. 2 (May 1961), 212–26.
[45] Ricardo, David, The High Price of Bullion (1809) (McCulloch edn, 1846).
[46] Ricardo, David, Reply to Mr. Bosanquet’s Practical Observations on the Report of the Bullion Committee (London, 1811).
[47] Ricardo, David, Principles of Political Economy and Taxation, 3rd edn, 1821.
[48] Rueff, Jacques, ‘Mr. Keynes’ Views on the Transfer Problem: I. A Criticism’, Economic Journal, xxxix, no. 155 (September 1929), 388–99.
[49] Taussig, F. W., ‘International Freight and Prices’, Quarterly Journal of Economics, xxxii (February 1918), 410–14.
[50] Viner, Jacob, Studies in the Theory of International Trade (New York, Harper, 1937).
[51] Whewell, W., ‘Mathematical Exposition of Some Doctrines of Political Economy, Second Memoir’, Transactions of the Cambridge Philosophical Society, ix, pt I (1856).
[52] Wicksell, Knut, ‘International Freights and Prices’, Quarterly Journal of Economics, xxxii (February 1918), 404–10.