RDP 1977-07: Money and the Balance of Payments 2. Theoretic Background[3]

In classical closed-economy theory, an increase in money will, in the long run and as a first approximation, raise only the price level, although it is recognised that output may be increased during the transition to a new equilibrium. The parallel result for a small open economy with a fixed exchange rate is that the full adjustment to an increase in domestic credit is borne by the balance of payments, while domestic prices are determined by world prices, and output is fixed at its equilibrium level.[4]

Just as the development in the last century of closed-economy monetary theory was primarily concerned to clarify the “transmission mechanism” which lay behind the traditional quantity equations,[5] the recent literature on the theory of money in the open economy has attempted to develop a theory of balance of payments determination which provides the dynamic underpinnings of the pioneering results of Johnson, Mundell, and others.

Some of this work has stressed that, in a world of highly integrated markets, adjustment to exogenous disturbances can occur sufficiently rapidly that the “long-run” results of the earlier writers are in fact descriptive of the real world.[6] These authors typically solve an open-economy portfolio balance model on the assumption that equilibrium is achieved within the observation period in financial markets, while income and prices remain fixed. In these models, with money demand, and the supply of money from sources other than capital flows (or the balance of payments), determined exogenously, the net capital inflow (or the change in international reserves) is simply equated to the excess demand for money.

The other major strand in this literature has focussed on the role of frictions, which are seen to be inherent in the operation of economic systems.[7] This work draws on the tradition of the classical “price specie–flow” mechanism, which envisaged adjustment to excess domestic credit taking place through rising domestic prices,[8] with the usual balance of trade effects followed by gold shipments which eventually eliminated the excess money. Thus this approach assumes, not only that the balance of payments effects of a monetary disturbance materialise over time, but that strong effects on domestic output and prices may be necessary to generate the long-run responses predicted by the simple equilibrium models. This should not be taken to simply, however, that the most general model in a frictional world could not involve both direct and indirect links between excess money and the balance of payments.

It is important to note, in this context, that if the transmission of a monetary disturbance to the balance of payments operates only through indirect channels, then the “monetary approach to the balance of payments” could be viewed as equivalent to the conventional absorption approach, with the crucial addition of a disequilibrium real balance effect on consumption spending.[9] The emphasis on what is claimed by proponents of the monetary approach to be the crucial role of domestic monetary disequilibrium as a causal determinant of the balance of payments, is probably best explained as a reaction to the absorption models which, by ignoring or downplaying real balance effects, implied that flow equilibrium in goods markets was sufficient for balance in international payments.

Some conventional models do allow monetary conditions to influence the balance of payments through interest rate effects on capital flows. However, even with a theoretically appropriate stock-adjustment model of capital flows and an endogenous trade account, with changes in international reserves feeding into the money supply process, conventional models do not usually assign a role to monetary disturbances which is sufficiently powerful to generate the short-run offsets suggested by many proponents of the monetary approach to the balance of payments.[10]

The issue of the speed with which changes in international reserves act to directly offset a domestic monetary disturbance is, as noted in the introduction, a crucial one, since the more rapid and complete the balance of payments offset, the less is the quantity of excess money which remains to influence domestic variables. One important question is whether all the effects of a monetary disturbance, both domestic and external, appear quickly enough to restore the money market to equilibrium in a short period of time.[11] This is a particularly relevant issue in highly disturbed periods, when successive monetary shocks at closely spaced intervals must be taken into account. The models presented in the following section allow the economy to be off the demand for money function in the short run, and in some versions provide an estimate of the average lag required for an initial excess of money balances to be eliminated through its effects, both direct and indirect, on income, prices, interest rates and the balance of payments.


A general theoretic review and history of the monetary approach to the balance of payments is provided in the introductory essay of Frenkel & Johnson (1976); the current discussion concerns issues which are of direct relevance to the empirical analysis of alternative adjustment mechanisms. [3]

Since the forces which determine the balance of payments under a given exchange rate will also influence the exchange rate when it is responsive to market pressure, the discussion here is readily adaptable to the analysis of alternative exchange rate regimes. [4]

This development is excellently described by Marget (1938), pp.81 ff. [5]

An important and influential example of this approach is provided by Kouri and Porter (1974); other similar studies are contained in Frenkel and Johnson (1976). [6]

Examples of this approach are contained in Laidler (1972), Johnson (1975), and several macroeconometric models with well developed balance of payments and monetary sectors. [7]

The price effects themselves emerged as a consequence of the initial effects on the level of activity; see Rotwein (1955) pp. 37–38. [8]

The absorption approach, as set out for example by Alexander (1952), allowed that real balance effects could influence spending and hence the balance of trade, but wealth effects were regarded as relatively unimportant in practice, from Patinkin (1956) onwards. Some recent writers, for example Jonson and Kierzkowski (1975), have provided a formal reconciliation of the monetary and conventianal theories of the balance of payments in disequilibrium models with both money and goods. This result suggests that a disequilibrium real balance effect on domestic spending is likely to be important in practice, and recent evidence, such as that presented in Johnson, Mussa and Wymer (1977), bears this out. [9]

An important exception is the RDX2 model of the Canadian economy, which Helliwell and Lester (1976) have shown to generate substantial monetary offsets through the balance of payments. This result, however, appears to rely on a specification of short-term capital flows which imposes a substantial balance-of-payments offset to changes in the supply of money. [10]

Another related issue, of the different implications of monetary disturbances arising from alternative sources, has been addressed in a recent paper by Borts and Hanson (1976) within a model which allows monetary factors to have short-run domestic effects, but which does not investigate in detail the alternative mechanisms through which the effects are produced. [11]