the keynesian and classical determination of the exchange rate

The Keynesian and Classical Determination of the Exchange Rate

By

Emil-Maria Claassen

C o n t e n t s : I . I n t r o d u c t i o n . - - I I . K e y n e s i a n A p p r o a c h . - - I I I . Classical A p p r o a c h . - - IV. C o n c l u d i n g R e m a r k s .

I. Introduction

T he exchange rate is a macroeconomic variable whose determinants

depend on the choice of the underlying macroeconomic model. Two criteria can be used for the selection of the appropriate

macroeconomic model:

- - B y recognizing tha t the existing macroeconomic models for open economies are still ra ther simple in comparison to those of a closed economy, one could choose between two extreme cases: the Keynesian assumption of a given price level and the classical assumption of a given employment level. Even though there are intermediate cases between a fixed-price model and a fixed-quantity model, for didactical reasons we shall expose the extreme ones in order to show tha t they are surprisingly similar with respect to the determinants of the real exchange rate.

- - One-country models are used for small open economies where foreign variables are considered as given and two-country models are constructed when domestic and foreign macroeconomic variables are interdependent. We know from the literature tha t floating exchange rates isolate the countries from each other only when there is no capital mobili ty (and b y ignoring the Laursen-Metzler effect). Con- sequently, there must be the highest degree of an international transmission of disturbances in the case of perfect capital mobility (in the sense of a perfect substi tutabil i ty of domestic and foreign financial assets). We will show such an international transmission mechanism of disturbance in the traditional terms of an expansionary monetary

Remark: This note represents parts of a research work supported by the Deutsche For- schungsgemeinschaft and elaborated within its special program oxt "Inflation and Employment ia Open Economies". The author would like to thank the unknown referee for his helpful comments.

2~

20 Emi l -Mar i a C l a a s s e n

policy in the home country, under the assumption of perfect capital mobility. Again, it is astonishing that the Keynesian and classical frameworks are similar with respect to this transmission mechanism: in the Keynesian model the foreign country will realize a decrease of its real national income and in the classical model the foreign country will observe a fall in its general price level. However, in the Keynesian framework there will be a rise in the real exchange rate whereas the latter will remain unchanged in the quantity-theoretical world, at least as far as the long-run equilibrium values are concerned.

In what follows we shall present firstly the Keynesian two-country model for the determination of the real exchange rate and afterwards the classical two-country model; the latter has to differentiate between the nominal and the real exchange rate.

H. Keynesian Approach

The Keynesian two-country model (under flexible exchange rates and perfect capital mobility) has been developed mainly by Mundell [1968 ] and popularized on a more comprehensive level, for instance, by Dorn- busch and Krngman [I976 , pp. 542---548] and Dornbusch [I98O, pp. i99----2o2 ] . Its characteristic is the mobility of capital which renders possible a disequilibrium in the current account, for example a surplus in the home country and a corresponding deficit in the foreign country, creating an income expansion in the first country and an income contraction in the second one. This international interdependence of income determination is valid a tortiori under perfect capital mobility.

The traditional two-country model of the Mundellian type can be written as:

domestic goods market (x) S (y) - - I (r) - - T (y , y * , e) = 0

Sy > 0, L < 0, T~ < 0, Ty. > 0, Te > 0

domestic money market L (y, r) = M (3)

I . > o , I ~ < o

foreign money market L* (y*, r) = M* (4)

L 'y , > 0, L* r < 0

foreign goods market S* (y*) ~ I* (r) + T (y, y*, e) = 0 (2)

S*y. > O,I* r < 0

Determination of the Exchange Rate 2 I

where S denotes saving, I investment, T the current account surplus, L the demand for money, M the supply of money, y the output, r the interest rate and e the exchange rate - - defined as the domestic currency price of foreign exchange, all variables being expressed in real terms. The asterisk stands for the foreign country.

The solution of the above equation system can be conceived analy- tically in terms of two successive stages even though they are realized simultaneously: (i) the world market for goods [(I) -b (2)] and the world market for money [(3) -4- (4)] determine the world interest rate and world income; (if) the distribution of world income is given by the equilibrium condition in the national money markets. In the case where the equilibrium values for r, y and y* still imply a disequilibrium in the national goods markets (for instance an excess supply of goods in the home country which must be equal to an excess demand for goods in the foreign country because of the equilibrium on the world market for goods), a change in the exchange rate will take place, equilibrating the national goods markets. In our example, a depreciation of the home currency arises which increases total demand in the home country and, correspondingly, lowers total demand in the foreign country. In this sense, the money market determines national income and the goods market the exchange rate.

In Figure I, the equihbrium condition of the domestic money market is described by the LM schedule and the equilibrium condition of the domestic goods market by the IS schedule. Given the world interest rate (ro), the money market fixes the equihbrium value for domestic output (Yo). Assume that there is an excess supply in the goods market such that the exchange rate is at the level U (correspondingly, the foreign country possesses an excess demand for goods). Consequently, the exchange rate will rise to point A by which part of world demand is switched from the foreign country to the home country in order to equilibrate the national goods markets. The positive slope of the IS schedule reflects the assumption that the Marshall-Lerner condition holds, i.e. that T, > 0.

In the upper right-hand panel of Figure I, the IRP line describes the interest-rate parity

r - - - - r* -b0[ (e - -e ) /e ] 0 • 0 ~< 1 (5)

where the last term in the above expression illustrates the expected change in the exchange rate. ~ is the exchange rate which is expected to reign in the long run. The slope of the IRP line is equal to 0e/e 2. At point A, the domestic interest rate corresponds to the foreign interest rate (r o ~ ro* ) such that eo ---- eo- We have included the interest-rate parity in order to show also the immediate effect of an expansionary monetary

2 2 E m i l - M a r i a C l a a s s e n

policy on the exchange rate and the movements of the latter towards its long-run equilibrium value.

An expansionary monetary policy in the home country shifts the LM schedule to the position LM 1. The effect on the exchange rate is described first for a one-country model of the type of the smaU-country hypothesis and later for a two-country world.

Figure I

LM LM r / / / I IRP('o) IRP~2)

. . . . . . . ,/_c__L . . . .

. . . . . . . . . . . . . . . . . . . . . . . ~ .

I = YO Y2 Y eo'e r el

I e 11 / 45+

~I5 I I

- / - , ' B I I / . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - r . . . . - ~ ' B /

'+// . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . t- . . . . ,,k c

/ i I / I /

. . . . . . y "-', . . . . . . . . . . . . . . . . . . . . . ,, ; :

i I /

Y

.,0=~

I. O n e - C o u n t r y Model

At the assumed unchanged interest rate r o, the disequilibrium in the money market will be eliminated by the rise of national income to y~ creating an excess supply of goods (absorption rises less than income). The latter will be reduced to zero by the depreciation towards ~ according to the IS line. Assuming that individuals fix the newly expected exchange rate for the long run at the level e2 after the monetary expansion has taken place, the IRP line shifts to the position IRP (e2)- This rise in output and in the exchange rate is relevant for the long run.

In the very short run, there is an interest-rate autonomy pushing down the domestic interest rate to r 1. As is usually done, we assume that financial assets markets adjust more quickly to any disequilibrium than goods markets such that there is an immediate movement from A to B in the upper left-hand panel of Figure I. In order to equalize the expected

D e t e r m i n a t i o n of t he E x c h a n g e R a t e 23

return between domestic and foreign financial assets, there must be an important depreciation - - (el) above the long-run one (ez) - - in order to create an expected rate of appreciation which has to be equal to the interest-rate differential (see point B in the upper right-hand panel).

The subsequent movement of the domestic interest rate and the exchange rate follows the path from B to C. While there is always a temporary equilibrium in the financial sector, an excess demand for goods exists in the real sector because the lower interest rate (rl) and the higher exchange rate (el) push up national income to y,. The first capital movements take place the moment the depreciation improves the current account because capital flows must be conceived as accommodating any current-account imbalance.

2. T w o - C o u n t r y Mode l

If one takes into account a two-country model, the domestic monetary expansion has the effect of reducing the world interest rate, for instance to r v Consequently, the adjustment path of the exchange rate and of national income will not be BC but B'D (Figure 2). In the long run, part of the initial excess supply in the domestic money market will be absorbed by a higher demand for money, such that the rise in domestic income will

Figure 2

o i LM(r-) LMI(rl) LMI(ro) /:5(rq) I Is{,i) N

~ . . . . . "~ i / / /

/ I (o I

I I I

I I

I r

YO 3/1 Y2 Y

be less in order to equilibrate the national market (see the LM 1 (rl) schedule and the lower equilibrium income Yl). On the other hand, the excess supply in the national goods market will be smaller since the lower interest rate raises total demand (see the IS (rl) schedule). Consequently, the depreciation necessary to equilibrate the national goods markets will be of a lower amount (ex) than that in the one-country model (e2). Because

2 4 E m i i - M a r i a C l a a s s e n

the long-run exchange rate will be lower (el) than the one of Figure z (~), the IRP-line shifts less to the right than in the upper right-hand panel of Figure I such that the immediate depreciation (to e x in Figure z) will be lower (for instance, to e'x); the adjustment path towards D begins at point B' in Figure ~ instead of point B. 1 Furthermore, it should be emphasized that during the adjustment process towards the long-run equilibrium, r must be replaced by r* in equation (z) and (4) so that all variables (r, r*, y, y*, e) are determined simultaneously by the five equations (r) N (5). Only for the long run (point D) is our former statement valid that the money market determines national income in each country and that the goods market fixes the exchange rate. For the very short run (point B'), it is the money market that determines the interest rate and, by this, via the interest-rate parity the exchange rate, while the goods market remains in disequilibrium by definition.

As is well-known from the Mundell model, the international transmission of the domestic monetary disturbance to the foreign country consists in the reduction of the latter's income in the long run. Even though, at the very beginning, an excess demand exists in the foreign goods market (because of the lower interest rate and higher demand for imports in the home country before the depreciation of its currency), the higher money demand in the foreign country as a consequence of the fallen interest rate cannot be satisfied, such that the money demand must be reduced via a lower national income. The excess demand for goods which still exists in the foreign country can only be eliminated by an appreciation of its currency.

TIT. Classical Approach

We have restated the Keynesian model of the Mundell type because it can be shown that, in principle, the classical model comes to the same long-run conclusion with respect to the determination of the real exchange rate by the national goods markets and with respect to the repercussion of an expansionary open-market operation in the domestic country on the

i The change of the long-run exchange rate in a one-country model amounts to:

d~ = [(s + m)/L r Td dM where s is the marginal propensity to save and m the marginal propensity to import. Without presentiu8 a rather complete formula for de in the context of a two-country model, a smaller change in ~ can already be shown within a o~e-country model by assuming a decrease in the interest rate (---dr):

de = [(s -}- m)/Ly To] dM - - {[Lr (s + m) Jr Ly IrJ/Ly Te} dr

The second term of the r ight-hand expression is negative for the case of a decline in the rate of interest.

Dete rmina t ion of the Exchange Ra te 25

foreign economy, now in terms of a reduced foreign price level instead of a lower foreign real income. The transmission mechanism is that of a lower world interest rate which pushes up the foreign demand for real cash balances, involving a fall in the foreign price level for a given foreign money supply.

The quantity-theoretical model, similar in its simplicity to the Keyne- sian two-country model of equations (I)--(4), regards real national incomes as given and dichotomizes the economy into the real and monetary sector. The equation system of the real sector can be written as:

domestic goods market (6)

foreign goods market

S (r, w) - - I (r) - - T (x) = 0 wherew = a + m Sr > 0 , Sw < 0 , Ir < o, T x > 0

S (r, w*) - - I* (r) + T (x) = 0 (7) S* > 0 , S** < 0 , I* < 0 wherew* = a * + m *

domestic market for real cash balances L (r) = m; L r < 0 (8)

foreign market for real cash balances L* (r) = m*; L* < 0 (9)

where w represents the real value of non-human wealth which is supposed to be composed by equities (a) and (outside) money holdings (m = M/P) ; in analogy, w* stands for the real value of non-human wealth held by the foreign country 1.

1 The wea l th of the home co tmt ry can be defined as:

w = y a x + k a 2 + m = T A l / r + ), A ~ / r + m

where a = Ta 1 + Xa 2 . Res iden ts in the home coun t ry hold domes t i c equi t ies (al) and foreign equi t ies (a2). y is the p ropor t ion of domes t ic equi t ies held b y the p r iva t e sector. Open-marke t opera t ions are a s sumed to be conducted in t e rms of domest ic equit ies, y is equa l to 1 for the case where money is only of the outs ide type. In an exclus ively inside money economy w = a I + ). a 2 since m = (l - - T) a l . ~. s t ands for the real exchange r a t e which will be defined la te r on. The symbol A signifies the p e r m a n e n t income s t r eam from equit ies. The wea l th com- pos i t ion of the foreign coun t ry

w* = T* a* + a~/), + m* = y* A~/r + A~/Xr + m*

should be in te rp re ted in ana logy to the composi t ion of w.

The real va lue of wea l th (w) can change for var ious reasons. There are quan t i ty - induced wea l th effects: a h igher s tock of p r i v a t e l y held equi t ies v ia domest ic i nves tmen t affecting A1,

26 E m i l - M a r i a C l a a s s e n

Real cash balances have been introduced in the saving behaviour following the wealth-saving relationship of Metzler [1951 ]. According to the Patinkin [1965] tradition, their introduction guarantees the deter- nfinacy of the general price level, x denotes the real exchange rate or the terms of trade defined as

X = e P*/P (Io)

i.e. as the relation between the foreign prices (P*) of imported goods expressed in home currency and the domestic currency prices of home goods (P); e stands for the nominal exchange rate. Both the real and nominal exchange rates were identical in the context of the Keynesian approach because the price levels were assumed to be constant 1.

The solution of the equilibrium system can be conceived in ternls of three analytical stages. The first two concern the determination of the real variables r, m, m* and ~.

(i) As in the Keynesian case, the interaction of the world market for goods [(6) + (7)] and of the world market for real cash balances [(8) + (9)] determines the world interest rate and the world stock of real cash balances (m plus m*).

(if) Under the hypothesis that the national goods markets still remain in disequilibrium, an adequate change in the real exchange rate will provide their equilibrium [equation (6) on (7)]. As in the Keynesian framework, an excess supply in the domestic goods market (S-- I > T) which is necessarily equal to an excess demand in the foreign goods market (I* - - S* > T) will be eliminated by the depreciation of the home currency in real terms, i.e. by a rise in x. *

a higher stock of foreign equities held by domestic residents arising from a current-account surplus and increasing A~, and a higher volume of outside money which does not affect the coefficient T. More subtle are the price-induced wealth effects, i.e. the interest-rate-induced wealth effect and the wealth effect induced by the real exchange rate. Macroeconomic analysis always becomes cumbersome when the value-induced wealth effects are taken into account. In order to simplify the exposition, we shall ignore them because the long-rtm equilibrium values are not necessarily significantly affected by them. The price-induced wealth effects

are treated in Claassen [x983].

x We follow the traditional terminology with respect to the nominal and real exchange rate as i t has been established in the literature over the last years. However, i t would be more accurate to express e as the exchange rate which is the product of a real component

(),) and of a nominal component (P/P*).

t r, m, m* and ), are detexmined simultaneously by equations (6)--{9) if one takes into

aceouat explicitly the wealth definitions described in footnote 1, p. 25.

D e t e r m i n a t i o n of the E x c h a n g e R a t e 27

(iii) The model is closed with the monetary sector by the derivation of the price levels for a given domestic and foreign nominal money supply

P = M/m ( I I )

P* = M*/m* (12)

and by the nominal value of the exchange rate according to the relationship (IO) between the domestic and foreign price level for a given equilibrium value of the real exchange rate.

In Figure 3 the domestic market for real cash balances is illustrated geometrically by the LM schedule, and the domestic goods market by the IS schedule. I t is assumed that the equilibrium value of the world interest rate is r o. At the interest rate to, nlo real cash balances are demanded which will be supplied by a movement of the general price level for whatever nominal value the money supply may have. The IS schedule has a negative slope (equal to Sw/Tx), provided that the Marshall-Lerner condition holds (T x > 0). While the domestic money market is equilibrated

Figure 3

LM r .~Ap(RO )

i [" 'x.,,._ ~ - ~ ' ~ ~1

t , : [

I', i

\ ' ''B J , /" . . . . . . . v ~ , - - - ~ . . . . . . . . . . . . . r . . . . . ; ~ . . . . . . . . . . . . . . . . . . . . . . . .

t i / J , L

r% m 2 m

//t,5" / / /

28 E m i l - M a r i a C l aa s se n

by the domestic price level, any disequilibrium in the domestic goods market is eliminated by a change in the real exchange rate. Thus, for instance, an excess supply of goods at point U involves a depreciation towards point A which increases total demand for the products of the domestic economy 1.

The initial equilibrium is established at point A. We assume that Po = P* = 1 such tha t Xo = e o and we assume xo = eo = l . The interest- rate pari ty of equation (5) holds also for the classical model, but it has to be reformulated explicitly in terms of the real exchange rate since the interest-rate pari ty refers to the real interest rarest:

r = r * - 9 0 ( ~ - - X ) / x 0 < 0 ~< 1 (13)

i, is the real exchange rate which is expected to reign in the long run. The real-interest rate pari ty is represented in the upper right-hand panel and its interpretation is analogous to tha t of Figure I.

As in the Keynesian model, we shall examine the exchange-rate effects of an expansionary monetary policy of the domestic country. We consider the impact effect, the intermediate effect and the long-run effect on the exchange rate first within a one-country model and then within a two- country model.

a Dorabusch's quantity-theoretical two-country model [x976 ] is similar to ours with respect to the saving function. Assuming investment in each country equal to zero, equilibrium in the world goods market and in the national goods markets is realized without any change in the real exchange rate: S (r, w) = - - S* (r, w*) = T. This is a possible case to the extent that one uses a one-commodity model for the world economy. C.onsequeatly, our model assumes that each country produces a differentiated good such that there is an endogenous relative price between domestic and foreign products.

Another possibility is that in both countries (homogeneous) tradable goods and non- tradable goods exist. An excess supply of goods in the home country (equal to an excess demand for goods in the foreign country because of the equilibrium in the world goods market) does not need to cover exclusively the category of tradable goods - - homogeneous or differentiated ones - - such that a change in the relative price of tradable and non-tradable goods has to take place in both countries. Under this aspect, ~, would represent the relative price between tradable goods (T) and non-tradable goods (N) in the two countries respectively

x = (P~/P~)ItP~IPT~.

t We abstract from continuous increases in the general price level mad, by this, from the inflationary phenomenun~ However, a once-and-for-all increa~ in the price level, to the extent that it is expected to take place during the period under consideration, will affect the nominal interest rate of that period whereas in all future periods, the nominal interest rate will be identical to the real interest rate. For that reason, we have not included the nominal interest rate in the demand functions for money, since they refer to the long run. For the description of the short-run adjustment mechanism, the transitional phenomenon of a higher nominal interest rate could be taken into account in the functions of the demand for money, complicating the adjustment process a bit more.

Determination of the Exchange Rate 29

I. O n e - C o u n t r y Model

Suppose that the expansion of the domestic quantity of money is of the outside money type. The increase in the money supply is assumed to be equal to mo ms. In the long run, even in a two-country model, the interest rate would not be affected, such that the domestic price level would increase ill proportion to the quantity of money. Consequently, the change in the long-run nominal exchange rate will be indicated fully by the purchasing power parity while the long-run real exchange rate remains unchanged (xo = xo).

The impact effect of the monetary expansion on the domestic interest rate (rx) and on the real exchange rate (xl) is indicated by point B. Since the financial sector of the economy is assumed to react first to any disequilibrium, there is a fall in the interest rate creating an interest-rate differential which has to be equal to an anticipated rate of appreciation which is brought about by the depreciation towards e 1. To the extent that, at the very beginning, the domestic price level has still not moved, this depreciation is identical to a depreciation of the real exchange rate (Xx). At point B in the lower left-hand panel, there is an excess demand for goods because of the lower interest rate and of the higher real exchange rate, both factors pushing up the domestic price level to Pv The adjustment path towards the long-run equilibrium position is that of BA in the upper four panels and that of BC in the bottom panel.

We shall now assume an expansion of the domestic quantity of money of the inside type in terms of open-market operations: one part of equities (a) held by the private sector is exchanged against the newly created money. The initial composition of wealth (w) is modified without changing total wealth, such that the IS schedule of Figure 3 shifts to the position IS 1 (ro) by the amount of the additional money mo my This shift takes place under the assumption of a one-country model where the "small country" is not able to influence the world interest rate r o .

Due to the fact that the long-run equilibrium in the money market must again be at point A in the upper four panels, the long-run nominal exchange rate wiLl again rise to e 2 and the long-run real exchange rate will remain at X o. The impact effect on the domestic interest rate and on the exchange rate is indicated again by point B. After the impact effect, the domestic price level begins to rise inducing the same adjustment path from B to A in the upper four panels and from B to C in the bottom panel.

3 ~ Emi l -Mar i a C laassen

2. T w o - C o u n t r y Model

If the increase of the money supply takes place in terms of outside money, a two-country model does not add any new information. Since, in the long run, the interest rate remains unchanged, the adjustment process of the nominal and real exchange rate remains that of the one- country model.

However, in the case of inside money (open-market operations), over the long run there will be a decline of the world interest rate (even if the domestic country is a small one). The fall in the interest rate raises the domestic as well as the foreign demand for real cash balances. Consequently, the increase in the domestic price level is lower than in the one-country model, whereas there is also a fall in the foreign price level in order to satisfy the foreign excess demand for real cash balances. In the long run, the nominal exchange rate depreciates by the amount of the rise in the domestic price level and the fall in the foreign price level. Furthermore, if the interest elasticity of the foreign demand for money is higher than the one of the home country, the depreciation of the nominal exchange rate will exceed the rate of increase of the domestic money supply.

As far as the evolution of the real exchange rate is concerned, we come to the same conclusion as in the case of the one-country model. Due to the assumption that asset markets react more quickly to a disequilibrium than the goods markets, the impact effect on the exchange rate is the effect on the real exchange rate according to the interest-rate pari ty where the depreciation of the home currency overshoots the long- run real exchange rate. From the one-country model we know that a subsequent appreciation will take place accompanied by a rising domestic price level and a rising domestic interest rate. In the case of a two-coun- t ry model, after the impact effect has taken place, the foreign interest rate begins to decline which produces a lower interest-rate differential and, thus, a lower expected rate of appreciation of the home currency. This lower expected rate is induced by a subsequent higher (or quicker) appreciation than the one which is derived in a one-country model.

During the adjustment process towards long-run equilibrium, the real exchange rate exceeds its long-run value creating a series of current- account surpluses in the domestic country. The acquisition of additional foreign assets serves to compensate the loss in wealth due to the rise in the domestic price level. Similarly, in the foreign country, the increase in wealth as a consequence of the fall in the foreign price level gives rise to a corresponding cumulative foreign current-account deficit. When the initial wealth level has been attained in both countries, there will

3 2 E m i l - M a r i a C l aa s sen

already shifted to the left because of the lower interest rate) con "tmues to shift gradually towards point D 1.

IV. s Remarks

Under perfect capital mobility and under long-run equilibrium con- siderations, the interaction of the world markets for goods and money determines the interest rate within a Keynesian as well as a quantity- theoretical framework. The difference between both models lies in the role of the national money markets. For the Keynesian approach which assumes, in its simple version, the general price level as given, they fix national incomes, and for the quantity theory operating under the hypothesis of full employment, they determine the national price levels. Both approaches, when they are transposed into the context of the international economy, have one essential feature in common: the national goods markets indicate the equilibrium value of the real exchange rate; in addition, in the classical model, once the real exchange rate is known, the nominal exchange rate results from the relationship of the national price levels.

The present article also stated the effect of an expansionary monetary policy on the real exchange rate and on real income in the Keynesian framework and on the real and nominal exchange rate and the price level in the classical framework, both within a one-country and two-country model.

As far as the long-run effects within the context of a two-country model are concerned, there is a fall in the world interest rate, a rise in domestic income, a fall in foreign income and an increase in the (real identical to the nominal) exchange rate within the Keynesian model. The latter increase is necessary in order to eliminate the excess supply in the domestic goods market and the corresponding excess demand in the foreign goods market. For the classical approach, one has to differentiate between outside and inside money operations because only the latter ones affect the interest rate.

An expansionary monetary policy of the outside money type produces a corresponding increase in the domestic price level and, by this, a rise in the nominal exchange rate leaving all real variables unchanged.

- - W h e n the monetary expansion is conducted with inside money, there is a fall in the world interest rate, and the equilibrium value

z However, the country will experience a slight appreciation of the real exchange rate because it receives now more interest payments from abroad. Implicitly we assume that, at the be~nnln~7 the home country is a net creditor. Consequently, the trade balance deficit must be larger, which is brought about by an appreciation.

3 2 E m i l - M a r i a C l aa s se n

already shifted to the left because of the lower interest rate) con "tmues to shift gradually towards point D 1.

IV. s Remarks

Under perfect capital mobility and under long-run equilibrium con- siderations, the interaction of the world markets for goods and money determines the interest rate within a Keynesian as well as a quantity- theoretical framework. The difference between both models lies in the role of the national money markets. For the Keynesian approach which assumes, in its simple version, the general price level as given, they fix national incomes, and for the quantity theory operating under the hypothesis of full employment, they determine the national price levels. Both approaches, when they are transposed into the context of the international economy, have one essential feature in common: the national goods markets indicate the equilibrium value of the real exchange rate; in addition, in the classical model, once the real exchange rate is known, the nominal exchange rate results from the relationship of the national price levels.

The present article also stated the effect of an expansionary monetary policy on the real exchange rate and on real income in the Keynesian framework and on the real and nominal exchange rate and the price level in the classical framework, both within a one-country and two-country model.

As far as the long-run effects within the context of a two-country model are concerned, there is a fall in the world interest rate, a rise in domestic income, a fall in foreign income and an increase in the (real identical to the nominal) exchange rate within the Keynesian model. The latter increase is necessary in order to eliminate the excess supply in the domestic goods market and the corresponding excess demand in the foreign goods market. For the classical approach, one has to differentiate between outside and inside money operations because only the latter ones affect the interest rate.

An expansionary monetary policy of the outside money type produces a corresponding increase in the domestic price level and, by this, a rise in the nominal exchange rate leaving all real variables unchanged.

- - W h e n the monetary expansion is conducted with inside money, there is a fall in the world interest rate, and the equilibrium value

z However, the country will experience a slight appreciation of the real exchange rate because it receives now more interest payments from abroad. Implicitly we assume that, at the be~nnln~7 the home country is a net creditor. Consequently, the trade balance deficit must be larger, which is brought about by an appreciation.


of domestic and foreign real cash balances increases, involving a relatively smaller rise in the domestic price level and a decrease in the foreign price level. The nominal exchange rate adjusts in relation to the new national price levels. The real exchange rate, at least in the long run, remains unchanged because an excess supply in the domestic goods market and the corresponding excess demand in the foreign goods market will be eliminated, at the initial real exchange rate, when the former wealth level has been reconstituted by a cumulative current account surplus in the home country and by a corresponding cumulative current account deficit in the foreign country.

To the extent tha t one leaves the world of a Keynesian fixed-price model and of a classical fixed-quantify model and that one regards a situation of a flexible quanti ty and price model, the long-run effects of an expansionary monetary policy must be a mixture of both extreme cases: a rise in the domestic price and output levels, a fall in the foreign price and output levels and an increase in the nominal and real exchange rate. However, this model still has to be constructed.

References

Claassen, Emil-Maria, "The Nominal and Real Exchange Rate in a Quant i ty Theo- retical Two-Country Model". In : E.-M. Claa.ssen and P. Salin (Eds.), Recent Issues in the Theory of Flexible Exchange Rates. 5th Paris-Dauphine Conference on Money and In ternat ional Monetary Problems, i5 . - -x 7. 6. i981. Studies in Monetary Economics, Vol. 8, Amsterdam 1983, pp. 57--68.

Dornbuseh, Rudlger, "Capital Mobility, Flexible Exchange Rates and Macroeconomic Equi l ibr ium". In : E.-M. Claassen and P. Salin (Eds.), Recent Issues in Inter- national Monetary Economics. 3rd Paris-Dauphine Conference on Money and In ternat ional Monetary Problems, 28.--30. 3. 1974. Studies in Monetary Economics, Vol. 2, Amsterdam I976, pp. 261--278.

- - , Open Economy Maevoeconomics. New York 198o.

- - , and Paul Krugmma, "Flexible Exchange Rates in the Short Run" . Brookings Papers on Economic Activity, i976 , No. 3, PP. 537--584 �9

Kouri, Pentti J . K., "The Exchange Rate and the Balance of Payments in the Short Run and in the Long Run: A Monetary Approach". The Scandinavian Journal o] Economic, s, Vol. 78, 1976, pp. 28o--3o 4.

bletzler, Lloyd A., "'Wealth, Saving, and the Rate of In teres t" . The Journal o/ Political Economy, Vol. 59, I95 r, PP. 93 - - I I6 .

Mtmdeal, Robert A., International Economics. New York 1968.

Patiuldn, Don, Money, Interest, and Prices: An Integration o] Monetary and Value Theory. 2nd ed., New York I965.

S

Wdtwirt~affliahes &rc~iv Bd. CXIX. 3

34 Emil-Maria Claassen

Z u s a m m e n f a s s u n g : Die Bestimmung des Wechselkurses im keynesianischen und klassischen Model. w Im Modell des langtristigen Gleichgewichts wird bei voll- kommener Kapitalmobilit~t der Zinssatz dutch das Zusammenwirken der Weltn~rkte ffir Gfiter und Kapital bcstimmt, und zwar sowohl nach Keynes als auch nach der Quantit~tstheorie. Beide Ans~tze haben im Hinblick auf die internationalen Zu- sammenh~inge eines gemeinsam: Die nationalen Gliterm~rkte geben den Gleich- gewichtswert des realen Wechselkurses an. Dieser Aufsatz unterstreicht auch die Wirkung einer expansiven Geldpolitik auf den realen Wechselkurs und das Real- einkommen im Rahmen der keynesianischen Theorie und auf den rcalen und nomi- nalen Wechselkurs sowie das Preisniveau im klassischen Modell, und zwar sowohl bei nut einem Land als auch im Zwei-L/inder-Fall. Was die langfristigen Wirkungen be- trifft, so ergibt sich im Zwei-Lgmder-Fall des keynesianischen Modells eine Senkung des Zinssatzes auf dem Weltmarkt, eine Zunahme des heimischen und eine Ab- nahme des ausl~mdischen Einkommens sowie ein Anstieg des Wechselkurses, wobei der reale mit dem nomiualen identisch ist. Beim klassischen Ansatz mul3 man zwi- schen AuBengeld- und Innengeld-Operationen unterscheiden, weil nur letztere den Zinssatz beeinflussen. Eine expansive Geldpolitik gleich welcher Art l~tlt den realen Wechselkurs unver~ndert, jedenfalls aut lange Sicht.

R d s u m d : La ddtermination keyn6sienne et classique du taux de change. - - Donnde la mobilit~ parfaite du capital et l'dquilibre ~ long terme l ' interaction des marchds mondiaux des biens et de la monnaie ddtermine le taux d'int6r6t en cadre keyndsien aussi bien qu'en cadre de la th~orie quantitative. Les deux approches ont, s'ils sont transposds dans le contexte de rdconomie internationale, un aspect essentiel en commun: les marchds des biens nationaux indiquent la valeur d'dquilibre du taux de change r6el. De plus, cet article ddmontre l'effet d 'une politique mondtaire expansionniste sur le taux de change r~el et sur le revenu r~%l en cadre keyn~sien, et sur le taux r6el et nominal aussi bien que sur le niveau de prix en cadre classique, dans un module ~ seulement un pays et ~ deux pays. Regardant les effets ~ long terme dans le contexte d 'uu module k deux pays du type keyndsien, le taux d'int&r6t mondial et le revenu dtranger baissent, le revenu int6rieur et le taux de change (r6el identique au taux nominal) montent. Au cas de l 'approche classique il faut diffdrencier entre des oi~rations du ,outside, et *inside money*, parce qu'elles ne sont que les derni~res qui affectent le taux d'int~r~t. Une politique mondtaire expansionniste de deux types n'influence pas le taux de change rfiel, au moins ~ long terme.

R e s u m e n : La determinaci6n keynesiana y cl~sica de la tasa de cambio. - - Bajo movilidad perfecta del capital y bajo consideraciones de equilibrio de largo plazo, la interacci6n de los mercados mundiales de bienes y dinero determina la tasa de inter6s tanto dentro de un marco keynesiano como de uno te6rico-cuantitativo. Ambos planteamientos, cuando son transpuestos a un contexto de la economla internacional, tienen una caracteristica esencial en comdn: los mercados de bienes nacionales indican el valor de equilibrio de la tasa de cambio real. En este artlculo tambidn se establece el efecto de una poHtica monetaria expansiva sobre la tasa de cambio real y el ingreso real en el marco keynesiano, y sobre la tasa de cambio real


y n o m i n a l y el n ivel de precios en el marco cl~sico, a m b o s den t ro de u n mode lo de uno y de dos palses. E n c u a n t o a los efectos de largo plazo se refiere, den t ro del con tex to de u n mode lo keynes i ano de dos parses h a y u n a ca ida en la t a s a de inter6s m u n d i a l y el ingreso ex te rno , y u n a lza en el ingreso in fe rno y la t a s a de cambio (real id6n t ica a la nomina l ) . P a r a el p l a n t e a m i e n t o cl~sico se debe d i ferenciar en t r e operaciones m o n e t a r i a s e x t e r n a s e in te rnas , po rque s61o las f i l t imas a fec t an la t a s a de inter6s. U n a pol i t ica m o n e t a r i a e x p a n s i v a de a m b o s t ipos de ja la t a s a de cambio real in- var iable , por lo m e n o s en el largo plazo.

3*

the keynesian and classical determination of the exchange rate

Documents