I. Introduction

VICENTE GALBIS ^*

Keynesian theory is basically a theory of the short-run determination of income and the interest rate in a closed economy (Hicks, 1937). The money supply is assumed to be exogenously given and is the monetary policy instrument. How this monetary instrument should be used is still a matter of dispute, with some advocates assigning to it a major role in dealing with cyclical fluctuations said to be inherent in the free enterprise economy (Ritter, 1963), while others blame its inappropriate manipulation for many of the cyclical fluctuations that are actually observed (Friedman, 1960). It is generally agreed, however, that the effects of monetary policy on income are powerful.¹ Friedman and others would add the proviso that these effects are also highly unpredictable because they tend to work with a long and variable lag.

The effects of monetary policy actions on income are far more complicated and uncertain in an open economy. For one thing, in such an economy the policymakers are faced with the additional task of stabilizing the balance of payments and/or ensuring that movements in the foreign exchange rate are not detrimental to the stability of income. Mundell (1962) addressed himself to this question, deriving the classic proposition that monetary policy should be better addressed to problems of external stability while fiscal policy should play a larger role in domestic stabilization (see also Fleming, 1962). Thus he showed that monetary policy lacked independence as an instrument of income stabilization in an open economy.

Subsequent developments on the theory of stabilization policies in an open economy distinguished sharply between such policies under fixed and floating exchange rate systems (Mundell, 1968; Sohmen, 1969).² By means of a simple extension of Keynesian analysis allowing for international capital movements, these contributions further deflated the importance of monetary policy as an instrument of income stabilization in an open economy with a fixed exchange rate, but they also showed that this policy could regain full leverage under a floating exchange rate system. An opposite set of conclusions was found with regard to fiscal policy.

Other theoretical developments have further demonstrated that even the concept of monetary policy in an open economy is not well defined, since it is not generally possible to control the money supply. Thus net credit of the central bank, not money supply, is a more true instrument of control (Guitián, 1973). But even the credit instrument is hampered as a means of stabilizing income under a fixed exchange rate; its main or exclusive effect is on capital flows and the balance of payments.

This paper has a twofold purpose. First, it presents succinctly the basic theory of monetary and exchange rate policies in the context of a small open economy under both fixed and flexible exchange rates; this theory has been developed by Mundell (1962; 1968) and others (see Whitman, 1970). An open economy is defined here as one that carries on trade in goods and services with other countries and, more importantly, participates in international capital movements. Furthermore, the simplifying assumption is made, representing a strictly polar case, that capital mobility is perfect in the absence of direct government controls.³ The economy is small in the sense that its influence on incomes and interest rates in other countries is negligible.⁴ Second, this paper attempts to expound further on some neglected aspects of the theory of monetary and exchange rate policies in a small open economy under a floating exchange rate system and on the effects of capital controls under both fixed and flexible exchange rate systems.⁵

The assumptions are framed in terms of a simple Keynesian-type model which can be adapted easily to the operations of the external sector, whether under a fixed or floating exchange rate system and with or without capital controls (Section II). The assumptions are kept as simple as possible for the macroeconomic conditions not directly related to the characteristics of the small open economy, as distinct from a closed economy, without sacrificing the introduction of all relevant constraints and policy instruments—especially with regard to monetary policy specified in terms of the net domestic credit policy of the central bank and the reserve requirement ratio of the banks. Thus, the theoretical results are slightly more general than those previously derived in the literature. The principal limitations of the model are that it is based on a flow theory, rather than on a stock adjustment theory, of capital movements; that it does not contain a government budget constraint, so that no restrictions or interconnections between fiscal and monetary policies are recognized; and finally, that it implicitly assumes that prices and wage rates are sticky. However, this last assumption is of no immediate relevance to the discussion because it is phrased in terms of movements in nominal income rather than in terms of real income and employment.

Section III deals with a fixed exchange rate situation and shows that the theory and the conclusions with regard to the effects of monetary policy are rather different from those that have been applied to a relatively large and closed economy such as that of the United States. Thus, inferences drawn from the literature based primarily on the U.S. experience have been frequently and erroneously applied to the small open economy such as that of Canada. Section IV adjusts the model to a floating exchange rate situation and asserts that monetary policy regains much of its freedom to cope with domestic income problems. However, it is also shown that freeing the exchange rate cannot automatically solve the dilemma of the monetary authorities on whether to address themselves to domestic or foreign considerations, because actions taken with a view as to domestic considerations also induce developments in the foreign exchange, capital flows, and external real sector. The “dirty float” case, in which the float is not entirely accompanied by a policy of nonintervention in the foreign exchange market, is also considered. The model is subsequently modified to explore the consequences of applying foreign capital controls under both fixed and floating exchange rate systems (Section V). Finally, Section VI summarizes the main conclusions of the paper and focuses on some of the major qualifications for their empirical validity.

II. Structure of a Small Open Economy

Equations (1) through (9) are assumed to describe succinctly the general structure of a small open economy. They are specified in sequence. A brief discussion is then followed by a summary and restatement of the basic structure so as to highlight the point that all cases covered in this study can be derived from this general structure by appropriate choice of additional restrictions on the equations and/or the variables.

where

Af	=	net foreign assets of the central bank
$\begin{array}{c}{A}_{-1}^f\end{array}$	=	net foreign assets of the central bank, lagged one period
B	=	current account balance of the international balance of payments, a function of national income, foreign income, and the foreign exchange rate
C	=	private consumption expenditure, a function of disposable income
CD	=	ratio of currency to deposits held by the private nonbank sector with the banks
e	=	foreign exchange rate defined as the number of units of foreign currency per unit of national currency, so that an increase (decrease) in e means an appreciation (depreciation) of the national currency in terms of the foreign currency
G	=	total government expenditures
H	=	high-powered money or monetary base
I	=	private investment, a function of the interest rate
K	=	net international capital flows
M	=	money stock
Md	=	money demand
Ms	=	money supply
MULT	=	money multiplier
NDA	=	net domestic credit provided by the central bank
r	=	domestic interest rate
rf	=	foreign interest rate
RE	=	ratio of excess reserves of the banks to their deposit liabilities
RQ	=	ratio of required reserves of the banks to their deposit liabilities
T	=	taxes
Y	=	national income
Yf	=	foreign income

View Table

Af	=	net foreign assets of the central bank
$\begin{array}{c}{A}_{-1}^f\end{array}$	=	net foreign assets of the central bank, lagged one period
B	=	current account balance of the international balance of payments, a function of national income, foreign income, and the foreign exchange rate
C	=	private consumption expenditure, a function of disposable income
CD	=	ratio of currency to deposits held by the private nonbank sector with the banks
e	=	foreign exchange rate defined as the number of units of foreign currency per unit of national currency, so that an increase (decrease) in e means an appreciation (depreciation) of the national currency in terms of the foreign currency
G	=	total government expenditures
H	=	high-powered money or monetary base
I	=	private investment, a function of the interest rate
K	=	net international capital flows
M	=	money stock
Md	=	money demand
Ms	=	money supply
MULT	=	money multiplier
NDA	=	net domestic credit provided by the central bank
r	=	domestic interest rate
rf	=	foreign interest rate
RE	=	ratio of excess reserves of the banks to their deposit liabilities
RQ	=	ratio of required reserves of the banks to their deposit liabilities
T	=	taxes
Y	=	national income
Yf	=	foreign income

and m, g, and h are used to represent functions, of which the partial derivatives are denoted by the respective symbols followed by subscripts 1, 2, and so on (for instance, B₁ means δB/δY).

Equation (1) represents in summary form a simple Keynesian process of national income determination. Income is broken down by expenditure components into private consumption, which is a positive function of disposable income (income minus taxes); private investment, which is a negative function of the interest rate; exogenously given government expenditures; and the foreign current account balance, which is a negative function of both national income and the level of the exchange rate and a positive function of foreign income.

Equation (2) is a typical Keynesian demand for money function; the demand for money is positively related to income and negatively related to the interest rate. Equation (3) defines the supply of money as the product of the monetary base or high-powered money and the money multiplier. Equation (4) is an identity which breaks down the monetary base into domestic credit and foreign asset components. Equation (5) summarizes the process of determination of the net foreign assets component of the monetary base in any given period; this is determined by foreign assets in the previous period plus the net surplus or deficit in the current account balance, which is endogenously determined, as explained earlier, plus the net capital balance, which is also endogenous to the model in the absence of capital controls. In equation (6) the money multiplier is expressed in terms of the currency-deposit ratio of the banks and their required reserve ratio.⁶ Equations (7) and (8) express, respectively, the currency-deposit and the excess reserve-deposit ratios as negative functions of the interest rate, which is taken as the relevant measure of the opportunity cost of holding currency and reserves. Finally, equation (9) is an equilibrium condition of the model equating the money supply and the demand for money.

Classification of the variables as endogenous or exogenous and then as policy or nonpolicy variables within the exogenous group requires specification of further conditions. However, since there are several suitable sets of alternatives leading to the variety of situations studied in this paper, a unique classification of all the variables is inappropriate, although some play a fixed role: G and T are fiscal policy instruments; NDA and RQ are monetary policy instruments; Y^f and r^f are nonpolicy exogenous variables (by the small economy assumption). Other variables, such as r, K, A^f, and e, are susceptible to direct manipulation as policy instruments but may also be endogenous to the system. Finally, Y and M must be viewed as essentially endogenous.

The general structure of the small open economy may be summarized in these four equations:

All the cases covered in this paper can readily be derived from this general structure by appropriately selecting four of the variables as the unknowns under a given situation and analyzing the reaction of these unknowns to the policy variables available under that situation. The cases studied are:

Case 1. Perfect capital mobility (Sections III and IV). The domestic interest rate is equal to, and determined by, the foreign interest rate.⁷

Case 1a. Fixed exchange rate case (Section III). The exchange rate is fixed and therefore exogenous as a result of direct intervention by the authorities in the foreign exchange market. The model determines Y, M, K, and A^f, given the exogenous variables r, Y^f and the policy variables e, G, T, NDA, and RQ.

Case 1b. Floating exchange rate case (Section IV). Absence of intervention in the foreign exchange market makes e endogenous, but A^f becomes constant and therefore exogenous. The model determines Y, M, K, and e, given the exogenous variables r, Y^f, A^f and the policy variables G, T, NDA, and RQ.

Case 1b–1. Floating with a view as to the appropriate exchange rate (Section IV). Absence of intervention in the foreign exchange market (A^f is constant), coupled with the desire to have a constant exchange rate, calls for fiscal and monetary policies to be coordinated in a unique way if a desired income goal is to be achieved. The model for this case is not different from that of the freely floating exchange rate. Simply, the analysis is reversed so that, instead of studying the effects of instruments on targets, two targets are now specified (income maintenance and exchange rate constancy) and appropriate levels of the instruments are derived.

Case 1b–2. Dirty float (Section IV). The authorities make use of limited intervention in the foreign exchange market to help break “excessive” movements in the foreign exchange rate while letting the rate drift in any direction in the long run. The model determines Y, M, K, e, and A^f, given the exogenous variables r, Y^f and the policy variables G, T, NDA, and RQ. It should be noted that five rather than four endogenous variables are determined in this case as a result of the specification of an independent intervention policy reaction function on the part of the monetary authorities, relating movements of A^f to e. This is therefore a more general case and is perhaps more realistic.

Case 2. Capital controls (Section V). Introduction of capital controls makes K an exogenous variable and forces r to be endogenous.

Case 2a. Fixed exchange rate (Section V). The model determines Y, M, r, and A^f, given the exogenous variable Y^f and the policy variables K, e, G, T, NDA, and RQ.

Case 2b. Floating exchange rate (Section V). The model determines Y, M, r, and e, given the exogenous variable Y^f and the policy variables A^f, K, G, T, NDA, and RQ.

III. The Model with a Fixed Exchange Rate

As indicated earlier, the assumption in this paper of an open economy means that it carries on trade in goods and services with other countries and that its capital market is perfectly integrated with the world capital market in the absence of government controls. This assumption implies that arbitrage operations in the international capital market prevent the domestic interest rate from departing systematically from that of the world capital market.⁸ Furthermore, the assumption of a small economy means that the world interest rate is exogenous or, in other words, that the domestic interest rate adjusts and is equal to the foreign interest rate.

The model in equations (1) through (9), together with condition (14) and the further assumption that the economy operates under a fixed exchange rate system,¹⁰ simultaneously determines the interest rate, national income, the balance of payments, and the stock of money, given the values of the policy and nonpolicy exogenous variables. Superficially, this appears to be a mere extension of the Keynesian model of the simultaneous determination of the interest rate and income. However, as the analysis below reveals, in a small open economy the determination of the money stock, income, and the rate of interest is strikingly different from that in a closed economy. It will be shown that monetary policy is rendered ineffective in controlling the interest rate, income, and even the money stock. Monetary policy can show a powerful effect, however, in determining capital flows and, therefore, the balance of payments.

Basically, the working of the model is as follows. The domestic interest rate is determined by the foreign interest rate as a result of direct arbitrage operations in the integrated capital market (equation (14)). Given the interest rate, the exchange rate, and the level of foreign income—together with exogenously determined government expenditures and taxes—the level of income, private consumption, and the current account balance of the international balance of payments are determined (equation (1)). It follows, therefore, that income determination is independent of monetary policy; this is a consequence both of the openness of the economy (which causes the domestic interest rate to be determined by the foreign interest rate rather than the domestic money supply and the demand for money, as would be the case in a closed economy) and of the fixed exchange rate regime. Furthermore, given income and the interest rate, the effective demand for money is also determined independently of monetary policy (equation (2)). Finally, given equilibrium condition (9), it also follows that the effective money supply is determined by, and adjusts to, the demand for money. The nature of this adjustment is implicit in the interest rate equalization, condition, which requires capital to flow freely so as to equate the money supply and the demand for money at the given interest rate.

The endogenous determination of the net amount of capital flows needs to be spelled out in some detail, as this alone bears the brunt of the adjustment of the supply to the effective demand for money. Such adjustment cannot take place through endogenous changes in the money multiplier, since this is fixed for any given level of the interest rate and the required reserve ratio. Neither can the adjustment take place through movements in the current balance, which are independent of money supply.

Substitution of equations (4) and (5) in equation (3) and the use of the equilibrium condition (9) yield

M = (NDA + A^f_–1 + B + K) MULT

which can be solved for K as

K = M/MULT – A^f_–1 – B – NDA

and expressed in terms of its underlying determinants in equations (6) through (8) as follows:

Equations (1), (2), and (15), with conditions (9) and (14), constitute the reduced form of the model and can be used to trace the effects of monetary, fiscal, and exchange rate policies on the economic targets (see Table 1). As already noted, neither of the two monetary policy instruments (NDA nor RQ) has any effect on the stock of money, the interest rate, or income; their only effect is on the net capital flows. Thus a tightening of monetary policy either by a reduction in the net domestic credit of the central bank or by an increase in the required reserve ratio has the effect of increasing the volume of net capital inflows (or decreasing the volume of net capital outflows). Since the current balance is unaffected, net foreign assets of the central bank will increase.

Table 1.

Effects of Policy Instruments in a Small Open Economy with a Fixed Exchange Rate ¹

¹See text for explanation of symbols. The expressions in this table represent the total derivatives of each endogenous variable with respect to each policy variable, holding the other policy variables constant. For example, the expression in column 1, row 3 represents the effect of e on Y when NDA, RQ, G, and T are held constant; its negative sign indicates that currency revaluation reduces income.

Table 1.

Effects of Policy Instruments in a Small Open Economy with a Fixed Exchange Rate ¹

Policy Instruments		Policy Targets
		Y	M	K	B	Af
Monetary policy
	NDA	0	0	–1	0	–1
	RQ	0	0	M/(1 – g(r)) > 0	0	M/(1 – g(r)) > 0
Exchange rate
	e	$\frac{B_3}{\left(1-c_1-B_1\right)}<0$	$\frac{m_1{B}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1+\left(1-C_1\right)\right\}B3}{\left(1-C_1-B_1\right)}<0$	$\frac{B_3\left(1-C_1\right)}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1+2\left(1-C_1\right)\right\}B3}{\left(1-C_1-B_1\right)}<0$
Fiscal policy
	G	$\frac{1}{\left(1-C_1-B_1\right)}>0$	$\frac{m_1}{\left(1-C_1-B_1\right)}>0$	$\frac{\left(1/\mathit{MULT}\right)m_1-B_1}{\left(1-C_1-B_1\right)}>0$	$\frac{B_1}{\left(1-C_1-B_1\right)}<0$	$\frac{\left(1/\mathit{MULT}\right)m_1}{\left(1-C_1-B_1\right)}>0$
	T	$\frac{C_2}{\left(1-C_1-B_1\right)}<0$	$\frac{m_1{C}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1-B_1\right\}{C}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{B_1{C}_2}{\left(1-C_1-B_1\right)}>0$	$\frac{\left(1/\mathit{MULT}\right)m_1{C}_2}{\left(1-C_1-B_1\right)}<0$

¹See text for explanation of symbols. The expressions in this table represent the total derivatives of each endogenous variable with respect to each policy variable, holding the other policy variables constant. For example, the expression in column 1, row 3 represents the effect of e on Y when NDA, RQ, G, and T are held constant; its negative sign indicates that currency revaluation reduces income.

View Table

Table 1.

Effects of Policy Instruments in a Small Open Economy with a Fixed Exchange Rate ¹

Policy Instruments		Policy Targets
		Y	M	K	B	Af
Monetary policy
	NDA	0	0	–1	0	–1
	RQ	0	0	M/(1 – g(r)) > 0	0	M/(1 – g(r)) > 0
Exchange rate
	e	$\frac{B_3}{\left(1-c_1-B_1\right)}<0$	$\frac{m_1{B}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1+\left(1-C_1\right)\right\}B3}{\left(1-C_1-B_1\right)}<0$	$\frac{B_3\left(1-C_1\right)}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1+2\left(1-C_1\right)\right\}B3}{\left(1-C_1-B_1\right)}<0$
Fiscal policy
	G	$\frac{1}{\left(1-C_1-B_1\right)}>0$	$\frac{m_1}{\left(1-C_1-B_1\right)}>0$	$\frac{\left(1/\mathit{MULT}\right)m_1-B_1}{\left(1-C_1-B_1\right)}>0$	$\frac{B_1}{\left(1-C_1-B_1\right)}<0$	$\frac{\left(1/\mathit{MULT}\right)m_1}{\left(1-C_1-B_1\right)}>0$
	T	$\frac{C_2}{\left(1-C_1-B_1\right)}<0$	$\frac{m_1{C}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{\left\{\left(1/\mathit{MULT}\right)m_1-B_1\right\}{C}_2}{\left(1-C_1-B_1\right)}<0$	$\frac{B_1{C}_2}{\left(1-C_1-B_1\right)}>0$	$\frac{\left(1/\mathit{MULT}\right)m_1{C}_2}{\left(1-C_1-B_1\right)}<0$

¹See text for explanation of symbols. The expressions in this table represent the total derivatives of each endogenous variable with respect to each policy variable, holding the other policy variables constant. For example, the expression in column 1, row 3 represents the effect of e on Y when NDA, RQ, G, and T are held constant; its negative sign indicates that currency revaluation reduces income.

As regards fiscal policy, an increase in taxes reduces income and the money stock, and the same effect results from a reduction in government expenditures; these classic Keynesian conclusions are not altered as a result of the openness of the economy when the exchange rate is fixed. The capital balance deteriorates as the effective demand for money decreases, while the improvement¹¹ in the current balance brings an addition to the money supply. The overall balance of payments also deteriorates as the deterioration in the capital balance more than offsets the current account improvement. Opposite conclusions follow in the case of a reduction in taxes or an increase in government expenditures.

Finally, consider the effects of exchange rate policy. Suppose that the authorities decide to devalue the national currency so that e is reduced. This improves the current account balance and increases income and the money stock. The capital balance also improves as the rise in the effective demand for money more than offsets the rise in money supply derived from the current account improvement; the result is an overall improvement in the balance of payments.

IV. The Model with a Floating Exchange Rate

One of the policy choices open to the policymakers in a small open economy is to allow the level of the exchange rate to be determined by market forces (supply and demand). The rules of a “pure float” require that the monetary authorities abstain completely from intervening in the foreign exchange market, so that the value of A^f is frozen at a given level, thereby allowing the foreign exchange rate to be determined purely by market forces. Nevertheless, this does not mean that the monetary authorities cannot influence the market indirectly through domestic credit and reserve policies. Indeed, they may form a view as to what should be regarded in the country as the “appropriate” level of the exchange rate, given the country’s growth goals, and may follow this up with domestic monetary adjustments. Furthermore, in practice the authorities may not relinquish completely their right to intervene directly in the foreign exchange market; in this case the exchange system becomes a cross between the fixed and floating rate systems in the sense that movements both in official reserves and in the exchange rate take place simultaneously (often referred to as dirty float). All these cases are analyzed below.

open float

Consider first the case of a pure float and assume that the authorities do not have a particular view as to the appropriate level of the exchange rate,¹² so that

$\begin{array}{cc}{A}^f-{A^f}_{-1}=B\left(Y,Y^f,e\right)+K=0,& \left(16\right)\end{array}$

and e is endogenous.

Freeing the foreign exchange rate has important consequences with regard to the determination of income and the stock of money. It will be shown that monetary policy is no longer powerless, as changes in this policy affect simultaneously the level of the exchange rate, the current account balance, income, and the equilibrium stock of money. On the other hand, fiscal policy becomes powerless as a means of controlling income and the money stock; it affects only the exchange rate and the current account balance. As in the case of an open economy with a fixed exchange rate, these results can best be discussed using the reduced form of the model.

Equations (16) and (3) through (9) imply the following money supply function:

$\begin{array}{cc}M=\left[(A^f+\mathit{NDA})\right]\left[(1+g\left(r\right)\right]/[g\left(r\right)+h\left(r\right)+\mathit{RQ}]\mathrm{.}& \left(17\right)\end{array}$

This function shows that, given the interest rate as determined by condition (14),¹³ the money supply is completely determined by the monetary authorities. Indeed, the money supply function becomes essentially identical to that in a closed economy (where A^f would, of course, be equal to zero). This indicates that in the case of a floating exchange rate regime, the demand for money must adjust to the supply of money through the mechanism of the exchange rate, if the equilibrium condition (9) is to be fulfilled.

Equations (1), (2), and (17), together with conditions (9) and (14), summarize the model with a clean and open floating exchange rate. Its operation may be described as follows. Given the domestic interest rate as determined by arbitrage operations in the integrated capital market and given the stock of money as determined jointly by the behavioral responses of the public and banks and by monetary policy, the foreign exchange rate, the current account balance, and national income move to that level which adjusts the demand for money to the stock of money.

The effects of monetary and fiscal policies on the economic targets are as follows (see Table 2). A tightening in monetary policy, either by a reduction in the net domestic credit of the central bank or by an increase in the required reserve ratio, has the effect of appreciating the national currency, worsening the current account balance,¹⁴ and reducing the stock of money and the level of income; the effect on the current account balance is negative, since the deterioration in this balance resulting from the appreciation of the national currency is only partially offset by the induced decline in national income. Fiscal policies have no effect on either the stock of money or national income. It has already been noted that the stock of money in a floating rate regime is determined solely by the monetary authorities, given the behavioral relations relating to the nonbank public and banks. Furthermore, with the interest rate given exogenously, national income adjusts to the level at which the demand for money equates the stock of money; its level is therefore determined independently of fiscal policies. Restrictive fiscal policies (either an increase in taxes or a reduction in government expenditures) only depress the level of the exchange rate and improve the current account balance.

Table 2.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate ¹

¹See Table 1, footnote 1, for the interpretation of the entries in this table.

²Effects on capital flows are exactly the negative of the effects on the current balance, since by assumption there is no change in foreign reserves. For example, dK/dNDA = –MULT(1 – C₁)/m₁ < 0, so that dK/dNDA + dB/dNDA = dA^t/dNDA = 0.

Table 2.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate ¹

Policy Instruments		Policy Targets
		Y	M	e	B 2
Monetary policy
	NDA	$\frac{\mathit{MULT}}{m_1}>0$	MULT > 0	$\frac{\mathit{MULT}\left(1-C_1-B_1\right)}{B_3{m}_1}<0$	$\frac{\mathit{MULT}\left(1-C_1\right)}{m_1}>0$
	RQ	$-\frac{H\cdot \mathit{MULT}}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}<0$	$-\frac{H\cdot \mathit{MULT}}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)}<0$	$-\frac{H\cdot \mathit{MULT}\left(1-C_1-B_1\right)}{B_3\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}>0$	$-\frac{H\cdot \mathit{MULT}\left(1-C_1\right)}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}<0$
Fiscal policy
	G	0	0	$-\frac{1}{B_3}>0$	–1
	T	0	0	$-\frac{C_2}{B_3}<0$	–C2 > 0

¹See Table 1, footnote 1, for the interpretation of the entries in this table.

²Effects on capital flows are exactly the negative of the effects on the current balance, since by assumption there is no change in foreign reserves. For example, dK/dNDA = –MULT(1 – C₁)/m₁ < 0, so that dK/dNDA + dB/dNDA = dA^t/dNDA = 0.

View Table

Table 2.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate ¹

Policy Instruments		Policy Targets
		Y	M	e	B 2
Monetary policy
	NDA	$\frac{\mathit{MULT}}{m_1}>0$	MULT > 0	$\frac{\mathit{MULT}\left(1-C_1-B_1\right)}{B_3{m}_1}<0$	$\frac{\mathit{MULT}\left(1-C_1\right)}{m_1}>0$
	RQ	$-\frac{H\cdot \mathit{MULT}}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}<0$	$-\frac{H\cdot \mathit{MULT}}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)}<0$	$-\frac{H\cdot \mathit{MULT}\left(1-C_1-B_1\right)}{B_3\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}>0$	$-\frac{H\cdot \mathit{MULT}\left(1-C_1\right)}{\left(\mathit{CD}+\mathit{RE}+\mathit{RQ}\right)m_1}<0$
Fiscal policy
	G	0	0	$-\frac{1}{B_3}>0$	–1
	T	0	0	$-\frac{C_2}{B_3}<0$	–C2 > 0

¹See Table 1, footnote 1, for the interpretation of the entries in this table.

²Effects on capital flows are exactly the negative of the effects on the current balance, since by assumption there is no change in foreign reserves. For example, dK/dNDA = –MULT(1 – C₁)/m₁ < 0, so that dK/dNDA + dB/dNDA = dA^t/dNDA = 0.

dirty float

From the point of view of the small open economy, preference for a dirty float exchange system (in which movements both of the exchange rate and of foreign assets held by the central bank are allowed) is understandable, since such a system permits the authorities to lean in favor of either alternative depending on the circumstances. On the other hand, this pragmatic arrangement has often been criticized by those who think that it gives the national authorities too much discretionary power which may be used to pursue exchange rate policies of a beggar-thy-neighbor nature and detrimental to the world community. Nevertheless, this criticism of the dirty float is not entirely correct since, as will be discussed below, there is no effect from direct intervention policy that could not also have been obtained through the appropriate manipulation of monetary policy under a pure float. The issue is more one of practicality in a world with fast-changing domestic and international conditions.

Assume first that the monetary authorities have unrestrained intervention power, so that A^f becomes a genuine policy instrument.¹⁸ In order to raise the level of income, the authorities may purchase foreign exchange (that is, raise A^f), thereby inducing a depreciation of the national currency and an improvement in the current account balance. Nevertheless, the same result could have been obtained by means of an expansionary monetary policy under a pure float,¹⁹ as shown in Table 2. It follows then that intervention policy cannot be used to neutralize changes in the exchange rate derived from monetary policy without making it self-defeating as a device for stabilization of income. Again, although intervention policy could be used to offset the exchange rate effect arising from fiscal policy so as to make this policy effective under a floating system, the stability of the exchange rate could also be achieved through an expansionary domestic credit policy; the choice is a matter of practicality. Intervention policy might be used, however, as an expeditious means of correcting excesses in the country’s own stabilization measures based on monetary policy.²⁰

Consider now the special case in which intervention in a floating rate system is limited to softening the destabilizing effects on the exchange rate arising from either domestic stabilization policies or from other exogenous shocks to the economy. This is sometimes known as intervention policy limited to “maintaining orderly market conditions,” which means that the authorities are intent on breaking sharp day-to-day fluctuations in the exchange rate while letting it drift in any direction in the longer run. In practice, this policy implies that the central bank stands ready to purchase (sell) foreign reserves whenever the exchange rate tends to appreciate (depreciate). Instead of equation (16), the following equation now applies:

$\begin{array}{cc}{A}^f-A_{-1}^f=E\left(e-e_{-1}\right);E_1>0& \left(19\right)\end{array}$

where E is a functional form. With this change, equation (17) becomes

$\begin{array}{cc}M=\left[E\left(e-e_{-1}\right)+A_{-1}^f+\mathit{NDA}\right]\left[1+g\left(r\right)\right]/\left[g\left(r\right)+h\left(r\right)+\mathit{RQ}\right]\mathrm{.}& \left(20\right)\end{array}$

Finally, from equation (19) and the balance of payments identity, equation (5), it also follows that

$\begin{array}{cc}B\left(Y,Y^f,e\right)+K=E\left(e-e_{-1}\right)\mathrm{.}& \left(21\right)\end{array}$

The system of reduced form equations consists of (1), (2), (20), and (21), with the explicit endogenous variables Y, e, K, and M. (B and A^f are also determined endogenously.) Exogenous to the system are the fiscal and monetary policy instruments (r,Y^f) and the lagged values of e and A^f.

The solution to this system for the effects of changes in the policy variables on the endogenous variables always yields results which are between those of the pure floating and fixed exchange rate systems (see Table 3). In this hybrid system, domestic credit policy is still effective as a means of determining income but less so than in the case of the pure float, since part of the effect is leaked through intervention policy, which prevents the exchange rate from changing as much as it does in the pure float and thereby reduces its impact on the current account balance. On the other hand, fiscal policy regains some of its effectiveness as an income-stabilizing device to the extent that the offsetting movements in capital flows, the exchange rate, and the current balance, which are characteristic of the pure float, are counteracted by intervention policy; the money supply is affected in the same direction as the initial change in fiscal policy.

Table 3.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate Tempered by Movements of Reserves (Dirty Float) ¹

¹See Table 1, footnote 1, for the interpretation of the entries in this table. Note that D = (1 – C₁ – B₁) MULT E₁ – m₁B₃ > 0.

²The effects on the economic targets from changes in the alternative monetary and fiscal policy variables (RQ and T, respectively) have been omitted from this table, since they are necessarily opposite to those shown here.

Table 3.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate Tempered by Movements of Reserves (Dirty Float) ¹

Policy Instruments		Policy Targets
		Y	M	e	K	B	Af
Monetary policy2
	NDA	$\frac{-\mathit{MULT}{B}_3}{D}>0$	$\frac{-\mathit{MULT}{m}_1{B}_3}{D}>0$	$\frac{-\left(1-C_1-B_1\right)\mathit{MULT}}{D}<0$	$\frac{\mathit{MULT}\left\{\left(1-C_1\right)B_3-\left(1-C_1-B_1\right)E_1\right\}}{D}<0$	$\frac{-\mathit{MULT}\left(1-C_1\right)B_3}{D}>0$	$\frac{-\left(1-C_1-B_1\right)E_1}{D}<0$
Fiscal policy 2
	G	$\frac{\mathit{MULT}{E}_1}{D}>0$	$\frac{\mathit{MULT}{m}_1{E}_1}{D}>0$	$\frac{m_1}{D}>0$	$\frac{-\mathit{MULT}{E}_1{B}_1-m_1\left(B_3-E_1\right)}{D}>0$	$\frac{\mathit{MULT}{B}_1{E}_1+B_3{m}_1}{D}<0$	$\frac{m_1{E}_1}{D}>0$

¹See Table 1, footnote 1, for the interpretation of the entries in this table. Note that D = (1 – C₁ – B₁) MULT E₁ – m₁B₃ > 0.

²The effects on the economic targets from changes in the alternative monetary and fiscal policy variables (RQ and T, respectively) have been omitted from this table, since they are necessarily opposite to those shown here.

View Table

Table 3.

Effects of Policy Instruments in a Small Open Economy with a Floating Exchange Rate Tempered by Movements of Reserves (Dirty Float) ¹

Policy Instruments		Policy Targets
		Y	M	e	K	B	Af
Monetary policy2
	NDA	$\frac{-\mathit{MULT}{B}_3}{D}>0$	$\frac{-\mathit{MULT}{m}_1{B}_3}{D}>0$	$\frac{-\left(1-C_1-B_1\right)\mathit{MULT}}{D}<0$	$\frac{\mathit{MULT}\left\{\left(1-C_1\right)B_3-\left(1-C_1-B_1\right)E_1\right\}}{D}<0$	$\frac{-\mathit{MULT}\left(1-C_1\right)B_3}{D}>0$	$\frac{-\left(1-C_1-B_1\right)E_1}{D}<0$
Fiscal policy 2
	G	$\frac{\mathit{MULT}{E}_1}{D}>0$	$\frac{\mathit{MULT}{m}_1{E}_1}{D}>0$	$\frac{m_1}{D}>0$	$\frac{-\mathit{MULT}{E}_1{B}_1-m_1\left(B_3-E_1\right)}{D}>0$	$\frac{\mathit{MULT}{B}_1{E}_1+B_3{m}_1}{D}<0$	$\frac{m_1{E}_1}{D}>0$

¹See Table 1, footnote 1, for the interpretation of the entries in this table. Note that D = (1 – C₁ – B₁) MULT E₁ – m₁B₃ > 0.

²The effects on the economic targets from changes in the alternative monetary and fiscal policy variables (RQ and T, respectively) have been omitted from this table, since they are necessarily opposite to those shown here.

V. Capital Controls and Their Effects on Other Economic Policies

This section analyzes the working of the economy under nonmarket policy restrictions on the international mobility of capital. The immediate consequence of a policy of capital controls is to prevent the functioning of the interest rate arbitrage mechanism, which means that the domestic interest rate will then be affected by decisions regarding the money supply, the exchange rate, and fiscal policy. It is assumed here that capital controls, if instituted, will apply to all sectors of the economy, so that no loopholes in the application of the policy are permitted.

Assume then that the policymakers wish to fix the net amount of capital inflows (or outflows) and that they do this by imposing a direct capital control regulation. This is represented by

where $\overline{K}$ is the net amount of capital flows allowed by the policymakers and condition (14) no longer applies. The consequences for the effectiveness of other policies under fixed and floating exchange rate regimes will first be analyzed and then the effects from the introduction of changes in capital controls themselves will be discussed.