On the Monetary Analysis of an Open Economy

The original contribution of the monetary approach to the balance of payments was to focus on conditions in the money market in explaining developments in the external sector of the economy. Payments’ surpluses or deficits were seen as arising from imbalances between the supply of and demand for money at a given exchange rate. Alternatively, under a free float, the exchange rate was viewed as varying to balance the demand for money with a fixed supply of money. Subsequent elaboration of this approach has broadened its perspective: alternative assets, including bonds and foreign currencies, have been introduced into the analysis together with more complex treatment of the formation of expectations and of differing rates of adjustment in asset and product markets. Using this asset-market approach, it has been possible to model the current and capital accounts separately and to analyze such phenomena as exchange rate overshooting.1 But money itself is still usually treated as merely cash balances and noninterest-bearing demand deposits. This fact seems curious, as the proportion of such narrow money in total liquidity has declined with the increasing attraction of quasi-money and the widening spectrum of near-money.


The original contribution of the monetary approach to the balance of payments was to focus on conditions in the money market in explaining developments in the external sector of the economy. Payments’ surpluses or deficits were seen as arising from imbalances between the supply of and demand for money at a given exchange rate. Alternatively, under a free float, the exchange rate was viewed as varying to balance the demand for money with a fixed supply of money. Subsequent elaboration of this approach has broadened its perspective: alternative assets, including bonds and foreign currencies, have been introduced into the analysis together with more complex treatment of the formation of expectations and of differing rates of adjustment in asset and product markets. Using this asset-market approach, it has been possible to model the current and capital accounts separately and to analyze such phenomena as exchange rate overshooting.1 But money itself is still usually treated as merely cash balances and noninterest-bearing demand deposits. This fact seems curious, as the proportion of such narrow money in total liquidity has declined with the increasing attraction of quasi-money and the widening spectrum of near-money.

The original contribution of the monetary approach to the balance of payments was to focus on conditions in the money market in explaining developments in the external sector of the economy. Payments’ surpluses or deficits were seen as arising from imbalances between the supply of and demand for money at a given exchange rate. Alternatively, under a free float, the exchange rate was viewed as varying to balance the demand for money with a fixed supply of money. Subsequent elaboration of this approach has broadened its perspective: alternative assets, including bonds and foreign currencies, have been introduced into the analysis together with more complex treatment of the formation of expectations and of differing rates of adjustment in asset and product markets. Using this asset-market approach, it has been possible to model the current and capital accounts separately and to analyze such phenomena as exchange rate overshooting.1 But money itself is still usually treated as merely cash balances and noninterest-bearing demand deposits. This fact seems curious, as the proportion of such narrow money in total liquidity has declined with the increasing attraction of quasi-money and the widening spectrum of near-money.

Less attention has been paid to improving the analysis of the liability side of the balance sheet. The standard assumption has been that all domestic assets are issued directly by the monetary authorities. The supply of money or bonds varies only if the authorities intervene in bond or foreign exchange markets, or indulge in “helicopter” operations. This view provides no role for the domestic private borrower in generating the basic demand for credit, or for the banking system in intermediating between borrower and lender.

This paper argues for a more sophisticated treatment of the supply of money. The simplifications that all money bears no interest and that all assets are claims on the government clearly limit the richness of the asset-market approach. The approach does not permit the analysis of the joint determination of money interest and exchange rates, or of how shifts in banking behavior or the demand for credit affect equilibrium. Such limitations may perhaps be forgiven at the early stages of development of a theory: some abstraction is, of course, necessary to facilitate the analysis of a complex situation. However, if the simplifications lead to policy conclusions that are seriously misleading or incomplete, the incorporation of a more satisfactory treatment becomes urgent.

The body of this paper has four main sections. Section I discusses ways in which recent institutional changes make the simplified treatment of money increasingly unrealistic and argues in favor of an approach in which the demand for credit is given the same prominence as the demand for monetary assets. Section II develops a Keynesian model of monetary equilibrium in a competitive banking system that provides the basic framework for the subsequent analysis. The following two sections study the behavior of this model from different angles: Section III looks at long-run equilibria in which prices and wages have adjusted to relative values accepted by firms and unions; Section IV focuses on short-run equilibria in which prices and wages are sticky so that relative prices may be disturbed by exchange rate movements. These two sections are not intended to sum to an exhaustive study of the situations considered,2 but rather to indicate the dangers of the standard approach and to suggest a preferable way of proceeding.

I. Interest Rates, Indeterminacy, and the Demand for Credit

The central point of this section is that standard asset-market models of the balance of payments or exchange rate determination cannot easily assimilate quasi-money into their framework. Once a variable rate of interest is paid on domestic liquidity in such models, the degrees of freedom for monetary policy increase; it seems that the authorities may simultaneously achieve two out of exchange rate, liquidity, and interest rate targets. This finding appears to conflict with conventional views on the scope for monetary policy.

The first part of this section argues that recent institutional changes make it essential to include the interest earned on money holdings as a relevant feature of a satisfactory model of exchange rate determination. The second part contests that such a model must then also include treatment of the demand for domestic credit and of the intermediation between borrower and lender in order to provide a realistic framework for the analysis of monetary policy.


A peculiarity of the asset-market approach is that it does not include any explicit interest payable on money holdings, as if the only types of money available were cash balances and demand deposits (i.e., narrow money) but not time and savings deposits (i.e., quasi-money). This neglect would be justified analytically if (a) the demand for all relevant assets were completely independent of the interest paid on quasi-money, or (b) the nominal rates of interest paid on quasi-money were fixed, or (c) an alternative asset were included in the model whose properties essentially mirrored the properties of quasi-money. In (a), the rates of interest paid on quasi-money would be immaterial to the determination of equilibrium, provided that the distribution effects of shifts in interest rates were of second-order importance. In (b), the nominal rates of interest would become part of the given economic environment and could be treated as implicitly factored into the description of people’s behavior. In (c), the omission of quasi-money would be an appropriate simplification without major analytical consequences unless the focus of interest was specifically quasi-money itself.

In many if not most open economies, the first two of these conditions no longer hold, if they ever did. The institutional changes that have occurred in financial markets in recent years have tended to increase both the own-interest elasticity of the demand for money and the flexibility of the rates of interest paid on quasi-money.

From first principles, one would conjecture that the demand for money would depend on the total demand for financial wealth, on the relative rates of return expected from the various forms of money and alternative assets, and on the variation of such characteristics as convenience, liquidity, security, and the anticipated stochastic distribution of real returns among these assets. These nonpecuniary characteristics of different assets give the wealth holder reason to differentiate between assets bearing the same expected returns so that, in particular, the demand for money would not be perfectly interest elastic. Indeed, the traditional theories of money, such as the quantity theory, go to the other extreme and implicitly see money as having nonpecuniary characteristics so unique that the demand for money can be treated as essentially independent of its own rate of return. But, in most countries, the own-interest elasticity of the demand for money is probably increasing over time as institutional changes occur that make the characteristics of money less special. The precise story will be different in each country, but the general trend seems clear.

First, new types of financial instrument have been introduced and familiar types have been revamped to increase the attraction and the range of alternative forms of money and near-money. Deregulation has permitted commercial banks to offer more flexible and remunerative types of account, while interest-bearing stores of value with reasonable liquidity are also offered now by credit unions, savings and loan associations, building societies, money market mutual funds, and many other institutions. Increasingly, these latter institutions extend check-writing facilities to their accounts to provide means of payment. In addition, credit cards and “easy access” overdraft facilities make available instant lines of credit that may also be used as means of payment.

Second, the progressive removal of domestic exchange controls and the increasing integration of world financial markets have substantially reduced the transactions cost of holding bank accounts and other assets abroad, and hence have made available a range of close foreign substitutes for domestic monetary assets. A monetary asset held in a foreign banking system has liquidity and security characteristics that are similar but not identical to those of a domestic deposit with an equivalent maturity. If denominated in a foreign currency, its pattern of expected real returns might be strongly differentiated from the domestic alternatives by uncertainty over future exchange rate variations. But the existence of forward currency markets and offshore banking facilities provides the opportunity of holding assets abroad without incurring any concurrent exchange risk.

Third, the growth of multinational operations and the accumulation of financial wealth owned outside the traditional financial centers have provided a substantial pool of highly mobile and professionally managed funds that are seeking attractive short-term investments. As a result, it is no longer true, at least in the major developed countries, that domestic money is held only by domestic residents. By its very nature, this foreign demand is likely to be particularly interest elastic; the international placement of funds will be closely related to comparative returns in different markets.

All in all, there exist many domestic and foreign substitutes for both narrow money and quasi-money except in countries that are at a rudimentary level of financial development and that have effective controls on capital mobility. Outside these cases, the demand for the various types of money and near-money will tend to be sensitive to the rates of interest payable on quasi-money.

Nevertheless, if nominal rates of interest payable on all forms of money were fixed and not subject to policy manipulation, then they would not need to be explicitly included as variables in an economic model, whatever the demand elasticities. However, in most developed countries, while interest still may not be paid on demand deposits, the interest rates paid on time and savings deposits and on other types of liquidity are increasingly flexible. This flexibility reflects in part a new propensity on the part of monetary authorities to relax restrictions and to allow interest rates to be set by the financial institutions concerned. This deregulation is usually accompanied by measures aimed at increasing the degree of competition for funds among banks and between banks and other financial institutions. In consequence, interest rates on at least the less liquid monetary instruments tend to be highly sensitive to market conditions.

The literature does contain, of course, many studies that include narrow money and some other domestic asset (usually bonds) that bears interest at a market-determined rather than a fixed rate. Why is more work needed? One reason is that a narrow money/bond model is inadequate for the analysis of monetary policy using broad money as an explicit target. A second reason is that the real interest rate on domestic bonds is often assumed to be fixed by a perfectly elastic international demand for bonds, as if domestic and foreign bonds were indistinguishable; this contrasts sharply with the companion assumption that narrow money is held only by domestic residents.3 Neither assumption captures the intermediate view of quasi-money as an interest-bearing differentiated asset held both for its liquidity and as a store of value. Models that do treat bonds as a differentiated asset,4 on the other hand, let the supply of bonds and money be determined separately by the authorities. Such analysis would be applicable to a narrow money/quasi-money framework only if the authorities were to develop policy instruments that were capable of controlling these two monetary aggregates separately; it also faces a serious theoretical problem (considered in the next subsection of this paper) if both types of asset earn interest.


Consider the following simple model of the financial sector. Let the nominal demand for domestic liquid assets, denominated in sterling (DL), be an increasing function only of domestic nominal income (Y) and the nominal interest rate paid on liquid assets (R). Let Y itself be a decreasing function of the exchange rate, e (the price of sterling in terms of special drawing rights—SDRs). Let the supply of liquid assets (L) be controlled by the government. Then the financial market equilibrium condition is written


Figure 1 illustrates this condition as the locus AA in L/R space for a given exchange rate. AA slopes up, as increasing the interest rate raises the demand for domestic liquidity; the more interest elastic is the demand for liquidity, the steeper the slope of AA. An exchange rate appreciation reduces nominal income and shifts AA down and to the right, so altering the feasible combinations of L and R. In this situation, the government apparently has two degrees of freedom in selecting its monetary policy: it may choose to set two of L, e, and R. This result is contrary to the conventional wisdom that the government has only one degree of freedom—that (for example) if it chooses to aim for a liquidity target, it must allow the exchange rate and interest rates to be market determined. According to the present model, the government could well achieve both liquidity and exchange rate targets, despite shifts in the demand for liquid assets, if it were willing to vary the interest rate as required.

The salient point is that, after considering the rate of interest on liquidity, there are not one but two prices of domestic assets: the interest rate and the exchange rate; this feature leads directly to the extra degree of freedom. The difficulty does not go away if one relaxes assumptions to broaden the picture of the asset market: introducing domestic bonds would lead to an extra market-clearing equilibrium condition and two extra variables—the rate of return on bonds and the aggregate supply of bonds—and so increase rather than reduce the available degrees of freedom. Introducing foreign assets would not alter the relevant equilibrium conditions if the home country were small in world markets; otherwise, it would add one more equation and rate of return per asset. Separating domestic liquidity into narrow money and quasi-money would have no analytical consequence in this context unless the supplies of narrow money and quasi-money could be controlled independently—which would then leave the degrees of freedom unaffected. Nor would it help to hypothesize, on the basis of an infinite willingness by wealth holders to arbitrage between domestic and foreign money markets, that the demand for domestic liquidity was perfectly interest elastic. Not only would this hypothesis be implausible, requiring wealth holders to be utterly indifferent to the currency composition and security of their portfolios, but it would leave the total demand for domestic liquid assets indeterminate. In consequence, equation (1) could not be solved for the exchange rate even if the supply of liquid assets were fixed; that is, e and L may be set independently although R is now fixed to the rate of return on foreign assets. Finally, allowing for the effect of anticipated movements of the exchange rate on the expected real return to domestic liquidity would give the authorities an additional lever on the present through their pronouncements on future monetary policy.

There are two alternative directions in which to proceed. One is to accept that the authorities may indeed determine two out of the three variables—supply of liquid assets, interest rate, and exchange rate—and to investigate the basis on which policy should then be decided. In developing countries, this approach has in fact received much attention. For example, the literature on the choice of interest rates takes for granted that a stable exchange rate can be achieved simultaneously with controlled interest rates.5 Implicit in these discussions is the belief that interest and exchange rates are independent policy variables. But, in developed countries, the experience has been that the authorities have limited power to control more than one of the key variables on a permanent basis. Attempts to meet the targets for both money supply and the exchange rate, or interest and exchange rates, have eventually foundered on external disequilibrium. Then it would seem that the asset-market approach is an insufficient theoretical framework for policy analysis in such countries, as its practical implications are ill-founded once it is recognized that money—like other assets—has a variable rate of return. In reacting to this conclusion, it would appear necessary to extend the asset-market approach to include additional factors that would effectively constrain the application of monetary policy.

The solution offered in this paper is to extend the analysis beyond the simplistic assumption that the supply of liquidity is under the direct control of the monetary authorities. Instead, monetary assets are seen as the counterparts of monetary liabilities, consisting of bank loans and advances to the private sector and other debt instruments. Monetary equilibrium in a competitive market is then modeled as requiring that the demands for monetary assets and liabilities be matched. Use of this extra condition reduces the degrees of freedom and provides the desired result that the authorities are able to determine only one of the three key variables.6

Clearly, this approach requires an analysis of the demand for monetary liabilities that is compatible with the treatment of the demand for monetary assets. Economists have generally paid far less attention to the study of the demand for credit and its components than they have to asset-portfolio selection. But, on first principles, it would seem plausible to approach the demand for monetary liabilities in a way that is analogous to the demand for monetary assets, that is, as depending on the total demand for finance, the relative rates of interest payable on the various forms of bank credit and alternative sources of funds, and the variation of such characteristics as convenience, liquidity, and the anticipated stochastic distribution of interest costs between these sources. As with the demand for monetary assets, the existence of these other characteristics gives the borrower reason to differentiate between sources requiring the same expected rates of return so that, in particular, the demand for monetary liabilities would not be perfectly interest elastic. In practice, however, this interest elasticity is probably increasing over time, as institutional changes occur that make the characteristics of bank credit less unique. These changes include a widening range of domestic financial institutions and increasing opportunities to raise funds overseas.

A model of monetary equilibrium that matches the demand for monetary assets and liabilities must also inevitably contain a submodel of the means of intermediation between these two, that is, the banking system. This submodel would ideally explain the ratio between bank reserves and lending and the spread between the rates of interest on deposits and lending. Also, this submodel should contain mechanisms by which the monetary authorities would be able to influence bank behavior to affect the position of monetary equilibrium. Of course, the appropriate modeling of the banking system would depend on the degree of competitiveness within the system, conventions of bank behavior, and the instruments of monetary control, and hence would vary between countries. This paper does not attempt a serious model of bank behavior, however, and uses an ad hoc approach that is compatible with both competitive and oligopolistic types of banking system.7

One of the messages of this section of the paper is that no single approach to the monetary analysis of an open economy is necessarily appropriate for the study of all countries. In some countries—in particular, developing countries in which interest rates are fixed and credit is allocated by rationing—the conventional monetary approach may indeed be adequate for many purposes. But elsewhere a more elaborate analysis may be necessary to capture the behavior of the system and to delineate the policy options that are open to monetary policy.

II. Analytical Framework

This section develops a model of an industrial economy, open to both trade and financial flows. This model is Keynesian in the sense that relative prices are not perfectly flexible and domestic output depends on demand. To simplify, the real economy is reduced to one sector, and the fiscal branch of the government is omitted. Savings and investment are ignored, as if consumers and firms always held at the start of the period the stocks of wealth and capital that they wished to hold at the period’s end. The financial sector consists only of an aggregate banking system that intermediates between the demands for domestic assets and domestic credit, both denominated in sterling.

The rudimentary nature of this model should be recognized—it is intended for illustrative purposes rather than for immediate application. A more useful version for policy studies would need to distinguish the expenditure, taxation, and financial activities of government from the private sector to allow analysis of the impact of, for example, an altered public sector deficit or open market operations. The effects of stock adjustment on the balance of payments and the exchange rate are also important and would need to be considered in determining the impact of shocks to the domestic economy.8


The domestic economy produces a single good that is differentiated from the output of the rest of the world. The price of this good is set by markup over costs on domestic and world markets; firms set production to meet the resulting domestic and export demand; employment is determined by the labor required to achieve this level of production. Wages are fixed by trade unions that seek to achieve an appropriate balance between the take-home pay and the employment of their members. The rest of the world is large relative to the domestic economy in the sense that activity levels and prices do not depend on the state of the domestic economy. Imports are in perfectly elastic supply and are priced by arbitrage with world markets.

Let real income be y, and production f (by convention, lowercase letters indicate real values with the price of the domestic good being used as the numeraire); y and f are related by


where w is real domestic wealth, k is the real domestic capital stock, and r* is the real rate of return on foreign investment.

It is assumed throughout that the economy is in static stock equilibrium: given activity levels and prices, firms and consumers hold just the stocks of physical capital and wealth that they desire. Domestic firms achieve their target capital stock (k), which is an increasing function of the output level


Domestic consumers achieve their target level of wealth (w), which is an increasing function of y


Firms sell what they produce, and consumers spend what they earn.

Let α be the price of the domestic good relative to imports, that is, a measure of the competitiveness of the domestic economy; its value depends on the interaction between the price and wage setting mechanisms. Arbitrage sets the domestic price of imports (PM) to the world price (PM*) adjusted by the exchange rate e (again the price of sterling in terms of SDRs)


Then, the domestic price of the domestic good (PD) is given by


The proportion of income allocated to the consumption of the domestic good (λ) is a decreasing function of α. So, demand for the domestic good on the domestic market is given by


Sales of exports are also a decreasing function of α


Then, production is given by


Solving equations (2)(4) and (9) gives output (f) and real income (y) as decreasing functions of α and exogenous variables


Foreign firms and consumers are also assumed to hold their target stocks of capital and wealth. These targets are fixed in terms of imported goods and may be expressed as α.k* and α.w*, respectively, in terms of domestic goods.

In the long run, relative prices and wages reach equilibrium in the sense that firms and unions, knowing the state of the economy and taking the actions of others as given, are content with the prices and wages, respectively, that they set. Movements of the exchange rate affect the overall level of prices and wages, via their effect on the nominal price of imports, but not relative prices. The level of the long-run real domestic lending rate (r) does, however, affect competitiveness: a rise in the cost of funds will increase firms’ markups, and hence raise α (¯indicates a long-run value); a may also shift owing to fundamental changes in technology, institutional attitudes, or other structural facets of the economic environment. So,


In the short run, wage/price equilibrium is not necessarily maintained. In particular, unanticipated exchange rate movements may affect the value of a, as the import price responds faster than wages or the price of the domestic good. At a single point in time, this stickiness may be expressed in the equation


where α(.) is an increasing function of e. Short-run interest rates are supposed not to affect α on the grounds that firms do not vary the value of their markups over the business cycle.

In the absence of future shocks, α increases over time at rate β toward its long-run equilibrium value, α; β is a decreasing function of α−α


The domestic rate of inflation (ρ), which is defined as the rate of change of the domestic price over time, depends on β and on the rate of depreciation of the exchange rate (γ), according to



The domestic financial sector consists of a banking system, regulated by a central bank.9 Domestic monetary assets and liabilities are denominated in the domestic currency, sterling. There are no controls on external financial flows to restrict domestic access to foreign financial markets or foreign access to domestic financial markets. Sterling is, however, differentiated from other currencies by its own particular nonpecuniary characteristics.

Domestic firms finance part of their capital (K) through sterling credit from domestic banks and part through loans from foreign banks; the remainder is provided by the firms’ owners. Firms pay a nominal interest rate of R on credit from domestic banks; they expect sterling to depreciate at rate γ and domestic prices to increase at rate ρ. The share of capital financed through sterling credit (c) is a decreasing function of the real interest rate (R − ρ) and of a measure of the cost of domestic loans relative to foreign loans (R − γ).10

Foreign firms may also borrow from domestic banks. They raise the share c* of their capital (K*) in sterling loans; this share also decreases with R − γ. So the total demand for sterling loans (DC) is given by


Domestic wealth holders divide their wealth (W) between holdings of noninterest-bearing sterling currency, sterling bank deposits bearing interest at rate R − δ (δ being the interest rate spread), property rights, and foreign assets. They also expect the domestic currency to depreciate at rate γ and domestic prices to increase at rate ρ. The portfolio share of sterling currency (kM1) is a decreasing function of the inflation rate and of the real interest paid on deposits (R − δ − ρ); the portfolio share of bank deposits kMQ) is an increasing function of ρ, R − δ − ρ, and R − δ − γ. Foreign wealth holders place kMQ* of their wealth (W*) in sterling deposits; kMQ* increases with R − δ − γ. So the total demand for domestic broad money (DM2) is given by


The domestic banking system intermediates between the demand for sterling deposits and that for sterling loans. Commercial banks offer facilities for interest-bearing accounts to depositors and allocate their funds either to loans or to reserve deposits with the central bank. The central bank’s liabilities—the monetary base—include these reserve deposits and also its issue of notes and coins; it deploys its assets either in lending to commercial banks or in foreign exchange reserves.

The central bank has various policy instruments at its direct disposal: it sets the required minimum ratio between commercial bank reserves and deposits, the rates of interest that it pays on required and on excess reserves, and the rate of interest that commercial banks pay to borrow from it. With these instruments, it may influence the behavior of commercial banks. A full account of the interactions taking place within the banking system is not attempted here; its details would depend on such factors as the degree of competition between commercial banks and the way in which commercial banks responded to risk. Briefly, the ratio of total sterling broad money to sterling bank credit (η) and the commercial bank interest rate spread δ depend on the required reserve ratio and the spread between the interest rates that the central bank pays on deposits and the rate that it charges for loans: increasing either of these will tend to raise both commercial banks’ desire for reserves and the cost of intermediation, hence to reduce η and increase δ. To simplify the analytical framework, the general presumption will be that the central bank operates to keep η and δ fixed and invariant to changes in macroeconomic variables. The central bank’s main policy instrument is the basic level of its interest rates, which is assumed to have a direct impact on R and hence on other potential macroeconomic variables.

Allocation within the banking system is competitive, in the sense that neither borrowers nor lenders face quantity constraints on their transactions at given prices. The sterling interest rate and/or exchange rate (depending on the government’s monetary policy) adjust to ensure that the money market clears. The authorities are sufficiently knowledgeable to achieve monetary targets by setting appropriate values for the available instruments.

III. Long-Run Monetary Equilibrium

In the long run, domestic competitiveness is expected to remain constant at α at a value depending on r(≡R − ρ); the expected rate of inflation (ρ) and the expected rate of exchange rate depreciation (γ) are then equal. Domestic consumers and firms achieve their stock targets, w and k, respectively, both of which are functions of long-run income, y. Using equations (3)(6), (10)(12), and (1617), the long-run conditions for money market equilibrium are as follows:


Equation (18) is the condition that the demand for sterling assets equal the stock of broad money; equation (19) requires that the demand for sterling loans be compatible with the stock of broad money; equation (20) sets the value of domestic competitiveness.

Provided that the authorities follow a monetary policy that is compatible with stable prices—hence, a stable exchange rate—p is set to zero; 8 and t) are taken as behavioral parameters of the banking system, while PM*, w*, and k* are exogenous variables. Equations (18), (19), and (20) can then be solved for the real interest rate, relative prices, and the stock of real broad money, m2(≡M2/.(PM*/e)]). Domestic prices and other nominal values depend on the authorities’ choice of monetary target; they may set either the exchange rate or the broad money stock (or some other nominal value).11 Within this framework, the government cannot affect real variables but has one degree of freedom to affect nominal values.

Figure 2 illustrates long-run equilibrium in r/m2 space. AA represents equations (18) and (19) in r/m2 space; it slopes upward provided that the shift in portfolio composition toward broad money as r increases outweighs the negative impact of r on income and hence the wealth target. BB represents equations (19) and (20) in r/M2 space; it must slope down as both the substitution and the income effect reduce the demand for loans as r increases. The equilibrium real lending rate is r, while r′ + δ′ is the equilibrium real deposit rate and m2δ is the equilibrium stock of real broad money. Despite financial openness and monetary stability, r′ + δ′ may well deviate from r* because of the imperfect substitutability of foreign currency for sterling; the stronger is the currency preference for sterling over other currencies, the less interest elastic the demand for sterling assets and liabilities, the shallower the AA and BB loci, and the more the scope for divergence between r′ + δ′ and r*.

What is the effect of an exogenous improvement in competitiveness representing, for example, a reduction in wage demands or an improvement in domestic technology? Assuming that the income effects on stock targets of a reduction in α are stronger than the disinflationary price effects, the loci AA and BB both shift upward in m2/r space. The resulting equilibrium has a higher real money stock. The impact on real interest rates is uncertain, depending on the relative slopes of AA and BB, that is, on the relative income and interest elasticities of the demand for sterling deposits and loans. Other comparative-static effects, such as the impact of a shift in portfolio preferences or in stock targets, may be deduced in a similar manner.

If the authorities are willing to sacrifice price stability or to manipulate the behavior of the banking system, that is, to alter the values of ρ, δ, or η, then they may regain some influence over real variables:

(i) The value of the rate of inflation would matter if and only if it would affect the total domestic demand for broad money even though real interest rates were kept constant. For instance, if the shift away from noninterest-bearing money that was caused by inflation were to reduce the total monetary share of wealth portfolios, then it would shift AA down to the right in r/m2 space. In equilibrium, an increased rate of expansion of M2 or rate of depreciation of e, leading to a higher ρ, would result in an increase in the real interest rate (hence, a fall in real income) and a reduction in the real money stock. However, one would expect that persistent inflation would eventually lead to the evolution of interest-bearing means of payment, such as those mentioned in Section I, to make it possible to hold liquid funds without incurring a negative return. This institutional development would limit the extent of the inflationary impact on the demand for broad money.

(ii) By increasing the cost of bank intermediation to engineer a rise in the interest rate spread (δ), the authorities may achieve a reduction of real monetary balances. In Figure 2, raising the spread from δ′ to δ″ reduces equilibrium real monetary balances to m2". The deposit interest rate falls while the lending rate rises. The real effect of such a policy is limited by the extent of any shift of intermediation from the domestic banking system elsewhere—to alternative domestic financial institutions or offshore.

(iii) Reducing the proportion of foreign reserve assets that are held against the sterling liabilities of the banking system (i.e., raising η) increases the availability of sterling loans, and hence shifts BB down to the left. In equilibrium, this action reduces domestic interest rates and increases real income. The extent of this effect is limited by the degree of interest elasticity of demand for sterling broad money balances: if interest elasticity is high, AA tends to be steep, and a shift in η will be reflected mainly in a reduction in real money balances rather than in a reduction in interest rates. Moreover, the scope for raising η is constrained by the country’s need-for foreign reserves to guard against short-run balance of payments difficulties caused by exogenous shocks.

IV. Monetary Equilibrium with Wage/Price Adjustment

This section investigates the short run during which relative prices and wages adjust slowly toward their long-run values. Movements of the exchange rate, which have an immediate impact on import and export prices, disturb the relative price of the domestic good on the home market from its long-run level. So a is an increasing function of the exchange rate as in equation (13), while its rate of increase β (which rises with (α − α), as in equation (14)) becomes a declining function of the exchange rate. The rate of domestic price inflation (ρ) depends on the rate of depreciation of the exchange rate (γ) and on β as in equation (15). Real capital and wealth stocks are increasing functions of short-run output and income, f and y, respectively, which depend on α as in equations (10) and (11).

Under these assumptions, the short-run monetary equilibrium conditions include


Taking δ and η as behavioral parameters, these conditions are three equations in five endogenous variables: M2, R, e, ρ, and γ. Clearly missing is an equation governing the formation of exchange rate expectations. But, in general, if expectations are to approach rationality, then it would be necessary to discover the complete adjustment path in order to model γ. To simplify, let the domestic portfolio share of broad money (kM1 + kMQ) be independent of the inflation rate. Then the equation system (21)–(23) may be expressed in terms of only four endogenous variables: M2; e; the real interest rate, R − ρ; and the relative interest rate factor, R − γ. In this formulation, real values in the present are determined irrespective of anticipated future nominal values; only the nominal interest, inflation, and depreciation rates cannot be discovered without a model of the complete adjustment path.

In either case, the authorities possess—in the short run—a degree of freedom to set a monetary target. Because prices are sticky, the value of this target has repercussions on the real economy as well as on purely nominal values. In particular, monetary policy can be seen as an important element of any macroeconomic stabilization strategy. Alternative fixed-target types of regime will provide the economic system with alternative insulation properties against exogenous shocks, while shifts in target could in theory be used to accelerate the process of wage/price adjustment.

To illustrate how the system would behave under different policy regimes, it is convenient to simplify the model further by assuming that the rate of wage/price adjustment is negligible; that is, β = 0. Provided that exogenous variables and policy targets are expected to remain at their present values, the equilibrium in the next period is the same, in both real and nominal terms, as equilibrium in the present. So ρ and γ can also be set to zero. Last, let δ equal zero. Then the short-run monetary equilibrium conditions can be reduced to


Rather than attempt a systematic survey of the comparative-static properties of this system,12 this section concentrates on a few topics to bring out some of the possibilities that are inherent in this framework of analysis.


If the authorities act so as to fix R to some target value, e and M2 then adjust to ensure monetary equilibrium in the short run. Figure 3 illustrates equations (24) and (25) in e/M2 space. The loci AA and BB—representing equations (24) and (25), respectively—both slope down as, with a fixed interest rate, an appreciation in the exchange rate leads to a fall in the demand for both sterling assets and liabilities. Equilibrium occurs at P, with an exchange rate of e′ and a money stock of M2′.

An increase in the interest rate target shifts AA out to the right and BB down to the left. Provided that the slope of AA is steeper than the slope of BB (i.e., provided the wealth target is more responsive than the capital target to shifts in the exchange rate), the new equilibrium at Q has a higher exchange rate and a lower money stock than originally. This result is compatible with the conventional wisdom that raising domestic interest rates leads to an exchange rate appreciation and a contraction of monetary aggregates.

In the United Kingdom, in the period 1980-81, the authorities had difficulty in controlling the growth of the money stock within the limits stated by their financial strategy. An explanation offered for this phenomenon was that raising interest rates in a recession might increase rather than reduce the demand for loans. Within the present analytical framework, this perverse response would indeed lead to the observed result: BB would shift not to the left but rather to the right; the new equilibrium at S then has a higher rather than a lower money stock than before.

All floating exchange rate regimes, the fixed interest rate regime included, have the property of shielding the economy from external price shocks: the exchange rate moves to offset any movement in PM* to keep domestic prices at their original levels. The fixed interest rate regime has the distinctive property that it also insulates the real economy from internal price shocks. Consider an exogenous increase in the domestic markup, hence in PD. If the exchange rate depreciates so as to increase import prices by the same proportion, a is unaffected by the shock, while M2 grows to accommodate the proportionally higher demand for sterling assets and liabilities. By contrast, fixing M2 or some other monetary aggregate rather than R would mean that, in equilibrium, e would have depreciated less than PD had risen, so that a would increase and real income and output levels would fall.


Figure 4 illustrates equations (24) and (25) in e/R space for a given M2 target. The locus AA, representing equation (24), slopes up as an increase in the interest rate is needed to counterbalance the deflationary effect of an exchange rate appreciation to keep the total demand for sterling assets fixed at the target. The locus BB, representing equation (25), slopes down as a reduction in the interest rate is required to offset the impact of an increase in the exchange rate to maintain the demand for sterling liabilities consistent with the target. Equilibrium is at P.

In this framework, it is clearly important to keep track of the direct effect of the exogenous shock on the demand for sterling liabilities as well as on the demand for sterling assets. For example, a rise in foreign interest rates will reduce the demand for sterling assets but will raise the demand for sterling liabilities, shifting AA up to the left and BB up to the right. There must be a corresponding rise in the sterling interest rate as equilibrium shifts to Q. By contrast, a general rise in foreign income levels, raising foreign wealth and capital stock targets, will raise foreign demand for both sterling assets and liabilities, shifting both AA and BB to the right. The result is an appreciation in the exchange rate with an indeterminate and small effect on the sterling interest rate as equilibrium shifts to S.

A variety of other monetary aggregates may also be used as the target for monetary policy, including the narrow money stock, M1, the domestic component of broad money, M2(−), which excludes foreign holdings of domestic deposits; and domestic credit, DC, the domestic component of sterling lending. The choice between alternative targets has consequences for the response of the economic system to different types of exogenous shock. Each of these may be analyzed after appropriate rearrangement of equations (24) and (25).

Consider a favorable income shock, such as, for example, the exploitation of a newly discovered oil reserve, which shifts the functions y (α,.) and f (α, .) upward. Under all four regimes, the exchange rate appreciates to restore monetary equilibrium at the original monetary target; this appreciation reduces domestic competitiveness and, hence, partially neutralizes the original shift’s positive effect on real income. However, the extent of this appreciation is less under the M2 target regime than under any of the other three. Under the former regime, the aggregates M1, M2 (−), and DC may all increase as the exchange rate rises, as this appreciation reduces foreign demand for sterling assets and liabilities. Hence, the squeeze on the non-oil economy that is required to restore equilibrium is rather less.

The effects of various financial shocks may be qualitatively as well as quantitatively different under the alternative regimes:

(1) A shift in domestic portfolio preferences from narrow money to quasi-money has no effect on equilibrium under M2(−), M2, or DC target regimes. However, with an M1 target, an exchange rate depreciation and a fall in the interest rate are required to restore balance.

(2) A shift in foreign portfolio preferences toward sterling assets under a fixed M2 or DC regime leads to a fall in the domestic interest rate and an exchange rate depreciation; under a fixed M1 or M2(−) regime, the same shift leads to a fall in the domestic interest rate with an exchange rate appreciation.

VI. Conclusion

Recent institutional developments, particularly in developed economies, have led to increasingly competitive money markets dominated by the adjustment of price rather than quantity. The theme of this paper is that in such a situation monetary aggregates, interest rates, and the exchange rate are jointly determined, and it is then essential to pay as much attention to modeling the supply of money as to the demand. Otherwise, the analysis of monetary policy is sometimes misleading and at best incomplete.

If an interest rate on money balances is simply introduced into a standard model of asset-market equilibrium, it appears that the authorities have two degrees of freedom, being able to set two from the set of monetary aggregates, interest rates, and the exchange rate; this conclusion is at odds with the conventional wisdom on the limits to sustainable monetary policy. But the money market-clearing condition—that the demand for monetary assets be compatible with the demand for bank borrowing—provides an additional restraint on the power of monetary policy. In the long run, in which nominal prices are perfectly flexible and stocks are at their target values, the authorities may control the price level but not real activity levels, relative prices, or real interest rates in monetary equilibrium, unless they are prepared to manipulate the behavior of the banking system. In the short run in which nominal prices are sticky, the authorities may exert some influence over the real economy through their choice of monetary policy, but this policy cannot hope to achieve more than one independent monetary target without altering bank behavior.

The analysis contained in the body of this paper is also intended to bring out some of the comparative-static implications of models that treat the liability as well as the asset side of the money market. These include: (1) the conventional view that raising interest rates leads to an exchange rate appreciation and a contraction of the money stock rests on (i) the demand for money balances being more responsive to shifts in the exchange rate than the demand for bank borrowing, and (ii) the substitution effect of an increase in lending rates on bank borrowing outweighing the income effect; (2) the qualitative effect of a financial shock on monetary equilibrium depends on (i) the impact of the shock on the demand for bank borrowing as well as on the demand for money balances, and (ii) the precise choice of monetary target.


The Macroeconomic Effects of Changes in Barriers to Trade and Capital Flows: A Simulation Analysis—MOHSIN S. KHAN and ROBERTO ZAHLER

(pages 223-82)

The liberalization of foreign trade and capital movements, or what is commonly termed opening up the economy, is a subject of increasing interest to developing countries, including many in Latin America. While the longer-run costs and benefits of such a strategy may be well known, there has been little systematic study of the short-run effects on the economy as it changes from a relatively closed system to a more outward-oriented one. Analysis of some of the short-term to medium-term economic effects of reducing barriers to trade and capital flows is the basic focus of this paper.

The short-term issues related to opening up are handled within the framework of a dynamic general-equilibrium model constructed specifically for this purpose. This model, which concentrates on the determination of major macroeconomic variables, has its roots in both theoretical and empirical general-equilibrium models, as well as in the monetary-oriented models that have been developed to study the macroeconomics of open economies. At the same time, care is taken to explicitly incorporate certain phenomena that have emerged in country experiences with liberalization, such as persistent deviations from the law of one price and from interest rate parity. This approach makes the resulting model more representative of a “typical” Latin American country, although it should be stressed that the model is quite general and therefore is not entirely applicable to a particular country. Dynamics are introduced to enable one to trace the transition paths of variables—such as economic activity, employment, prices, the balance of payments, external indebtedness, and interest rates—from one equilibrium position to another.

Using a representative set of parameter values, the model simulates various types of opening-up strategy, including (1) gradual and sudden reduction in tariffs, (2) gradual and sudden removal of restrictions on capital flows, (3) simultaneous gradual removal of both of these barriers, and (4) gradual removal of both types of barrier in different sequences. It is believed that such experiments cover most types of policy actually adopted by those developing countries that have sought to open up their economies. Briefly, the results indicate that in all cases some transitory costs are incurred, including reductions in economic growth and employment, the emergence of current account deficits, and reductions in government revenues. The overall exercise yielded the following important insights. First, the effects of liberalizing foreign trade differ markedly from the effects of liberalizing capital flows, and the effects of a simultaneous liberalization of both are not the same as the effects of implementing the two liberalization policies separately. Second, the speed with which the reforms are instituted is important. As expected, a shock type of approach has a more pronounced impact at the beginning, but the adjustment to equilibrium is faster. Finally, and perhaps most interesting, it is shown that the choice of the type of policy to be implemented first is not a matter of indifference. Clear trade-offs emerge, allowing policymakers to choose particular strategies that are consistent with their own overall perspectives.

The Underground Economy in the United States: Annual Estimates, 1930-80— VITO TANZI (pages 283-305)

This article presents yearly estimates for the underground economy in the United States for the period 1930-80. The method used for calculation is one developed a few years ago by the author, and it has been used in many other countries. The main conclusions can be summarized as follows: (1) in 1980 the underground economy, expressed as a percentage of gross national product (GNP), was somewhere between 4.5 percent and 6.1 percent; (2) the only other period in which it might have been higher than that was during World War II; (3) it has been increasing since the mid-1960s; and (4) over the period 1965-80, it has grown by more than 2 percentage points of GNP (by almost 50 percent). The recent trend is disturbing, as it seems to have accelerated in recent years, especially since the mid-1970s. This trend was probably influenced by the substantial increase in marginal tax rates over the period 1975-80, caused by inflation and the absence of significant tax cuts. It is not possible at this point to speculate on whether the tax cuts enacted in 1981 have reversed the trend.

The results obtained in this paper should not be taken as precise measures of the underground economy; they are, at best, broad indications of trends and of orders of magnitude because they are sensitive to the assumptions made as well as to data revisions. However, it is comforting to realize that for 1974 and 1976 they are of the same order of magnitude as direct estimates by Simon and Witte (1980) and by the Internal Revenue Service (1979).

A word on what has been measured is necessary. The estimates attempt to measure the incomes that were associated with the excessive use of cash and that presumably were not reported to the tax authorities. Whether these incomes were or were not measured by the national accounts authorities cannot be determined. Presumably, part of these incomes not only evaded the tax net but also may have escaped the attention of the national accounts authorities—but, how large this part was cannot be assessed with the information at hand. Should this part be large, it would have serious implications for the conduct of economic policy that is based largely on changes in economic activity as reflected by the national accounts.

Measurement of the Public Sector Deficit and Its Implications for Policy Evaluation and Designwillem h. buiter (pages 306-49)

This paper studies budgetary, financial, and monetary policy evaluation and design using a comprehensive wealth or permanent income accounting framework. A set of stylized balance sheets and permanent income accounts is constructed for the public, private, and overseas sectors. These are then contrasted with the conventionally measured balance sheet and flow of funds accounts. This permits a new look at the issues of “crowding out” and the “eventual monetization of fiscal deficits.”

The conventionally measured public sector financial surplus, even when evaluated at constant prices or as a proportion of gross national product, presents a potentially misleading picture of the change in the real net worth of the public sector. One reason is that capital gains and losses on outstanding stocks of marketable financial assets and liabilities are not included in the flow of funds. This omission includes changes in the real value of public sector debt that are due to inflation and to exchange rate changes. A second reason is the omission of revaluations in nonmarketable (and often merely implicit) assets and liabilities, such as the expected future streams of tax receipts and of benefit payments. The need to distinguish between transitory (e.g., cyclical) and permanent (or structural) changes in the deficit and between current account and capital account deficits is reiterated.

The paper than proposes some general rules for the design of stabilization policy—measures to facilitate expenditure smoothing by avoiding or minimizing the incidence of capital market imperfections. Both national governments and international agencies should design fiscal, financial, and budgetary policies so as to induce an evolution of the conventionally measured balance sheet and flow of funds accounts that permits private agents and national economies, respectively, to approximate the behavior that would be adopted if comprehensive wealth or permanent income were the only binding constraint on economic behavior. This goal can be achieved by keeping disposable income in line with permanent income and by ensuring an adequate share of disposable financial wealth in total wealth.

Estimating Models of Financial Market Behavior During Periods of Extensive Structural Reform: The Experience of Chile—DONALD J. MATHIESON (pages 350-93)

This paper examines the problems involved in estimating portfolio balance models of the behavior of financial markets during a period of extensive structural reforms. The model encompasses specifications of the portfolio behavior of banks and nonbanks (firms and households) regarding holdings of currency, demand deposits, time deposits, and bank loans. In addition, the analysis describes the determinants of bank loan and time deposit rates and the linkages between inflation, the balance of payments, interest rates, and financial aggregates. The model is estimated for a sample of monthly Chilean data from July 1965 to June 1980. The estimation results are also contrasted with those obtained in previous work on Argentina.

Applied General-Equilibrium Tax Modeling—JOHN B. SHOVEN (pages 394-420)

This paper summarizes the progress that has been made in computational general-equilibrium (CGE) models designed for tax policy evaluation. After reviewing the structure of CGE models, the paper discusses the inclusion of taxes, foreign trade, and financial behavior. The data requirements and use are examined, as is the measure of economic welfare used for efficiency comparisons.

The U.S. model (constructed with the assistance of the U.S. Treasury Department) is reviewed, and results are presented for both the integration of the corporation and personal income taxes and the adoption of a personal consumption tax instead of the current income tax. Primary results indicate that the present value of the efficiency gain from integrating the two tax systems is $350-450 billion (in 1973 dollars), between 0.7 and 0.9 percent of the present value of national income plus leisure; the gain from moving to a consumption tax is larger, amounting to roughly $630 billion (1973 dollars), 1.25 percent. The paper also asks how long the long run is for such a policy change as the adoption of the consumption tax. The model suggests that it takes 30-40 years to adjust fully to the new tax program, indicating that policies designed to increase real wages by taxing saving more lightly to increase capital/labor ratios may take a long time to produce the desired effect.

The paper reports estimates of the total efficiency losses caused by the entire tax system. The hypothetical experiment that was involved replaces the existing set of federal, state, and local taxes with an equal-yield set of nondistortionary lump-sum levies. The estimated total losses of the existing system are 10 percent of the present value of future national income and 6.7 percent of the present value of national income plus leisure. These estimates are significantly higher than those previously published for individual components of the overall tax system.

The final results for the U.S. general-equilibrium model give an estimate of how much efficiency losses increase per dollar of increased tax receipts. The paper asserts that if all taxes are raised proportionally, the deadweight loss increases between $0.52 and $0.76 per additional dollar raised, implying that the private sector loses between $1.52 and $1.76 for each dollar raised. These results indicate that considerations of tax efficiency have important implications for the economy’s allocation of resources. The paper also briefly discusses the implications that tax efficiency considerations have for cost/benefit studies of governmental expenditure programs.

The paper concludes by asserting that applied general-equilibrium analysis is now sufficiently well developed to be a useful tool for policy analysis.

On the Monetary Analysis of an Open Economy—CHARLES COLLYNS

(pages 421-44)

Recent institutional developments in developed economies have led to increasingly competitive money markets that are dominated by price rather than quantity adjustment. In this situation, monetary aggregates, monetary interest rates, and the exchange rate are jointly determined. But, if an interest rate on money balances is simply introduced into a standard asset-market model of monetary equilibrium, it then appears that the authorities have two degrees of freedom, being able to set two from the set of monetary aggregates, interest rates, and the exchange rate. This conclusion is at odds with conventional views on the limits to sustainable monetary policy. This paper reconciles this discrepancy by requiring that, in equilibrium, the demand for bank loans be compatible with the demand for monetary assets. This procedure requires an explicit treatment of the money supply process, including the behavior of the banking system and of the demand for credit. In the context of this extended model, the monetary authorities can achieve only one independent monetary target. In the long run, monetary policy does not affect real variables, unless the authorities manipulate the behavioral characteristics of the banking system; in the short run, in which nominal prices are sticky, the choice of monetary policy does have real effects. The comparative-static implications of policy choices and exogenous shocks are seen to depend on the relative force and direction of their impact on the demand for bank borrowing as well as for monetary assets.


Effets macroéconomiques des modifications apportées aux obstacles aux échanges commerciaux et aux mouvements de capitaux : analyse de simulation—MOHSIN S. KHAN et ROBERTO ZAHLER (pages 223-82)

La libéralisation du commerce extérieur et des mouvements de capitaux, ou ce qu’il est encore communément convenu d’appeler l’“ouverture” de l’économie, est un sujet qui suscite un intérêt croissant dans les pays en développement, notamment dans un grand nombre de pays d’Amérique latine. S’il est vrai que les coûts et les avantages à long terme de cette stratégie sont sans doute bien connus, peu d’études systématiques de ses effets à court terme sur l’économie, à mesure que celle-ci passe d’un système relativement fermé à un système davantage axé sur l’extérieur, ont été entreprises. Le présent document est essentiellement consacré à l’analyse de certains effets économiques à court et moyen termes de la réduction des obstacles aux échanges commerciaux et aux mouvements de capitaux.

Les questions à court terme liées à l’ouverture de l’économie sont traitées dans le cadre d’un modèle dynamique d’équilibre général construit spécifiquement à cette fin. Ce modèle, qui se concentre sur la détermination des principales variables macroéconomiques, trouve ses racines dans des modèles d’équilibre général, tant théoriques qu’empiriques, ainsi que dans les modèles reposant davantage sur la théorie monétaire qui ont été conçus pour l’étude macroéconomique des économies ouvertes. Dans le même temps, les auteurs ont pris soin d’incorporer explicitement certains phénomènes qui se sont présentés dans plusieurs pays ayant fait l’expérience de la libéralisation, comme, par exemple, des écarts persistants par rapport à la loi du prix unique du marché et par rapport à la parité des taux d’intérêt. Cette méthode rend le modèle qui en résulte plus représentatif d’un pays “typique” d’Amérique latine, encore qu’il faille souligner que le modèle est plutôt général et, par conséquent, n’est pas entièrement applicable à un pays particulier. Sa dynamisation est introduite pour permettre de déterminer la trajectoire de transition suivie par les variables—telles que l’activité économique, l’emploi, les prix, la balance des paiements, la dette extérieure et les taux d’intérêt—lorsque l’économie passe d’une position d’équilibre à une autre.

Utilisant un ensemble représentatif des valeurs des paramètres, le modèle simule différents types de stratégies d’ouverture, parmi lesquelles : 1) une réduction progressive ou brusque du tarif douanier; 2) une suppression progressive ou brusque des restrictions aux mouvements de capitaux; 3) l’élimination progressive simultanée de ces deux types d’obstacles; et 4) l’élimination progressive en séquences de ces deux types d’obstacles. On estime que ces expériences recouvrent la plupart des types de mesures effectivement adoptées par les pays en développement qui ont cherché à ouvrir leur économie. En bref, les résultats montrent que, dans tous les cas, le processus entraîne des coûts temporaires, parmi lesquels une diminution de la croissance économique et de l’emploi, l’apparition de déficits extérieurs courants et une baisse des recettes publiques. Dans son ensemble, l’étude fournit les indications importantes suivantes. Premièrement, les effets d’une libéralisation des échanges commerciaux avec l’extérieur sont très différents des effets d’une libéralisation des mouvements de capitaux et les effets d’une libéralisation simultanée des uns et des autres sont différents des effets de politiques de libéralisation mises en oeuvre séparément dans ces deux domaines. Deuxièmement, la rapidité avec laquelle les réformes sont mises en oeuvre est importante. Comme on peut s’y attendre, une méthode de “choc” a une incidence plus prononcée dans un premier temps, mais le retour à l’équilibre se fait plus vite. Enfin, et c’est peut-être la chose la plus intéressante, les auteurs montrent que le choix du type de politique à mettre en oeuvre en premier lieu n’est pas indifférent. Des arbitrages apparaissent nettement qui permettent aux responsables de la politique économique d’opter pour les stratégies qui sont compatibles avec leurs propres perspectives générales.

L’économie clandestine aux Etats-Unis : estimations annuelles, 1930-80 —VITO TANZI (pages 283-305)

Le présent article fournit des estimations annuelles relatives à l’économie clandestine aux Etats-Unis pour la période 1930-80. La méthode utilisée pour effectuer les calculs a été mise au point par l’auteur il y a quelques années et a été appliquée dans de nombreux autres pays. Les conclusions principales peuvent être résumées comme suit : 1) en 1980, l’économie clandestine, exprimée en pourcentage du produit national brut (PNB), s’est située entre 4,5 % et 6,1 %; 2) la seule période pendant laquelle elle a peut-être représenté un pourcentage plus élevé est celle de la deuxième guerre mondiale; 3) la part de l’économie clandestine a augmenté depuis le milieu des années 60; et 4) au cours de la période 1965-80, elle a augmenté, en proportion du PNB, de plus de deux points de pourcentage (soit de près de 50 %). La tendance enregistrée récemment est préoccupante, étant donné que l’économie clandestine semble se développer à un rythme croissant, surtout depuis le milieu des années 70. Cette tendance a probablement été influencée par l’augmentation considérable des taux marginaux d’imposition au cours de la période 1975-80, augmentation due à l’inflation et à l’absence de réduction sensible des impôts. Il n’est pas possible à ce stade de déterminer si les réductions d’impôts adoptées en 1981 ont renversé ou non la tendance.

Les résultats figurant dans la présente étude ne doivent pas être considérés comme des mesures précises de l’économie clandestine; ils ne donnent, au mieux, que des indications générales des tendances et des ordres de grandeur parce qu’ils sont sensibles aux hypothèses retenues ainsi qu aux révisions des données. Toutefois, il est rassurant de constater que, pour 1974 et 1976, les chiffres sont du même ordre de grandeur que les estimations directes effectuées par Simon et Witte en 1980, et par l’administration fiscale des Etats-Unis (Internai Revenue Service) en 1979.

Il convient de préciser brièvement les éléments qui ont été mesurés. Les estimations cherchent à évaluer les revenus liés à l’utilisation excessive des règlements en espèces et qui n’ont probablement pas été déclarés à l’administration des impôts. Il n’est pas possible de déterminer si ces revenus ont ou non été évalués par les services responsables de la comptabilité nationale. Selon toute vraisemblance, une partie de ces revenus a non seulement échappé à l’impôt mais, en outre, a échappé à l’attention des services-de la comptabilité nationale; toutefois, il n’est pas possible d’évaluer, sur la base des renseignements disponibles, l’ampleur de cette partie des revenus. Si elle est effectivement importante, elle pourrait avoir des effets graves sur la conduite de la politique économique qui se fonde essentiellement sur les variations de l’activité économique telles qu’elles apparaissent dans la comptabilité nationale.

Mesure du déficit du secteur public et ses implications pour l’évaluation et l’élaboration de la politique économique — WILLEM H. BUITER (pages 306-49)

La présente étude analyse l’évaluation et l’élaboration des politiques budgétaire, financière et monétaire à partir d’un cadre comptable reposant sur le patrimoine global ou sur le revenu permanent. Un ensemble de comptes types de patrimoine et du revenu permanent est mis au point pour les secteurs public, privé et le reste du monde. Ces comptes sont ensuite comparés aux comptes de patrimoine et aux comptes d’opérations financières établis selon les méthodes traditionnelles. Cette comparaison jette un jour nouveau sur les questions d’envahissement du marché financier par le secteur public et de monétisation finale des déficits budgétaires.

L’excédent financier du secteur public, lorsqu’il est mesuré selon les méthodes traditionnelles, même s’il est calculé à prix constants ou en proportion du produit national brut, donne une idée de la variation de la valeur réelle nette du patrimoine du secteur public qui risque d’induire en erreur. L’une des raisons en est que les gains et pertes en capital sur les stocks existants d’actifs et de passifs financiers négociables ne sont pas inclus dans les opérations financières. Cette omission porte aussi sur les variations de la valeur réelle de la dette du secteur public dues à l’inflation et aux fluctuations de taux de change. Une autre raison est que sont également omises les réévaluations des actifs et des passifs non négociables (et souvent purement implicites), tels que les flux futurs actualisés de recettes fiscales et de dépenses de transferts. L’auteur réitère la nécessité d’effectuer une distinction entre variations temporaires (par exemple cycliques) et variations permanentes (ou structurelles) du déficit et entre déficits des opérations courantes et déficits des opérations en capital.

L’auteur propose ensuite quelques règles générales pour l’élaboration de mesures de stabilisation destinées à atténuer les variations des dépenses en évitant ou en minimisant l’incidence des imperfections du marché des capitaux. Les gouvernements comme les institutions internationales devraient mettre au point des politiques fiscales, financières et budgétaires propres à susciter une évolution des comptes de patrimoine et des tableaux d’opérations financières établis selon les méthodes traditionnelles, de façon à permettre aux agents économiques privés et aux économies nationales, respectivement, d’approcher le comportement qui serait adopté si le patrimoine global ou le revenu permanent était la seule limite contraignante au comportement économique. Cet objectif peut être atteint si le revenu disponible suit le revenu permanent et en rendant disponible une part suffisante du patrimoine financier à l’intérieur du patrimoine total.

Estimation de modèles du comportement du marché financier en périodes de réformes structurelles généralisées : l’expérience du Chili—DONALD J. MATHIESON (pages 350-93)

L’auteur de la présente étude analyse les difficultés que présente l’estimation de modèles du comportement du marché financier—axés sur l’équilibre des portefeuilles—en périodes de réformes structurelles généralisées. Le modèle spécifie le comportement en matière de portefeuilles des banques et des agents non bancaires (entreprises et ménages) au niveau des avoirs en monnaie, des dépôts à vue, des dépôts à terme et des prêts bancaires. L’analyse comporte également une description des facteurs déterminants des taux servis sur les prêts bancaires et les dépôts à terme ainsi que des liens existant entre l’inflation, la balance des paiements, les taux d’intérêt et les agrégats financiers. Le modèle est estimé au moyen d’un échantillon de données mensuelles observées au Chili pendant la période allant de juillet 1965 à juin 1980. Enfin, l’auteur compare les résultats à ceux obtenus dans le cadre d’une étude effectuée antérieurement pour l’Argentine.

Modèles d’équilibre général appliqués incorporant la fiscalité — JOHN B. SHOVEN (pages 394-420)

L’auteur fait état des progrès réalisés dans le domaine des modèles informatiques d’équilibre général conçus pour évaluer les politiques fiscales. Après avoir examiné la structure de ces modèles, il traite de la prise en compte, dans ces modèles, des impôts, du commerce extérieur et des comportements financiers. Il passe en revue les données qu’il convient de rassembler, l’utilisation qui en est faite, ainsi que la façon dont on mesure le bien-être économique pour effectuer des comparaisons d’efficacité.

L’auteur analyse le modèle élaboré pour les Etats-Unis (en collaboration avec le Ministère des finances des Etats-Unis) et présente les résultats donnés par la prise en compte de l’impôt sur les sociétés et de l’impôt sur le revenu des personnes physiques ainsi que par le remplacement de l’impôt actuel sur le revenu par un impôt sur la consommation des personnes physiques. Les premiers résultats indiquent que la valeur actuelle du gain d’efficacité dû à l’intégration des deux régimes d’imposition est d’environ 350-450 milliards de dollars E.U. (en dollars de 1973), soit entre 0,7 et 0,9 % de la valeur actuelle du revenu national plus les loisirs; le gain résultant de l’adoption d’un impôt sur la consommation est plus élevé puisqu’il est de l’ordre de 630 milliards de dollars E.U. (en dollars de 1973), soit 1,25 %. L’auteur cherche également à quantifier l’horizon temporel d’un changement de politique économique tel que l’adoption d’un impôt sur la consommation. D’après le modèle, il faut de 30 à 40 ans pour que l’économie puisse s’adapter complètement au nouveau programme fiscal, ce qui signifie que les mesures prises pour accroître les salaires réels en allégeant l’imposition de l’épargne de manière à accroître les ratios capital/travail peuvent ne produire l’effet désiré qu’après une longue période.

L’auteur donne un certain nombre d’estimations concernant les pertes d’efficacité causées par l’ensemble du système fiscal. L’auteur fait fonctionner le modèle en choisissant, par hypothèse, de remplacer les impôts fédéraux, d’Etat et locaux actuellement en vigueur par une série de prélèvements forfaitaires qui ont le même rendement et ne produisent aucune distorsion. Les pertes totales sont estimées, pour l’ensemble du système fiscal existant, à 10 % de la valeur actuelle du revenu national futur et à 6,7 % de la valeur actuelle du revenu national plus les loisirs. Ces estimations sont significativement supérieures à celles qui ont été publiées antérieurement pour les diverses composantes de l’ensemble du système fiscal.

Les résultats finals du modèle d’équilibre général pour les Etats-Unis fournissent une estimation de la proportion dans laquelle les pertes d’efficacité augmentent lorsque les recettes fiscales progressent d’un dollar. L’auteur affirme que, si tous les impôts sont relevés dans les mêmes proportions, la perte d’efficacité (“deadweight loss”) augmente d’un montant compris entre 52 et 76 cents des Etats-Unis pour chaque dollar de recette supplémentaire, ce qui signifie que le secteur privé perd de 1,52 à 1,76 dollar E.U. pour chaque dollar perçu au titre de l’impôt. Ces résultats montrent que les considérations relatives à l’efficacité de l’impôt ont d’importantes répercussions sur l’allocation des ressources au sein de l’économie. L’auteur examine ensuite brièvement l’implication que ces considérations d’efficacité de l’impôt ont pour les analyses coûts/avantages des programmes de dépenses publiques.

L’auteur conclut en affirmant que les modèles appliqués d’équilibre général sont maintenant suffisamment au point et constituent un instrument utile aux fins de l’analyse des divers aspects de la politique économique.

Analyse monétaire d’une économie ouverte— CHARLES COLLYNS (pages 421-44)

Certaines mutations institutionnelles récentes au sein des économies développées se sont traduites par des marchés monétaires de plus en plus compétitifs, dominés par des ajustements de prix plutôt que par des ajustements de quantité. Dans une telle situation, les agrégats monétaires, les taux d’intérêt monétaires et les taux de change sont déterminés conjointement. Mais, si un taux d’intérêt sur les encaisses est simplement introduit dans un modèle typique, fondé sur le marché des avoirs d’équilibre monétaire, il apparaît que les autorités ont deux degrés de liberté puisqu’elles sont à même de définir deux éléments de l’ensemble constitué par les agrégats monétaires, les taux d’intérêt et le taux de change. Cette conclusion est en contradiction avec l’opinion traditionnelle concernant les limites d’une politique monétaire soutenable. Le présent document concilie ces points de vue divergents en posant comme condition que, en situation d’équilibre, la demande de prêts bancaires soit compatible avec la demande d’avoirs monétaires. Cette procédure requiert un traitement explicite du processus d’offre de monnaie, et notamment du comportement du système bancaire et de la demande de crédit. Dans le contexte de ce modèle élargi, les autorités monétaires ne peuvent atteindre qu’un objectif monétaire indépendant. A long terme, la politique monétaire n’affecte pas les variables réelles à moins que les autorités ne manipulent les caractéristiques du comportement du système bancaire; à court terme, étant donné la viscosité des prix nominaux, le choix de la politique monétaire a des effets réels. On constate que les implications, sur les positions statiques comparées, des choix de politique économique et des chocs exogènes dépendent de la force relative et du sens de leur impact sur la demande de prêts bancaires et d’avoirs monétaires.


Efectos macroeconómicos de cambios en las barreras al comercio y al movimientos de capitales: Un modelo de simulación—MOHSIN S. KHAN Y ROBERTO ZAHLER (páginas 223-82)

La liberalización de los intercambios comerciales y de los movimientos de capital, llamada comúnmente la apertura de la economía, es un tema de creciente interés para los países en desarrollo, incluidos muchos países de América Latina. Aunque los costos y beneficios a largo plazo de esta estrategia son bien conocidos, se han hecho pocos estudios sistemáticos sobre los efectos a corto plazo en la economía, al pasar ésta de un sistema relativamente cerrado a un sistema orientado hacia el exterior. Este trabajo se centra en el análisis de algunos efectos económicos a corto y a mediano plazo de la reducción de barreras al comercio y a los movimientos de capital.

Las cuestiones de corto plazo relacionadas con la apertura se estudian en el marco de un modelo de equilibrio general dinámico elaborado específicamente para este fin. Este modelo, centrado en la determinación de las principales variables macroeconómicas, se basa en modelos teóricos y empíricos de equilibrio general, así como en modelos de orientación más monetaria que se han elaborado para estudiar la macroeconomía de una economía abierta. Al mismo tiempo, el trabajo se ha ocupado de incorporar explícitamente determinados fenómenos ocurridos en los países en que se ha llevado a cabo un proceso de liberalización, como es el alejamiento continuo de la ley del precio único y de la paridad del tipo de interés. Este enfoque hace que el modelo resultante sea más representativo de un país “típico” de América Latina, aunque debe señalarse que es bastante general y, por lo tanto, no es aplicable exactamente a un país determinado. Se introduce en el modelo la dinámica para que se pueda trazar la senda de transición de las variables—actividad económica, empleo, precios, balanza de pagos, deuda externa y tipos de interés—desde una posición de equilibrio a otra.

Utilizando un conjunto representativo de valores de los parámetros, se simulan en el modelo diversos tipos de estrategia de apertura, incluyendo: 1) reducciones graduales y repentinas de los aranceles, 2) abolición gradual y repentina de las restricciones a los movimientos de capital, 3) abolición gradual simultánea de ambas clases de barreras y 4) abolición gradual de ambas clases de barreras en diferentes secuencias. Se considera que tales experimentos abarcan la mayoría de las políticas adoptadas por los países en desarrollo que han intentado abrir sus economías. En resumen, los resultados señalan que en todos los casos existen algunos costos transitorios, como una reducción del crecimiento económico y del empleo, la aparición de déficit en cuenta corriente y reducciones de los ingresos públicos. El trabajo global produjo las siguientes aclaraciones importantes. Primero, los efectos de una liberalización de los intercambios comerciales pueden ser muy diferentes de los efectos de una liberalización de los movimientos de capital, y los efectos de una liberalización simultánea de ambos no son los mismos que los efectos producidos si se ejecutan ambas políticas de liberalización en forma separada. Segundo, es importante el ritmo de aplicación de las reformas. Como cabía prever, un método de choque tiene un efecto más pronunciado al inicio, pero el ajuste al equilibrio es más rápido. Por último, y quizás esto sea lo más interesante, se demuestra que no es indiferente la elección del tipo de medidas de política a aplicar en primer lugar. Surgen relaciones de correspondencia claras, lo que permite que quienes deciden las medidas de política puedan elegir estrategias determinadas compatibles con sus propias perspectivas globales.

La “economía subterránea” de Estados Unidos: Estimaciones anuales, 1930-80 — VITO TANZI (páginas 283-305)

En este estudio se presentan estimaciones anuales de la “economía subterránea” de Estados Unidos durante el período 1930-80. El método empleado para el cálculo es el ideado hace unos años por el autor y se ha aplicado ya en muchos otros países. Las conclusiones principales serían las siguientes: 1) en 1980 la “economía subterránea”, medida en porcentaje del producto nacional bruto (PNB), se situó entre 4,5 por ciento y 6,1 por ciento; 2) el otro período en que podría haber sido superior fue el correspondiente a la segunda guerra mundial; 3) “la economía subterránea” ha venido creciendo desde mediados del decenio de los sesenta, y 4) durante el período 1965-80 ha aumentado más de 2 puntos porcentuales del PNB (casi 50 por ciento). Esta última tendencia es alarmante ya que parece haberse acelerado en los últimos años, especialmente a partir de mediados del decenio de los setenta. Probablemente esta propensión se haya visto influida por el gran aumento de las tasas impositivas marginales durante el período 1975-80 causado por la inflación y la falta de reducciones tributarias significativas. De momento no se puede saber si las reducciones tributarias de 1981 habrán invertido la tendencia.

Los resultados obtenidos en este artículo no deben entenderse como medidas precisas de la “economía subterránea”; en el mejor de los casos son una indicación general de tendencias y de órdenes de magnitud que a su vez están influenciados por los supuestos utilizados y posibles revisiones de datos. Sin embargo, resulta alentador comprobar que para 1974 y 1976 son del mismo orden de magnitud que las estimaciones directas de Simon y Witte (1980) y de la Dirección de Impuestos de Estados Unidos (1979).

Conviene puntualizar qué ha sido objeto de medición en este estudio. Las estimaciones tratan de cuantificar ingresos asociados al uso excesivo de dinero en efectivo y que puede suponerse no fueron declarados a las autoridades tributarias. No puede determinarse si estos ingresos han sido o no contabilizados por las autoridades encargadas de las cuentas nacionales. Cabe suponer que parte de estos ingresos no sólo evadieron ilegalmente la red tributaria sino que además escaparon de la atención de las autoridades a cargo de las cuentas nacionales. Sin embargo, con la información disponible, no es posible saber de qué porcentaje se trata. Si este porcentaje de ingresos es importante, tendría serias consecuencias sobre la aplicación de la política económica, ya que ésta se basa en gran medida en las variaciones de la actividad económica tal como se reflejan en las cuentas nacionales.

Medición del déficit del sector público e implicaciones para la evaluación y formulación de ciertas políticas—WILLEM H. BUITER (páginas 306-49)

En este trabajo se estudian la evaluación y la formulación de las políticas presupuestaria, financiera y monetaria en un marco conceptual de riqueza en sentido amplio y contabilidad de la renta permanente. Se elabora un conjunto de balances generales y cuentas de renta permanente para los sectores público, privado y exterior. Luego se cotejan con el balance general medido de forma convencional y con las cuentas de flujos de fondos. Esto da una nueva visión de los problemas de “desplazamiento o exclusión” y de la “monetización final de los déficit fiscales”.

El superávit financiero del sector público medido de forma convencional, aun cuando se lo evalúe a precios constantes o como proporción del producto nacional bruto, da una idea potencialmente engañosa de la variación del patrimonio real neto del sector público. Uno de los motivos de esto es que las pérdidas y ganancias de capital de la masa existente de activos y pasivos financieros negociables no están incluidas en la corriente de fondos. Esta omisión se extiende también a las variaciones del valor real de la deuda del sector público debidas a la inflación y a las fluctuaciones del tipo de cambio. Un segundo motivo es la omisión de revaluaciones de los activos y pasivos no negociables (y a menudo solamente implícitos), tales como las corrientes futuras previstas de ingresos tributarios y pagos de beneficios. Se reitera la necesidad de establecer una diferencia entre las variaciones transitorias (es decir, cíclicas) y permanentes (o estructurales) del déficit, y entre los déficit en cuenta corriente y en cuenta de capital.

Se proponen luego algunas normas generales para la formulación de medidas de política de estabilización para facilitar la fluidez del gasto evitando o minimizando la incidencia de las imperfecciones del mercado de capitales. Tanto los gobiernos nacionales como los organismos internacionales deberían formular políticas fiscales, financieras y presupuestarias a fin de inducir una evolución del balance general medido de forma convencional y cuentas de flujos de fondos que permitan a los agentes privados y las economías nacionales, respectivamente, acercarse al comportamiento que se adoptaría si la riqueza en sentido amplio o el ingreso permanente fueran la única restricción forzosa al comportamiento económico. Esta meta puede alcanzarse manteniendo el ingreso disponible en armonía con el ingreso permanente y haciendo que la riqueza financiera disponible constituya una proporción adecuada de la riqueza total.

La estimación de modelos de comportamiento del mercado financiero durante períodos de reforma estructural amplia: La experiencia de Chile—DONALD J. MATHIESON (páginas 350-93)

En este estudio se abordan los problemas de la estimación de modelos de equilibrio de cartera del comportamiento de los mercados financieros durante un período de amplias reformas estructurales. El modelo incluye especificaciones del comportamiento de cartera de la banca y del sector no bancario (empresas y unidades familiares) en cuanto a las tenencias de moneda, depósitos a la vista, depósitos a plazo y préstamos bancarios. También se describen los factores determinantes de los tipos de interés de los préstamos bancarios y los depósitos a plazo, y los vínculos entre inflación, balanza de pagos, tipos de interés y agregados financieros. Se estima el modelo para una muestra de datos mensuales chilenos que abarca desde julio de 1965 hasta junio de 1980. Se comparan también los resultados de la estimación con los obtenidos en un trabajo anterior sobre Argentina.

Modelos tributarios aplicados de equilibrio general—JOHN B. SHOVEN

(páginas 394-420)

En el estudio se resumen los avances logrados en modelos de cálculo de equilibrio general concebidos para evaluar la política tributaria. Tras examinar la estructura de estos modelos, se estudia la inclusión de los impuestos, el comercio exterior y el comportamiento financiero. También se estudian las necesidades y empleo de datos, así como la cuantificación del bienestar económico utilizado para comparaciones de eficiencia.

Se estudia el modelo de Estados Unidos (elaborado con la ayuda del Departamento del Tesoro) y se presentan los resultados tanto para la integración de los impuestos sobre la renta de las personas físicas y de sociedades como para la adopción de un impuesto personal sobre el consumo que sustituya el actual impuesto sobre la renta. Los resultados primarios indican que el valor actual de la mayor eficacia conseguida al integrar los dos sistemas tributarios es de $350.000 millones a $450.000 millones (dólares de 1973), lo que supone entre 0,7 y 0,9 por ciento del valor actual de la renta nacional más el ocio. La ventaja de pasar a un impuesto sobre el consumo es mayor, ascendiendo aproximadamente a $630.000 millones (dólares de 1973), equivalentes a 1,25 por ciento. El autor también se pregunta sobre la duración posible del período de dicha modificación de política, como es la adopción del impuesto sobre el consumo. El modelo indica que llevará de 30 a 40 años el ajuste total al nuevo programa tributario, y que las medidas dirigidas a aumentar los salarios reales mediante una imposición más ligera sobre el ahorro a fin de incrementar los coeficientes capital/trabajo requerirán probablemente mucho tiempo para producir el efecto deseado.

Se ofrecen en el estudio estimaciones de las pérdidas de eficiencia total ocasionadas por el sistema tributario en su totalidad. El experimento hipotético en cuestión sustituye el conjunto actual de impuestos federales, de los estados y locales por una serie de igual rendimiento de gravámenes a tanto alzado no distorsionantes. Las pérdidas totales estimadas del sistema vigente son el 10 por ciento del valor actual del ingreso nacional futuro y 6,7 por ciento del valor actual del ingreso nacional más el ocio. Estas estimaciones son considerablemente más altas que las publicadas anteriormente correspondientes a componentes individuales del sistema tributario global.

Los resultados definitivos del modelo de equilibrio general de Estados Unidos dan una estimación de cuánto aumentan las pérdidas de eficiencia por dólar de incremento de los ingresos tributarios. En el estudio se afirma que si se elevan proporcionalmente todos los impuestos, la pérdida por falta de eficiencia aumentará entre $0,52 y $0,76 por cada dólar adicional recaudado, lo que significa que el sector privado pierde entre $1,52 y $1,76 por cada dólar recaudado. Estos resultados también indican que las consideraciones de eficiencia tributaria tienen importantes repercusiones en la asignación de recursos dentro de la economía. En el estudio también se abordan brevemente las consecuencias que los aspectos de eficiencia tributaria tienen en relación con los estudios de costos-beneficios de los programas de gasto público.

Finalmente se afirma que el análisis de equilibrio general aplicado está ya suficientemente desarrollado para ser un instrumento útil de análisis de política.

Del análisis monetario de una economía abierta—CHARLES COLLYNS

(páginas 421-44)

La reciente evolución institucional de las economías desarrolladas ha dado origen a mercados monetarios cada vez más competitivos dominados por el ajuste de los precios más que por el ajuste cuantitativo. En este contexto, los agregados monetarios, los tipos de interés y el tipo de cambio se determinan conjuntamente. Pero si se introduce simplemente un tipo de interés para los saldos monetarios en un modelo normal de equilibrio monetario del mercado de activos, entonces parecería que las autoridades tendrían dos grados de libertad y podrían fijar dos magnitudes dentro del conjunto formado por los agregados monetarios, los tipos de interés y el tipo de cambio. Esta conclusión difiere de las opiniones convencionales respecto a los límites de una política monetaria sostenible. En el trabajo se neutraliza esa discrepancia exigiendo que, en estado de equilibrio, la demanda de crédito bancario sea compatible con la demanda de activos monetarios. Esta lógica requiere un tratamiento explícito del mecanismo de oferta monetaria, incluyendo el comportamiento del sistema bancario y de la demanda de crédito. En el contexto de este modelo ampliado, las autoridades monetarias sólo pueden alcanzar un objetivo monetario independiente. A largo plazo, la política monetaria no afecta a las variables reales, a menos que las autoridades manipulen las características de comportamiento del sistema bancario; a corto plazo, con precios nominales rígidos, la selección de una política monetaria tiene efectos reales. Se considera que las implicaciones comparativas y estáticas de las diferentes opciones de política y de los choques exógenos dependen de la intensidad relativa y de la dirección de su impacto en la demanda de crédito bancario así como de activos monetarios.

In statistical matter (except in the résumés and resúmenes) throughout this issue,

Dots (…) indicate that data are not available;

A dash (—) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist;

A single dot (.) indicates decimals;

A comma (,) separates thousands and millions;

“Billion” means a thousand million;

A short dash (-) is used between years or months (e.g., 1977-79 or January-October) to indicate a total of the years or months inclusive of the beginning and ending years or months;

A stroke (/) is used between years (e.g., 1978/79) to indicate a fiscal year or a crop year;

Components of tables may not add to totals shown because of rounding.

International Monetary Fund, Washington, D.C. 20431 U.S.A.

Telephone number: 202 477 7000

Cable address: Interfund


  • Branson, William H., “Exchange Rate Dynamics and Monetary Policy,” Ch. 8 in Inflation and Employment in Open Economies, ed. by Assar Lindbeck (Amsterdam, 1979), pp. 189224.

    • Search Google Scholar
    • Export Citation
  • Coats, Warren L., Jr., and Deena R. Khatkhate (eds.), Money and Monetary Policy in Less Developed Countries: A Survey of Issues and Evidence (Oxford and New York, 1980).

    • Search Google Scholar
    • Export Citation
  • Collyns, Charles, “Macroeconomic Policy in the ‘British-Type’ Open Economy” (unpublished doctoral dissertation, Oxford University, 1981).

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rudiger (1976), “Expectations and Exchange Rate Dynamics,” Journal of Political Economy, Vol. 84 (December 1976), pp. 116176.

    • Search Google Scholar
    • Export Citation
  • Dornbusch, Rüdiger (1980), Open Economy Macroeconomics (New York, 1980).

  • Frenkel, Jacob A., and Michael Mussa (1983), “Asset Markets, Exchange Rates and the Balance of Payments: The Reformulation of Doctrine,” in Handbook of International Economics, ed. by Ronald W. Jones and Peter B. Kenen (Amsterdam, 1983).

    • Search Google Scholar
    • Export Citation
  • Kouri, Pentti J.K., “Monetary Policy, the Balance of Payments, and the Exchange Rate,” in The Functioning of Floating Exchange Rates: Theory, Evidence, and Policy Implications, ed. by David Bigman and Teizo Taya (Cambridge, Massachusetts, 1980).

    • Search Google Scholar
    • Export Citation
  • Mundell, Robert A., International Economics (London and New York, 1968).


Mr. Collyns, economist in the Western Hemisphere Department, was in the Central Banking Department when this paper was prepared. It has been developed from the author’s doctoral thesis and has benefited from comments by Mr. Peter Oppenheimer and Mr. Peter Sinclair of Oxford University, as well as colleagues in the Fund.


Dornbusch (1980) and Frenkel and Mussa (1983) provide useful sources for much of this material.


See Collyns (1981, Chs. 6–9) for a more extensive analysis of these and related models.


See, for example, Mundell (1968) and Dornbusch (1976).


See, for example, Branson (1979), Dornbusch (1980, Chs. 10 and 11), and Kouri (1980).


An equivalent approach in a model featuring only high-powered money and bonds would be to introduce the government’s own demand for credit as a dependent variable rather than as a given.


See Collyns (1981, Ch. 7).


See Collyns (1981, Ch. 9).


There is no stock market; ownership titles to firms are not transferable.


The home country’s demand for foreign assets and liabilities is assumed to be negligible in world markets; the foreign interest rate can then be taken as exogenous.


Price stability requires that the nominal interest rate be set equal to the equilibrium real interest rate.


See Collyns (1981, Chs. 7 and 8).