Growth-Oriented Adjustment Programs: Growth-Oriented Adjustment Programs A Conceptual Framework

A conceptual framework is proposed as an input to the design of growth-oriented adjustment programs. The two building blocks of the model are the well-known monetary approach to the balance of payments and a variant of the open-economy neoclassical growth model. The integrated model combines growth, inflation, and the balance of payments and links these objectives to government policies and the availability of foreign financing. The principal advantage of the framework is its simple structure, which enables it to be applied relatively easily to a variety of developing countries. [JEL 113, 430]


A conceptual framework is proposed as an input to the design of growth-oriented adjustment programs. The two building blocks of the model are the well-known monetary approach to the balance of payments and a variant of the open-economy neoclassical growth model. The integrated model combines growth, inflation, and the balance of payments and links these objectives to government policies and the availability of foreign financing. The principal advantage of the framework is its simple structure, which enables it to be applied relatively easily to a variety of developing countries. [JEL 113, 430]

A conceptual framework is proposed as an input to the design of growth-oriented adjustment programs. The two building blocks of the model are the well-known monetary approach to the balance of payments and a variant of the open-economy neoclassical growth model. The integrated model combines growth, inflation, and the balance of payments and links these objectives to government policies and the availability of foreign financing. The principal advantage of the framework is its simple structure, which enables it to be applied relatively easily to a variety of developing countries. [JEL 113, 430]

There is a broad consensus in both academic and policymaking circles that a healthy and sustained rate of economic growth is central to an adjustment strategy intended to achieve long-term viability in the balance of payments and a permanent reduction in the rate of inflation. Indeed, it can be argued that adjustment can only be judged successful if it brings about a rate of growth of output that allows for a steady improvement in per capita income and living standards. Although there may be general agreement on the concept of “growth-oriented adjustment,” however, designing a policy package that will simultaneously eliminate the macroeconomic imbalances in the economy and raise the growth rate turns out to be no easy task. The desire to develop a framework for growth-oriented adjustment has, thus, begun to preoccupy researchers and practitioners alike. Evidence of this interest is reflected in, among others, a growing number of academic papers (see Khan (1987) and the references therein), the recent joint World Bank-International Monetary Fund symposium on the subject of growth-oriented adjustment programs (see Corbo, Goldstein, and Khan (1987)), and, at a more policy-related level, the approach suggested in the report of the Group of Twenty Four (1987).

The need for a consistent framework in which policies can be linked to growth, inflation, and the balance of payments becomes particularly pressing as one moves to make the concept of growth-oriented adjustment operational. There are, of course, several empirical models available that attempt to capture the relationships among the principal objectives of adjustment programs, but these models tend to be either too complicated or too country-specific to apply easily across countries.1 The development of a conceptual framework that incorporates the most important macroeconomic policy instruments and targets, and at the same time can be tailored to the circumstances and the structural characteristics of the individual country, is now clearly one of the main priorities of research on developing countries.

This paper proposes a model that can serve as a starting point for the development of a generalized framework that can analyse issues in adjustment with growth. The two main building blocks of this model are the monetary approach to the balance of payments, used often in designing short-term stabilization programs geared toward balance of payments and inflation targets, and a simple version of the open-economy neoclassical growth model.2 Because our intent is purely expositional—that is, to describe the properties of a model combining two well-known models—there is no pretense here of presenting a “new” macroeconomic model for developing countries.3

The model that is developed has several noteworthy features. First, it combines growth, inflation, and the balance of payments within an integrated framework: unlike the earlier model of Khan, Montiel, and Haque (1986), however, it does not utilize restrictive assumptions about the growth process, foreign debt and debt service, or the endogeneity of exports. Second, the model is able to link these objectives of adjustment directly to a variety of government policies. Third, the model can be used to calculate the effects of changes in external financing on monetary and real variables in the economy. This last feature is of particular importance in present circumstances, in which developing countries are seriously constrained in the amount of foreign resources they can attract. Finally, because the model has a simple structure and requires a minimal amount of information, it can be applied in a fairly straightforward manner to a wide variety of countries. It thus has an operational advantage that more complex models tend to lack.

To analyze the properties of the model, we examine the impact effects on growth, inflation, and the balance of payments of a variety of exogenous and policy-induced shocks. The policy shocks include changes in domestic credit, the exchange rate, fiscal policies, and structural policies that alter the private savings rate and the efficiency of investment. Other shocks considered are changes in the income velocity of money and variations in foreign capital flows. The results obtained from this analysis turn out to be broadly consistent with widely held beliefs.

The remainder of the paper proceeds as follows. In Section I we briefly describe the structure of the model. Section II, which is the key component of the paper, presents comparative-static exercises undertaken to demonstrate the workings of the model. The concluding section summarizes the main points of the paper and highlights some of the directions in which the proposed framework could be extended.

I. A Model of Adjustment and Growth

The model formulated in this section contains a growth block and a monetary block that together determine growth, prices, and the balance of payments. This section begins by describing a simple macroeconomic framework for a representative developing economy. The growth and monetary components of the model are specified separately and are then combined to yield the complete model. This model assumes continuous equilibrium; because the model is specified in discrete time, all adjustments take place in one period.

The Macroeconomic Framework

Consider a small open economy that maintains a fixed exchange rate. In this economy the private sector is assumed to own all factors of production. It receives income from production (Y) and uses it to pay taxes (T), to consume (C), and to save (Sp). The private sector’s budget constraint is therefore:4


Private sector savings are devoted to the accumulation of physical capital,5 to increasing money balances, and to reducing liabilities to the banking system:


where PD is the price of domestic output. Because the model is formulated in discrete time, the symbol d is used to denote the change in a variable from the last period (y0) to the present (y); that is,


Consequently, dk is the change in the real private capital stock, dMd is the change in the nominal demand for money (hoarding), and dDp is the change in net domestic credit to the private sector extended by the banking system.

The government, in turn, consumes output (G), pays interest on its foreign borrowing (ieF), collects taxes from the private sector (T), receives the profits of the central bank (TB), and finances its deficit by borrowing from the domestic banking system (dDg) and from abroad (edF). We use e to denote the nominal exchange rate and F to represent the foreign-currency value of the government’s foreign debt.6 The government budget constraint is given by


Finally, the banking system, which is assumed to consist solely of the central bank, accumulates reserves with foreign-currency value R, extends credit to the government and the private sector, and issues liabilities in the form of money. With changes in the domestic-currency value of foreign reserves arising out of exchange rate movements being sterilized, the change in the money supply would have the form


where dMs is the change in the nominal money supply, dR is the foreigncurrency value of the change in foreign reserves (the balance of payments), e0 is the exchange rate in the previous period, and dD is the change in total domestic credit, which is defined as


The portion of the central bank’s profits that is transferred to the government is given by 7


Because this economy is open and both owns foreign assets (R) and owes liabilities to foreigners (F), gross national product (GNP), which we denote Y¯, will differ from the value of domestic production Y by an amount equal to net interest payments abroad:


Given these budget constraints, balance-sheet relationships, and definitions, we can proceed to describe the analytical model.

The Growth Component

Most of the modern literature on sources of growth in developing countries takes as its starting point the neoclassical production function.8 Thus, growth of capacity depends on increases in total factor productivity, in the size of the labor force, and in the capital stock. Treating the first two of these as exogenous (albeit possibly responsive to efficiency-enhancing policies), we can formulate the expansion of capacity as a linear function of real investment:


where y is real output (gross domestic product, GDP).9 The coefficient of investment, α1, is the marginal product of capital, and the constant term, α0, captures the combined effects of increases in total factor productivity and the change in the size of the labor force. In empirical analyses of growth in developing countries, an even simpler form of equation (8) is sometimes used, in which α0 is set equal to zero. The result is the familiar “incremental capital output relationship” (ICOR) associated with, among others, Chenery and Strout (1966). The ICOR is also the key relationship in the Revised Minimum Standard Model (RMSM) used by the World Bank to calculate external financing needs for developing countries.10 Here we will work with the less restrictive growth specification, described by equation (8), which allows for productivity changes. The marginal product of capital, however, can be taken as given for our purposes.11

The counterpart to equation (8) for GNP is given by

dy¯=dyPD0[ie(dFdR) +i(F0R0)de]ie0(F0R0)dPDPD0PD=α0 + α1dkPD0[ie(dFdR)+ i(F0R0)de]ie0(F0R0)dPDPD0PD .(9)

Thus, the change in real national income depends both on the change in domestic production and on the change in the real value of net interest payments abroad. If the external interest rate is held constant, the latter depends on changes in net international indebtedness and in the real value of foreign interest payments brought about by changes in the real exchange rate.

The second element in the simple growth model is the identity that relates aggregate investment to aggregate savings. In our framework, this can be written as


where sp, t, g, and ie(FR)/PD are the real values of private saving, taxes, government spending, and net foreign interest payments, respectively, measured in units of domestic goods.12 The first term in equation (10) is, therefore, real private saving, the second is real public saving, and the third is the real current account deficit (that is, real foreign saving).

The third relationship in the model is that for private savings behavior. The simplest way of representing this relationship is to make real private saving proportional to real private disposable income:13


We assume that t and g are exogenous in real terms and that dF is exogenous in foreign-currency terms. Using equation (11), we can rewrite equation (10) as


Finally, because y=y0+dy and PD=PD0+dPD, and using the growth equation (8), we can derive the following relationship between real output growth and changes in the price level in the simple growth model:

dy=(1sa1)1.{α0+α1[s(y0t)+(t g)+e[dFdR i(FR)]PD0+dPD]}.(12)

Since αl <1 and 0 <s <1, the term (1sα1)1 will be positive. This implies that if the trade balance is assumed to be in deficit—that is, [dFdRi(FR)]>0—an increase in dPD will reduce dy, given dR. The economic reason for this inverse relationship is that, since the supply of foreign savings is given in foreign-currency terms, an increase in prices (dPD>0) will result in a real exchange rate appreciation that will reduce the real value of foreign savings measured in terms of domestic goods. With reduced real savings—the other components of aggregate savings are unchanged—real investment falls, and, from equation (8), real output would decline.

Note however that, even after treating t, g, and dF as exogenous, the simple growth model is underdetermined because equation (12) contains three endogenous variables (dy, dR, and dPD).14 Two additional restrictions on these three variables are necessary to close the model.

The Monetary Component

The simple open-economy monetary model has proven to be a very useful device for analyzing balance of payments questions, and variants of this model are used to design financial programs to support the Fund’s lending to its member countries (see International Monetary Fund (1977, 1987)). The basic monetary model, as described in Robichek (1967) and in the papers contained in Frenkel and Johnson (1976), involves three relationships.

The first relationship is the flow supply of money, as given by equation (4):


The second relationship defines the flow demand for money. In general this would be derived from a stock demand for money function that included as explanatory variables real income, expected inflation, expected changes in the exchange rate, and interest rates, among others. A restricted version of this model is the constant-velocity money demand specification, which in flow terms is given by15


where P is the aggregate price level, to be defined presently, and v is the inverse of the income velocity of money.

The third key relationship in this model is the assumption of money market equilibrium, also in flow terms:


Substituting the flow demand for money, given by equation (13), and the flow supply of money (4) into the equilibrium condition (14) and solving for the change in reserves, we have


Equation (15) is the fundamental equation of the monetary approach to the balance of payments (Frenkel and Johnson (1976)). In this model, given the flow demand for money, increases in the rate of domestic credit expansion will he exactly matched by a deterioration in the balance of payments.

The aggregate price level that appears in equation (15) can be expressed as a weighted average of the price of importables (PZ) and the price of domestic output (PD). If the weight of importables is given by θ, we can approximate the change in the aggregate price level as


assuming that θ is constant and that e0=Pz0=PD0=1. Furthermore, if the law of one price holds and Pz*=1 is the (constant) foreign-currency price of importables, we also have


Using equations (16) and (17) to substitute out dP from equation (15), we obtain

dR=vdy +vy0θde +vy0(1θ)dPD+vθdedy+v(1θ)dPDdydD.(18)

Equation (18) summarizes the monetary model. As was the case in the growth model, this model is also underdetermined. Even if de and dD are taken as exogenous policy variables, equation (18) contains three unknowns—dR, dy, and dPD. Again, two additional restrictions among these endogenous variables are required to close this system.

The Merged Model

With these two well-known models in hand, we can proceed to combine them into a merged model that will yield simultaneous solutions for growth, prices, and the balance of payments.

Because both the growth model and the monetary model are incomplete as they stand, they are supplemented by an additional relationship—common to both models—and an ancillary assumption specific to each of the models. The additional relationship is the balance of payments identity:


where X and Z are the foreign-currency values of exports and imports, respectively.

Defining the balance of trade deficit in foreign-currency terms B=ZX, we assume that the change in the trade balance (that is, dB=BB0) is a function of changes in the real exchange rate and income:


In other words, the trade balance improves (B falls) in foreign-currency terms when the real exchange rate depreciates (e/PD>1) and deteriorates (B rises) when real output increases. Using equations (19) and (20), we can write the equation for dR as


Because F=F0+dF and R=R0+dR, equation (21) can be rewritten as


where a=(dFB0)+a(e/PD1)bdyi(F0F0). and B0=B0/(1i)

These equations can now he used to close the growth and monetary models. Note that capital flows in equation (21) are assumed to be exogenous. In a more realistic setting one would expect that capital flows would also be affected by the exchange rate. But if one assumes that capital flows have an autonomous component—corresponding, say, to dF—then the analysis will carry through even if a part of the capital flows is endogenous and responds to exchange rate changes.

Consider first the growth model. We can use equation (21) to eliminate dR from the growth equation (12). The result is

dy=(1sα1be/PD)1.(α0+α1{s(y0t) +(tg) +ePD[B0a(e/PD1)]}).(22)

By adding an ancillary assumption about dPD, this model is able to explain dy. It can be seen that an increase in the private savings rate, government savings, or in the exogenous component of the current account deficit would increase real output growth, in each case by increasing aggregate saving and, therefore, investment.16 The effect of a devaluation is, however, ambiguous. Evaluated at dy0=0, it is given by






Devaluation simultaneously increases the real value of the initial level of foreign saving and, by discouraging imports and encouraging exports, reduces that level. The first effect increases real investment, whereas the second decreases it. If substitution effects dominate (that is, if the parameter a is sufficiently large), the effect of devaluation on growth will be negative (η< 0) in this model.17 This result is a property of any open-economy Harrod-Domar model in which devaluation improves the trade balance. In essence, a real devaluation causes private agents to accumulate assets other than domestic capital, thereby reducing the growth rate of domestic output.

Equations (12) and (21) can be used, together with an assumption about prices, to apply the growth model in a “policy mode.” Given a target value for the balance of payments, say dR*, equations (12) and (21) can be solved simultaneously to determine the output growth rate (dy) associated with each level of external financing (dF). The World Bank’s RMSM model is usually applied in this way to calculate the foreign exchange and financing gaps.

Combining the balance of payments equation (21a) with reserve change equation (18), and assuming that the change in real output is exogenous (dy=dy¯) the monetary model can be solved for dR and dPD. The result is a variant of the solution obtained in the well-known Polak (1957) model.18 The model’s implicit solution for the change in the price of domestic output can be determined from

dpD[dFB0i(F0R0)](b+v)dy¯vy0θdevθdedy¯+a(e/PD1)v(1θ)(y0 +dy¯)= 0.(23)

It can be shown that the domestic price level increases with an increase in the flow of bank credit and decreases with an increase in output. The decrease follows from flow money market equilibrium. A one-unit increase in dy produces an incipient flow excess demand for money, since it both increases demand (by ν) and (through an increase in net imports of b’) reduces the supply of money. Equilibrium requires a reduction in the price of domestic output, which creates a flow excess supply of money both by reducing money demand (by v(1θ)y0) EOM and by increasing money supply (by the amount a’) through an improved trade balance. The effect of a devaluation on the price level is ambiguous. If initial balance of payments equilibrium is assumed and if dy is for simplicity set equal to zero, a devaluation will at the same time raise the flow supply of money by improving the trade balance and also the flow nominal demand for money by increasing the domestic-currency price of importables, and thus the aggregate price level. The net result on the flow excess supply, and therefore on dPD, will depend on the sign of the expression (avy0θ).

The policy mode of the monetary model would be implemented by using equations (18) and (21a), as well as the assumptions about the change in real ouput (dy) and a balance of payments target (dR *). One would then be able to solve for the domestic price level and the implied rate of domestic credit expansion.

Because the growth model assumes exogenous prices and the monetary model keeps real output exogenous, a natural alternative closure for the two models is to dispense with the ancillary assumptions made for each model (about the respective exogeneity of dPD and dy) and to combine the two models to solve for dPD and dy simultaneously. This is readily done by combining equations (22) and (23), which already incorporate equations (12), (18), and (21). The resultant” merged”model is described graphically in Figure 1. Equation (22), corresponding to the growth model, traces a positively sloped locus in dPDdy space (labeled GG in Figure 1), on the assumption that substitution effects are dominant. The slope of GG at the point dy=dPD=0 is

Figure 1.
Figure 1.

Macroeconomic Equilibrium in the Merged Model

Citation: IMF Staff Papers 1989, 002; 10.5089/9781451947045.024.A001

Increases in the domestic price level tend to he associated with increases in output in the growth model because higher domestic prices increase the trade deficit, and the associated increase in foreign saving results in increased investment.

Equation (23), which relates to the monetary model, traces out a negatively sloped locus, labeled MM in Figure 1. The slope of MM at dy = 0 and dPD = 0 is given by


where γ=a+v(1θ)y0>0. In the monetary model, increases in dPD and dy are negatively associated because both tend to increase the flow excess demand for money. It follows that an increase in dPD must he offset by a reduction in dy to maintain flow money market equilibrium.

The intersection of the GG and MM loci at point A in Figure 1 determines the equilibrium values of output changes dy* and domestic inflation dP*D. The model can be condensed into four equations—equations (22) and (23) as well as the GNP and balance of payments equations (9) and (21a). As summarized in Table 1, the model contains four endogenous variables—the growth of output and GNP, inflation, and the balance of payments. There are nine exogenous and predetermined variables. The five policy instruments include fiscal policy variables (t and g), monetary policy variables (dD and dDp), and the exchange rate (de). Finally, the model contains seven parameters.

Table 1.

Structure of the Merged Model

article image

II. Properties of the Merged Model.

Even though the merged model has a fairly rudimentary structure, particularly in terms of its behavioral content, it nevertheless has some interesting properties. This section will examine the impact effects of changes in policy instruments (domestic credit, the exchange rate, government spending), changes in key parameters (the private savings rate, total factor productivity, velocity), and, finally, changes in capital inflows. It is worth stressing that the results obtained from the comparative-static exercises are conditional on the underlying structure of the model and are not necessarily general propositions. Nevertheless, they do yield useful insights into the relationships among growth, inflation, and the balance of payments that one would need to take into account in designing adjustment programs in which growth is an explicit objective.

Changes in Domestic Credit

An increase in the rate of domestic credit expansion creates an excess flow supply of money. At a given level of output, this excess can only be absorbed by an increase in the price level, which induces an offsetting flow excess demand for money through two channels: first, there is an increase in demand for money through the constant-velocity relationship; second, as net imports rise there is a fall in reserves and a consequent reduction in the rate of monetary expansion. In the context of Figure 2 the MM schedule shifts upward to a position such as M’M’. If it is assumed that the increase in the supply of domestic credit leaves the government deficit unaffected,19 the GG schedule would remain stationary. However, the real exchange rate appreciation at point B, by creating a current account deficit, induces an increased use of foreign saving, which increases investment. But the associated rise in output increases the demand for money, thereby reducing the increase in the price level required to restore flow equilibrium in the money market. The end result is that both growth and inflation rise as a consequence of the increase in the rate of domestic credit expansion: that is, we move from A to C in a Figure 2.

Formally, the increase in the rate of inflation is given by



Ω=1α1η(b '+v)βγ>0
Figure 2.
Figure 2.

Effects of an Increase in Domestic Credit

Citation: IMF Staff Papers 1989, 002; 10.5089/9781451947045.024.A001

and, as previously defined,


The change in the growth rate is given by


To solve for the change in the balance of payments, notice from equation (21a) that

δ(dR)δ(dD) =aδ(dPD)δ(dD)bδ(dy)δ(dD)=(bα1βηa)(Ωγ)1<0.(26)

As is well known, in the monetary model of the balance of payments a credit expansion must cause the balance of payments to deteriorate. Overall, the results are consistent with the standard view that an expansion in domestic credit will raise prices, increase output, and worsen the balance of payments.


A devaluation has conflicting effects on the money market. Given the initial price of domestic goods, a devaluation will create a flow excess demand for money by increasing the aggregate price level as the price of imported goods rises, and a flow excess supply as the higher relative price of imports and exports discourages the former and encourages the production of the latter, thus leading to an improvement of the balance of payments.20 The net result will depend on the share of importables in the aggregate price index (θ), on the velocity of money (ν), and on the substitutability between home and traded goods (a). If the degree of substitutability is high, the reduction in net imports will dominate, and an excess flow supply of money will result. If, in contrast, import prices have a large weight in the aggregate price index and the income velocity of money is low, there would be an excess flow demand for money. In Figure 3, which analyzes the effects of a devaluation, it is assumed that there is an excess flow supply of money, which is eliminated by an increase in dPD, causing the MM schedule to shift upward to M’M’. It is important to notice that the upward shift in MM must be less than de. This can be verified from equation (23). If PD rises by de, an excess demand for money will result. Only if dPD <de— and substitution effects dominate—will the implied real exchange rate depreciation generate a sufficient flow supply of money (through the trade balance) so as to restore money market equilibrium.

Inspection of equation (22) verifies that if the real exchange rate is unchanged by a nominal devaluation—that is, if dPD=de—output growth will be unchanged. Thus, at a given dy, a nominal devaluation will shift GG upward by dPD=de, to a point such as B in Figure 3.

Under these circumstances a devaluation will unambiguously increase domestic prices while reducing real output.21 Intuitively, this is because growth can only increase if real savings increase. Since substitution effects are dominant in foreign trade, this requires that domestic prices increase more than in proportion to the devaluation. But if this were the case, there would be a flow excess demand for money. Thus, domestic prices must rise less than in proportion to the devaluation (the real exchange rate must depreciate); with the improvement in the trade balance, real domestic saving, and thus investment, must decrease. In terms of Figure 3, there must be a flow excess demand for money at point B, since the domestic price level has risen by dPD=de and no other variables affecting the market have changed. Consequently, the new M’M’ curve must pass below B. At point C, for example, the money market clears both because the increase in the flow demand for money is less than at B and because the depreciation in the real exchange rate at this point causes an improvement in the trade balance, which increases the flow supply of money. Since the upward shift in the MM curve is therefore smaller than that of the GG curve, the new equilibrium must be to the left of AB, implying that dy* <0.

Figure 3.
Figure 3.

Effects of a Devaluation

Citation: IMF Staff Papers 1989, 002; 10.5089/9781451947045.024.A001

The above assertions can be formally verified by inspecting the derivatives of dPD and dy with respect to a change in the exchange rate. For the change in prices, the derivative is


For growth, the derivative is


Finally, the effect of devaluation on the balance of payments is given by


The balance of payments improves both because the real exchange depreciates and because real output falls. Because prices are fully flexible in this model, aside from a constant factor (given by νyo) the effects of devaluation are equivalent to those of a decrease in domestic credit (equation (26)).

Changes in Government Spending

We now consider the effects of a reduction in real government spending, with real tax revenue and total domestic credit expansion held constant. This implies a reduction in the fiscal deficit, with the supply of credit freed up by the government being rechanneled to the private sector. Because domestic saving rises as a result of the increase in public sector saving, investment and output would rise.22 In Figure 4, therefore, the GG schedule would shift to the right to G’G’. Because domestic credit expansion remains constant, the money market is unaffected, and the MM schedule remains stationary. The increase in output puts downward pressure on domestic prices, and at the new equilibrium—at point C—growth is higher and inflation lower.

Figure 4.
Figure 4.

Effects of a Reduction in Government Spending

Citation: IMF Staff Papers 1989, 002; 10.5089/9781451947045.024.A001

The precise impact effects of a change in real government spending on inflation and growth are


so that a reduction in g decreases dPD, and


so that dy increases when g falls. Because output increases and prices fall, the effect on the balance of payments is ambiguous:


If substitution effects on the trade balance are strong relative to income effects (a’ is large relative to b’) and price effects are large (the value of (b’ + v)/ʏ is large), the balance of payments will improve.

A reduction of the fiscal deficit through an increase in taxes would operate in a similar manner, except that the impact effect on domestic saving would be 1—s rather than unity because there is a reduction in private saving as private disposable income declines.

Changes in the Private Savings Rate

The effects of an increase in the private savings rate in this model are quite similar to those resulting from an increase in public sector saving.23 A rise in the private savings rate increases domestic saving, thereby creating an excess of saving over investment. As investment rises output would be increased, shifting the GG curve in Figure 4 to the right. As the increased investment leads to higher output, however, a further increase in private saving is induced. This multiplier effect causes the shift in the GG schedule to be larger than would be the case when public saving is increased. If it is assumed that the increase in private saving is matched by investment and does not go into hoarding—leaving MM unaffected—there will be a higher rate of growth and a lower rate of inflation.24 The new equilibrium would be at a point such as C in Figure 4, but (as mentioned above) this point would lie further to the right than would be observed when government spending is reduced.

The effects of an increase in the private saving rate are given by




The balance of payments outcome is ambiguous, since prices and output move in opposite directions:


Changes in Total Factor Productivity

Consider the effects of an improvement in total factor productivity—that is, an increase in a0. Because at the original level of investment a larger amount of output would be forthcoming, an incipient excess of saving over investment would ensue. This new saving would be channeled into investment, and output would rise still further. The GG schedule would shift rightward, as in Figure 4, and once again the MM schedule would be unaffected. Thus a rise in factor productivity, like an increase in saving, is in effect a positive supply shock. Again, the increase in output exerts downward pressure on prices and the economy moves from A to Gin Figure 4. (We again ignore the changes in the slope of GG.) The price and output effects are




but the balance of payments effect is ambiguous:


Changes in Velocity

Suppose there is an exogenous increase in the demand for money, in the form of a reduction in the income velocity of money, given a positive initial output growth rate dyo. This increase would give rise to flow excess demand in the money market so that, at a given dPD, the MM curve would shift to the left, say to M’M’ in Figure 5. In other words, growth must fall to reduce the flow demand for money and to restore flow equilibrium in the money market. The GG curve is unaffected, and the economy moves to a new equilibrium on this curve with lower growth and reduced inflation, as at point B in Figure 5. The formal effects of a change in velocity on prices, output, and the balance of payments are

Figure 5.
Figure 5.

Effects of Reduced Velocity

Citation: IMF Staff Papers 1989, 002; 10.5089/9781451947045.024.A001


Because domestic prices and output both fall in this case, an increase in the demand for money causes the balance of payments to improve.

Changes in Capital Inflows

An increase in capital inflows, with the exogenous components of the current account held constant, must in the first instance be transformed into accumulation of reserves. Thus, such inflows will exert their impact on the domestic economy through the money supply—that is, they give rise to a flow excess supply of money. As long as these capital inflows go into international reserves, the economy’s use of foreign saving is unchanged. Thus, the MM curve in Figure 5 would shift to the right and the GG curve would remain stationary. Both qualitatively and quantitatively, this case is identical to that of an increase in domestic credit to the private sector (Figure 2). The reason for this equivalence is that both the increase in domestic credit and the increase in capital inflows function as money supply shocks, with the former operating through an increase in the central bank’s domestic assets and the latter through an increase in its foreign assets.

A more interesting exercise is one in which a new capital inflow is matched by an increase in imports—that is, d(dF)=dB0. Because such an inflow has no immediate monetary consequences, the MM curve remains stationary. The GG curve, in contrast, shifts to the right because the increased supply of foreign saving increases investment and thus output. The results are increased output and lower prices. The equilibrium is similar to that illustrated in Figure 4. Quantitatively, the effects on domestic prices, output, and the balance of payments are identical to those of a reduction in government spending, since both represent an infusion of savings. In this case, it is also of interest to investigate the effects on real national income (GNP). Using equation (9) and assuming initial current account balance (dF0dR0=0), we have


Using equations (30)–(32), we obtain


These results can readily he given an intuitive interpretation. A one-unit capital inflow used for investment increases output on impact by al (the marginal product of capital). But, as equation (43) shows, the total effect on domestic output differs from αi in this model by a factor of (Ωβ)’1, which consists of two parts, corresponding to the induced effects on saving generated by the initial capital inflow. First, the term C3–1 indicates that the original inflow of saving will be magnified by the effects of positive marginal propensities to save and to import (recall that β=1α1(s+b)<1), so that a one-unit increase in saving increases total saving by β–1, at a given dPD. Second, since domestic prices fall, this will also affect real saving (this corresponds to the movement from B to C in Figure 4). This effect is captured by the factor Ω–1, which is less than unity in absolute value. Thus the increase in domestic output may exceed or fall short of the marginal product of capital α1.

According to equation (43), the change in national income is derived by subtracting from the change in output the change in real factor payments abroad, which is the product of the interest rate and the change in real net international indebtedness. The change in real net international indebtedness, in turn, consists of the difference between the additional capital inflow and any induced changes in foreign exchange reserves, plus the change in the real value of interest payments. The induced balance of payments effect of the capital inflow may be positive or negative. To clear the flow money market, the increase in output must be accompanied by a decrease in prices. These changes have conflicting effects on the balance of payments, however, with the former reducing the overall surplus and the latter increasing it. The balance of payments surplus is likely to increase the larger are substitution effects (a’) relative to income effects (b’). To the extent that the balance of payments goes into deficit, net international indebtedness will increase by more than the one-unit capital inflow. If a surplus results, however, net indebtedness will increase by less than one unit. Finally, because domestic prices fall, the real value of the original level of interest payments must rise. This increases the real burden of servicing the debt and thus reduces real national income, accounting for the final term in equation (43).

Equation (43) thus represents a generalization of the familiar result that a capital inflow devoted to investment will only increase national income if the marginal product of capital (α1) exceeds the interest rate charged on the loan (1). The generalization involves taking into account induced effects on saving and on real net international indebtedness implied by the model.

Summary of Results

To obtain an overall perspective on the properties of the merged model, the results obtained from the comparative-static experiments are summarized in Table 2. This table shows the signs of the impact effects that changes in the various policy instruments, behavioral parameters, and exogenous variables have on prices, real output, and the balance of payments.

Table 2.

Impact Effects of Changes in Policy Instruments, Behavioral Parameters, and Exogenous Variables

article image

Matched by an offsetting change in imports.

III. Conclusions

The objective of this paper has been to outline a simple analytical framework—combining two popular models—that could prove useful in designing growth-oriented adjustment programs. This framework allows for the treatment of economic growth as an explicit policy objective, along with the objectives of balance of payments improvement and price stability that are usually parts of an adjustment strategy. The integrated model is thus able to address many of the issues that would arise in formulating a growth-oriented adjustment program. For example, given targets for growth, the balance of payments, and inflation, the model can be used to determine a set of demand management policies (domestic credit ceilings and reductions in the fiscal deficit), exchange rate policies, structural policies (policies to increase savings and the level and efficiency of investment), and external financing policies that would achieve these targets. The model outlined in this paper can also he used to evaluate the impact effects of certain domestic and foreign shocks on the economy.

The type of framework developed here should not, however, be interpreted as a comprehensive model of growth and adjustment, or as a completely realistic representation of the behavioral and structural characteristics of a developing economy. Although the model is internally consistent and includes many of the principal policy targets and policy instruments, there are several areas where further work is both necessary and desirable in order to introduce more realism into the analysis.

First, the behavioral equations in this model are rudimentary, amounting in some cases (for example, private savings and the demand for money) to simple rules of thumb. Furthermore, the institutional framework is extremely simplified. The monetary sector of the model, for example, does not cover demand for and supplies of different types of domestic and foreign financial assets, and more important, domestic interest rate determination. By excluding the interest rate, the model leaves out a potentially important channel through which monetary policy could affect the economy, as well as an important policy instrument to change savings and thus output. In addition, no attempt was made to formalize the interaction of the central bank, the fiscal sector, and the banking system. This is an area where further work would yield significant payoffs. Although it would obviously be desirable to incorporate behavioral and institutional relationships that are more realistic and analytically defensible, each such innovation would involve sacrifices in transparency and ease of application. The model described here does provide a first step in the direction of increased realism—by integrating capacity growth (albeit in a very simple way) into a standard monetary framework—but even this first step has been taken at some cost in terms of complexity.

Second, the model is a one-period model, and no allowance is made for lags and dynamics. This feature is particularly striking in the case of prices, which continuously adjust to clear the flow money market. The addition of short-run dynamic behavior—say, through lags in adjustment of prices to monetary disequilibrium or through slow revision of expectations of future inflation—while perhaps not changing the overall conclusions, would nonetheless yield useful information on the time path of prices. A satisfactory treatment of dynamics is necessary, since policy- makers are often as concerned with the time paths of the target variables as they are with the final outcomes. The model as presently formulated does not enable one to trace the transition of the economy from one equilibrium to another. Moreover, the model has an implicit long-run dynamic structure (through the accumulation of capital as well as financial assets) that implies that present actions have future consequences. Thus, a desirable extension of the present analysis would be to exploit the longer-term dynamics arising from asset and capital accumulation to determine if the impact effects obtained here hold up in both the transition and the steady state.

Third, growth in this model is entirely determined by supply factors, and the economy is assumed to be always operating at full capacity. As such, changes in aggregate demand have no effect, even in the short run, on the rate of capacity utilization. In a properly specified macroeconomic model, one would presumably wish to make distinctions among growth in productive capacity, growth of output resulting from more efficient use of productive capacity, and growth that results from increases in aggregate demand (when there is excess capacity in the economy). In the present full-capacity framework, one cannot treat important issues of employment and wage determination, and it is certainly possible that if one relaxed the full-employment assumption some of the results we have obtained would be altered.

Fourth, capital flows would in general respond to expectations of interest rates and exchange rates, which in turn would likely be affected by policy actions. Assuming foreign financing to be exogenous, as is done in the model here, is quite restrictive and may well lead to incorrect policy recommendations. Although there can be no argument about the appropriateness of treating capital flows as endogenous to domestic policies, it must, however, be recognized that modeling such relationships is quite difficult. As yet there have been few successful attempts to do this, reflecting both a lack of theory on what drives capital movements and an inability to properly model unobservable variables such as expectations.

In conclusion, the present model can be viewed as a first step in the development of a conceptual framework within which to design growth-oriented adjustment programs. The simplicity of the model is certainly a virtue; it is easily understood and can be readily applied with a minimal amount of information. From an operational perspective, particularly in countries where data are limited and of uneven quality, this latter feature takes on considerable importance. It would seem pointless to design complex models that cannot be applied because they have extremely demanding information requirements. Before this model is used, however, it is necessary to test the dynamic properties and the empirical relevance of the simple model. If the model passes these tests, it could serve as a foundation on which more elaborate and realistic structures, possibly along the lines mentioned above, can be built.


  • Chenery, Hollis B., and Alan M. Strout, “Foreign Assistance and Economic Development,” American Economic Review (Nashville, Tennessee), Vol. 56 (September 1966), pp. 679733.

    • Search Google Scholar
    • Export Citation
  • Corbo, Vittorio, Morris Goldstein, and Mohsin S. Khan, eds., Growth-Oriented Adjustment Programs (Washington: International Monetary Fund and World Bank, 1987).

    • Search Google Scholar
    • Export Citation
  • Fischer, Stanley, “Economic Growth and Economic Policy,” in Growth–Oriented Adjustment Programs, ed. by V. Corbo, M. Goldstein, and M. S. Khan (Washington: International Monetary Fund and World Bank, 1987), pp. 151178.

    • Search Google Scholar
    • Export Citation
  • Frenkel, Jacob A., and Harry G. Johnson, eds., The Monetary Approach to the Balance of Payments (London: Allen and Unwin, 1976).

  • Group of Twenty Four (Intergovernmental Group of Twenty Four on International Monetary Affairs), “The Role of the IMF in Adjustment with Growth” (New York, March 1987).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, The Monetary Approach to the Balance of Payments (Washington: International Monetary Fund, 1977).

  • International Monetary Fund, Theoretical Aspects of the Design of Fund-Supported Adjustment Programs, Occasional Paper 55 (Washington: International Monetary Fund, September 1987).

    • Search Google Scholar
    • Export Citation
  • Khan, Mohsin S., “Macroeconomic Adjustment in Developing Countries: A Policy Perspective,” World Bank Research Observer (Washington), Vol. 2 (January 1987), pp. 2342.

    • Search Google Scholar
    • Export Citation
  • Khan, Mohsin S., and Malcolm D. Knight, Fund-Supported Adjustment Programs and Economic Growth, Occasional Paper 41 (Washington: International Monetary Fund, November 1985).

    • Search Google Scholar
    • Export Citation
  • Khan, Mohsin S., and Malcolm D. Knight, Peter Montiel, and Nadeem Haque, “Adjustment with Growth: Relating the Analytical Approaches of the World Bank and the IMF,” Development Policy Issues Series Discussion Paper (unpublished; Washington: World Bank, October 1986).

    • Search Google Scholar
    • Export Citation
  • Lizondo, J. Saul, and Peter Montiel, “Contractionary Devaluation in Developing Countries: An Analytical Overview,” Staff Papers, International Monetary Fund (Washington), Vol. 36 (March 1989), pp. 182227.

    • Search Google Scholar
    • Export Citation
  • Michalopoulos, Constantine, “World Bank Programs for Adjustment and Growth,” in Growth-Oriented Adjustment Programs, ed. by V. Corbo, M. Goldstein, and M. S. Khan (Washington: International Monetary Fund and World Bank, 1987), pp. 1562.

    • Search Google Scholar
    • Export Citation
  • Polak, Jacques J., “Monetary Analysis of Income Formation and Payments Problems,” Staff Papers, International Monetary Fund (Washington), Vol. 6 (November 1957), pp. 150.

    • Search Google Scholar
    • Export Citation
  • Robichek, E. Walter, “Financial Programming Exercises of the International Monetary Fund in Latin America,” address to a seminar of Brazilian Professors of Economics, Rio de Janerio, September 1967.

    • Search Google Scholar
    • Export Citation
  • Robinson, Sherman, “Sources of Growth in Less Developed Countries: A Cross-Section Study,” Quarterly Journal of Economics (Cambridge, Massachusetts), Vol. 85 (August 1971), pp. 391408.

    • Search Google Scholar
    • Export Citation

Mr. Khan is Assistant Director in the Research Department. He is a graduate of Columbia University and the London School of Economics and Political Science.

Mr. Montiel is a Senior Economist in the Developing Country Studies Division of the Research Department. He is a graduate of Yale University and the Massachusetts Institute of Technology.


See Khan and Knight (1985) for a survey of selected macroeconomic models for developing countries.


The monetary approach to the balance of payments plays a key role in the formulation of the Fund’s financial programs (International Monetary Fund (1977, 1987)), and the World Bank uses a two-gap growth model to establish external financing needs and consistent projections across countries (Khan, Montiel, and Haque (1986)).


The model herein builds on the analysis of Khan, Montiel, and Haque (1986).


For simplicity, we assume that the private sector cannot borrow abroad and that domestic interest rates are zero. Although the introduction of private capital flows and domestic interest payments would complicate the analysis, it is unlikely that it would alter the conclusions.


We assume that there is no public investment, so that all investment in the economy is undertaken by the private sector.


Because only the government can borrow abroad, F is also the economy’s foreign debt.


Capital gains from devaluation are assumed to be retained by the central bank.


See, for example, Robinson (1971) and, more recently, Fischer (1987).


Note that because yo is given, dy is for all intents and purposes equivalent to the growth rate dy/yo.


See Khan, Montiel, and Hague (1986). In practice Bank programs go well beyond what is implied by the RMSM. For an extended discussion of the economics of Bank programs for adjustment and growth, see Michalopoulos (1987).


From the standard neoclassical production function it follows that α1, would be a negative function of the previous period’s stock of capital. Given that this is a one-period model, the marginal product of capital is predetermined.


Lowercase letters denote real variables throughout the paper.


A more realistic specification would involve the addition of the interest rate as an explanatory variable. As long as the interest rate was exogenous or a policy instrument, however, the basic conclusions reached here would not he altered significantly.


Recall that F=F0+dFandR=R0+dR.


Using a more general formulation for the demand for money would not alter the analysis appreciably as long as the function is stable with respect to the explanatory variables.


This relationship assumes that the factor α1(s + b) is less than unity, a condition that is unlikely to be violated empirically.


This is simply a slightly modified version of the familiar Marshall-Lerner condition.


In the Polak model the endogenous variables are the balance of payments and nominal output. With real output assumed to be exogenous, the Polak model then determines the change in reserves and prices.


All the increase in domestic credit therefore goes to the private sector.


We assume that, in the absence of the devaluation, the balance of payments would have been in equilibrium. If not, the change in the domestic currency value of the flow of foreign exchange reserves would also exert monetary effects.


The result that devaluation is stagflationary is, of course, model specific. In the more general case there is a deal of ambiguity about whether devaluation would raise or lower output on impact. See Lizondo and Montiet (1989).


This outcome follows from the assumption that all government spending is devoted to consumption. It is certainly possible that, if we included government investment as well, a decline in government spending could reduce output.


This increase in the private savings rate could be brought in various ways—for example, by an increase in interest rates or in financial development.


This description ignores changes in the slope of GG, which are, however, captured in equations (30) and (31).

IMF Staff papers: Volume 36 No. 2
Author: International Monetary Fund. Research Dept.