7 Transmission of Effects of the Fiscal Deficit in Industrial Countries to the Fiscal Deficit of Developing Countries

Mario Bléjer, and Ke-young Chu
Published Date:
June 1989
  • ShareShare
Show Summary Details

Ahsan H. Mansur and David J. Robinson

I. Introduction

In recent years the aggregate fiscal deficits of both industrial and developing countries have increased markedly and have also tended to move in concert. Although this parallel movement may be partly coincidental, there are a number of reasons to believe that the fiscal outturn in developing countries (during the first half of the 1980s) was, at least partly, adversely affected by the deterioration in the aggregate fiscal deficit of industrial countries. Burdened with external debt obligations and subjected to external financing constraints, developing countries were vulnerable to external shocks in the form of higher interest rates and/or economic recession.1

Conventional analysis suggests that an increase in the fiscal deficit of industrial countries leads to an increase in their imports from developing countries and to an improvement in the latter’s terms of trade. However, for developing countries burdened with external debt and facing an external financing constraint, an expansionary fiscal policy in the industrial countries leading to an increase in the global interest rate may cause an increase in their interest payments, which could offset the favorable effects noted above, and worsen their balance of payments, lower their economic growth, and increase their fiscal deficits. This paper presents a theoretical discussion of the main transmission mechanisms involved, followed by a number of simulation exercises showing the short and medium-term effects of fiscal expansion in industrial countries on the economic growth and fiscal deficit of developing countries facing an external financing constraint.

The key finding of the paper is that an increase in the aggregate fiscal deficit of industrial countries that is not matched by increased private sector savings is likely to worsen the fiscal balance of developing countries. This observation differs somewhat from the standard analysis because in analyzing the transmitted effects of a higher fiscal deficit from industrial to developing countries, both the analytical and simulation models presented here highlight the role of the interest rate, the level of external debt, and the current account constraint, in addition to the conventionally emphasized transmission of the output effect.

The size of the transmitted effect depends on the size of external debt and policy reactions in developing countries. A fiscal expansion in industrial countries may have some initial positive effects on the exports of developing countries, their terms of trade, and output and, consequently, on their fiscal balance, but these are likely to be more than offset in the aggregate by the effect of higher interest rates. Given a current account constraint, higher debt-service payments lead to import compression and to a reduction in output growth. Compression of imports may be induced through quantitative restrictions and unchanged domestic product prices; but if domestic prices in developing countries are allowed to adjust without recourse to quantitative restrictions, either the relative prices of their products or their terms of trade will deteriorate to keep imports in line with the current account constraint. On the fiscal side, these developments imply lower revenues, increased government expenditure on interest payments, and larger fiscal imbalances. If developing countries allow the exchange rate to adjust in line with external developments, the medium-term costs (in the form of lower growth and higher fiscal deficit) could be significantly reduced. The simulations also indicate that if the industrial countries, as a group, had maintained their composite fiscal deficit at its 1977-79 level, the global interest rate would have been lower and the developing countries would have enjoyed slightly higher growth and lower fiscal imbalances both in the short and medium term. The simulations are also broadly consistent with the stylized facts that characterized economic developments during much of the first half of the 1980s.

The plan of the paper is as follows. In Section II we develop a simple two-country analytical model to illustrate the linkages through which the key endogenous variables (e.g., interest rate, output, terms of trade, and determinants of the current account balance) are affected by an exogenous shift in industrial countries’ fiscal deficit. The model, under certain assumptions, can explain the stylized facts characterizing financial developments in industrial and developing countries in the early 1980s. Section III briefly sets out the specifications of key behavioral relationships and the working of the complete simulation model, which allows for extensions, including the accumulation of capital and public sector debt (external and domestic) and the effects of these accumulations on the interest rate, exchange rate, output growth, and fiscal outturns in both the short and medium term. Section IV describes a number of simulation experiments designed to analyze the quantitative effects of shifts in the fiscal deficit in industrial countries and the sensitivity of these effects to changes in key parameters. Some concluding remarks are presented in Section V.

II. A Simple Model of the International Transmission of Fiscal Policies of Industrial Countries

This section sets out a simple macroeconomic model that emphasizes the effects of fiscal policy in industrial countries on the fiscal balance in developing countries. The model starts from the proposition that an autonomous increase in the fiscal deficit in industrial countries will boost their output in the short run, notwithstanding an increase in the real interest rate.

These developments affect the fiscal outturn of developing countries in a number of ways. On the expenditure side, interest payments increase on both external and domestic public sector debt, owing to the higher global interest rate; the foreign component of development expenditure also increases if the binding current account constraint leads to an exchange rate adjustment. On the revenue side, in the short run, revenue will increase if exports, imports, and domestic economic activities in developing countries are favorably affected by the shift in aggregate demand in industrial countries. However, the adverse effects of the higher interest rate may reverse the initial favorable effects of fiscal expansion in the industrial countries, and output and trade expansion may decline globally in both the short term and the medium term, with negative effects on the fiscal outturns for the developing countries.

The interrelationship between the fiscal situation in developing countries and fiscal policy in the industrial world may be highlighted by the specification of a simple fiscal deficit relationship for developing countries (D*), along with a simultaneous system of equations determining the interest and exchange rates and output of industrial and developing countries in terms of a two-country model.2

To highlight the linkages and qualitative effects on the endogenous variables, we start with a simple two-country model with four equations, determining interest rate, relative price, and output levels. The price of the industrial countries’ product is assumed to be fixed and used as the numeraire, and the relative price movement or the terms of trade effect arises from variations in the price of the developing countries’ product. The relative price is determined, inter alia, by the level of imports consistent with the current account constraint; if the relative price is maintained at a fixed level in the face of external shocks (for example, higher interest rates), the level of imports into the developing countries is determined quantitatively. For expositional clarity, we make a number of restrictive assumptions, some of which are relaxed in the empirical analysis presented in Section III.

We assume that the credit market is fully integrated among industrial countries and that a single rate of interest applies to all industrial countries. Ex ante, investment in industrial countries is assumed to depend on the real interest rate (R):

Private sector savings (S) in industrial countries depend on the real interest rate (R) and income (Y). In line with the consensus that changes in the public sector deficit (D) are likely to be offset, at least partly, by alterations in private savings behavior (the Barro-Ricardo effect), we define the measured private savings as

where θ = the Barro-Ricardo coefficient reflecting the private sector’s induced response to public sector dissaving.3

The expression S(Y, R) is the component of private savings that corresponds to net wealth accumulation and θ · D is the Barro-Ricardo component reflecting the private sector’s induced response to public sector dissaving.

Domestic output in industrial countries is assumed to be demand-determined, depending on domestic absorption (A), which is a function of income, the real interest rate, and the government deficit. This is a flexible output model with a fixed price for the output of industrial countries. Developments in developing countries have no effect on this output.

Macroeconomic equilibrium in industrial countries can be expressed in terms of a simple two-equation model of income and interest rate determination:

where SER*(R) is the service sector deficit of developing countries and CÁ is the external sector current account balance excluding the service sector; the current account constraint (CA¯) implies that

For developing countries, the current account constraint is assumed to be fixed and equal to the financing available from industrial countries.5 If the interest rate increases, the service account balance would deteriorate, implying a reduction in imports, if it is not offset by a favorable growth in exports. The reduction in imports could be induced either through a decline in the relative price (p) of developing countries’ product (in terms of industrial countries’ product) or by the imposition of quantitative restrictions (to avoid a deterioration of developing countries’ terms of trade). Quantitative restrictions, however, do not benefit exports, whereas a reduction in the relative price does and thus helps ease import compression. Depending on the exchange regime under consideration, the current account constraint may be specified as fixed price with quantitative control:

or as flexible price without quantitative control:


p is the relative price of developing countries’ output in terms of industrial countries’ output.

Output in developing countries is assumed to be constrained by the availability of imports (equation (4)). A relaxation of the current account constraint and/or lower interest payments and higher exports would allow higher imports and, thus, higher growth:

The fiscal deficit for developing countries may be specified as


where RY* = domestic-based revenue; RIMP* = revenue from import duty; REX* = revenue from export duty; G* = government expenditure and net lending; and tx, tm, and tY are, respectively, the tax parameters for export (X*), import (M*), and domestic-based taxes. Equation (5) indicates that the fiscal policies of industrial countries influence the fiscal outturns of developing countries through their impact on the latter’s interest rate, output, exchange rate, imports, and exports.

Equations (1)-(5) determine five endogenous variables: the output levels of industrial and developing countries (Y and Y*, respectively), industrial countries’ real interest rate (R), relative price (p), and the budget deficit of developing countries (D*). The exogenous variables are limited to the fiscal deficit in industrial countries (D), the current account constraint, the external debt of developing countries, and the factors (not specified) that may influence the five endogenous variables noted above. In its present form, the model is partly recursive. Given the constrained level of the current account balance and the outstanding external debt of developing countries, Y and R are determined simultaneously by equations (1) and (2); the relative price or the levels of imports and output in developing countries can be obtained by substitution in equations (3) or (3′) and (4), respectively. Finally, the fiscal balance of developing countries (D*) can be derived through substitution of these four endogenous variables in equation (5).

In this simple system the effect of a higher fiscal deficit on the output of industrial countries depends on the expansionary effects of the higher fiscal deficit, the service sector surplus, and the improvements in the terms of trade relative to the contractionary effect of a higher interest rate. If the private sector treats a portion of domestic government bonds as a component of its net worth, implying that full Barro-Ricardo equivalence does not hold,6 an increase in the fiscal deficit of industrial countries would create an imbalance between global savings and investment, and might lead to a rise in the interest rate to restore equilibrium.7 If the elasticity of investment with respect to interest is low, the crowding-out of private investment would be so small that an increase in fiscal expansion would have a positive effect on domestic absorption and output in industrial countries. (See Appendix I for details.) The conditions under which an expansionary fiscal policy leads only to a smaller increase in the interest rate also imply a greater expansionary effect on output.

Given the current account constraint as specified in equation (3) or (3′), an expansionary fiscal policy would worsen (improve) the terms of trade of developing countries if the favorable output effect from higher exports is less (more) than the increased external debt servicing owing to higher interest rates, resulting in a decline (increase) in developing countries’ imports and a reduction (increase) in their growth rate. Thus, on the one hand, developing countries whose external debt-servicing obligations are not tied to market interest rates may benefit from fiscal expansion in the industrial countries in the short run. On the other hand, for heavily indebted developing countries with a large proportion of debt contracted at floating market rates, a higher interest rate may lead to severe import compression, lower output growth, and a deterioration in the budget deficit in the absence of additional adjustment measures.

This simple model does not, of course, fully capture the long-run impact of fiscal policy changes. Although a higher real interest rate may be expected to continue, the effects on output, the exchange rate, and the current account may be altered or reversed substantially over time as the process of capital formation is adversely affected by higher interest rates and as the process of asset and wealth accumulation influences saving and investment behavior and balance of payments flows. These longer-term aspects of the effects of fiscal policy changes are examined through simulation experiments in Section IV.

The model presented in this section is broadly consistent with the stylized facts of the period 1979-85. The average real interest rate increased by more than 4 percentage points during 1979-85, when the composite fiscal balance of the major industrial countries increased by more than 2 percentage points in relation to gross national product (GNP). (See Table 1.) In line with the higher interest rate, debt-service payments of developing countries increased by 70 percent during 1979-85 in terms of U.S. dollars and, in relation to exports of goods and services, they increased rapidly from 14 percent in 1979 to 20.5 percent in 1985. The payment obligations of the heavily indebted countries increased at a much faster rate, from around 30 percent of exports of goods and services in 1980 to about 50 percent by 1982.8 Furthermore, a deterioration in the terms of trade and a decline in gross capital formation also contributed to a marked slowdown in economic growth in developing countries. Both reduced imports and slower economic growth contributed to a slower growth in revenue, and, together with an increase in expenditure through higher interest payments, they led to a doubling of the fiscal deficit to around 5 percent of gross domestic product (GDP).

Table 1.Selected Variables for the Industrial and Developing Countries, 1979-85
Fiscal balance (central government)1
Industrial countries
Unadjusted fiscal balance−2.8−3.3−3.6−4.6−5.4−5.0−4.9
Cyclically adjusted balance−3.0−2.9−2.7−2.7−3.5−3.7−3.7
Developing countries−2.4−1.5−3.6−5.4−5.6−4.7−4.6
Interest rate (nominal)29.912.714.
Interest rate (real)
Central government interest expenditure (in percent of total expenditure and net lending)
Industrial countries7.
Developing countries6.
Of which: non-oil developing countries(7.6)(6.9)(8.2)(9.8)(12.4)(14.8)(16.3)
Industrial countries43.41.31.4−
Developing countries54.
Of which: 15 heavily indebted countries(6.1)(5.0)(0.5)(−0.4)(−3.4)(2.2)(3.1)
Debt-service payments of developing countries
(percentage of exports of goods and services)14.112.916.219.518.920.120.5
Of which: 15 heavily indebted countries(34.7)(29.6)(38.0)(49.4)(42.5)(41.1)(38.7)
(billion U.S. dollars)82.6100.5127.4138.1127.6142.1140.3
Terms of trade for primary product exporters60.3−7.4−10.3−−3.7
Gross capital formation (in percent of GDP)
Developing countries25.925.925.524.323.322.922.4
Of which: 15 heavily indebted countries(24.9)(24.7)(24.5)(22.3)(18.2)(17.4)(16.5)
Sources: International Monetary Fund, World Economic Outlook, April 1987: A Survey by the Staff of the International Monetary Fund; Government Finance Statistics Yearbook, various Issues.

As a percentage of GNP/GDP; industrial country data cover the seven major industrial countries.

Weighted averages of short-term nominal interest rates of the seven major industrial countries.

The composite consumer price increase of the industrial countries has been used as the price deflator.

Annual percentage change in the composite real gross national product (GNP).

Annual percentage change in the composite real GDP.

Annual percentage change.

Sources: International Monetary Fund, World Economic Outlook, April 1987: A Survey by the Staff of the International Monetary Fund; Government Finance Statistics Yearbook, various Issues.

As a percentage of GNP/GDP; industrial country data cover the seven major industrial countries.

Weighted averages of short-term nominal interest rates of the seven major industrial countries.

The composite consumer price increase of the industrial countries has been used as the price deflator.

Annual percentage change in the composite real gross national product (GNP).

Annual percentage change in the composite real GDP.

Annual percentage change.

Notwithstanding the qualitative inferences that may be made from these preliminary observations, a number of empirical questions remain unresolved. First, how significant, in quantitative terms, is the effect of a change in the fiscal deficit in industrial countries on the fiscal outturn for developing countries? Second, what are the long-run effects on the key endogenous variables when the dynamic processes described above are taken into account? We need a dynamic empirical model to answer these questions, even in a very simplified way, and such a model is considered in the next section.

III. The Simulation Model

The simulations were carried out on a medium-sized model incorporating 12 behavioral equations and 30 definitional equations or identities. In essence, the structure of the model is very similar to the one in Section II. However, in order to take into account the longer-term effects of a fiscal expansion, the model has been expanded here to include important stock-flow constraints, and the role played by relative prices has also been expanded. For example, as shown in Section III.1 below, real interest rates and national income are still determined by the interaction of savings and investment decisions with the current account in a fashion similar to equations (1) and (2) above.9 To bring in dynamic factors, real GDP is made a function of potential output, which, in turn, depends on capital stock and investment; and consumption depends on consumers’ wealth, which reflects the size of previous government deficits, current account surpluses, and investment. The forms of the import and export equations in the empirical model are also similar to those embodied in equation (5) in Section II. The main difference here is that we explicitly allow for the effects of import compression on exports, as described in Section III.2 below. As was done for the analytical model, we consider both fixed- and flexible-price specifications for developing countries. We allow for some interaction among the financing of the fiscal deficit, money supply, and domestic price determination in developing countries.

The behavioral equations and estimates of their parameters are based on a survey of the existing empirical literature. In cases where the estimated parameter values tend to vary among countries, we generally used a value in the midrange. The data used are largely taken from various issues of the following Fund publications: International Financial Statistics, Government Finance Statistics Yearbook, and World Economic Outlook. Stock variables, which are endogenous to the model in a dynamic context (for example, private sector wealth, public sector debt, and capital stock), are estimated by an accumulation of the relevant flows. The baseline values of some key economic variables used in the simulations are set out in Table 2.

Table 2.Baseline Values of Main Economic Variables Used in the Simulations, 1977–84
Percentage of GDP
Industrial Countries
Fiscal deficit−3.3−3.4−3.0−3.5−3.9−4.4−5.7−5.0
Million U.S. dollars
Non-oil developing countries
Interest payments14.322.
Current account deficit−21.7−31.6−48.5−75.2−94.5−72.1−36.5−21.9
External debt359.0390.6439.1524.9621.1708.0748.5793.4
Percentage of GDP
Import taxes3.
Fiscal deficit−3.6−3.5−3.2−3.7−5.0−6.2−5.6−5.2

Exports to, and imports from, industrial countries.

Excluding grants.

Exports to, and imports from, industrial countries.

Excluding grants.

This section provides only a brief description of the structure of the model and its key equations; a full description of the simulation model and a discussion of the associated parameters are provided in Appendices II and III, respectively. The model can be broadly divided into three parts: the real sector in industrial countries, in which the real interest rate is determined; the trade sector, through which the effects of external shocks are transmitted to developing countries; and the real and fiscal sectors in developing countries, which describe how changes in the external environment affect these countries.

1. The Real Sector in Industrial Countries

The real sector in industrial countries is composed of three basic relationships determining output, consumption, and investment. Capacity output is derived from a simple Cobb-Douglas function containing the capital stock and labor, with the capital stock endogenously determined from the investment function described below and labor supply taken as exogenous. The output function essentially describes changes in output from the baseline level, with higher real capacity output (YCR) leading to higher real output over a period. Output growth can also be temporarily disturbed by changes in real government expenditure:

Based on empirical estimates for the United Kingdom made by Laidler and O’Shea (1980), a 1 percent increase in real government expenditure (G/P) is assumed to give rise to a 0.15 percent increase in real GDP, when θ, the Ricardian constant, is 0.5.

The consumption function is based on the formulation of Blinder and Deaton (1985):

Real private consumption (CP/P) is positively dependent on real wealth (W/P) and real disposable income (YDIS/P), and negatively dependent on nominal interest rates (NR) and expected inflation (EINF).10 Disposable income is defined to exclude the portion of savings that takes place to offset changes in the real government deficit (the Barro-Ricardo effect).

Since the interest elasticity of consumption, γ2, is a key parameter in the simulations, a brief discussion of its value may be helpful. On purely technical grounds, γ2 is expected to be negative in sign, as an increase in interest rates encourages saving.11 Blinder and Deaton, like other researchers, encountered significant difficulties in finding a stable and well-determined estimate. Their estimates (for the United States) vary from -2.3 to -0.8, with the former being slightly better in econometric terms than the latter. In our baseline simulation we assume γ2 = -0.8, which appears intuitively reasonable and closer to other results (for example, Masson and Knight (1986)). The sensitivity of the results to the value of this parameter is examined in Section IV below.

The investment function is based on Masson and Knight’s formulation:

Investment is positively related to the gap between actual output and capacity output, and negatively related to the real interest rate (R) and to the capital stock (K) in the previous period. This equation implies that real private investment adjusts, with a lag, to an optimal capital stock, dependent on both real interest rates and expected output (proxied by actual output).12

Finally, the identity

closes the system. With the current account also determined by export and import equations, as described below, this relationship can essentially be seen as the equation determining the real interest rate, bringing domestic absorption into line with the current account.

2. Trade Flows

Developing countries are assumed to face a rigid current account constraint in nominal terms, equal to the actual current account deficit in each year. With interest payments determined by interest rates and the outstanding debt, this determines the trade balance.

The price and volume of exports for developing countries are determined separately, following Khan and Knight (1981). Export volume is supply-determined, and export price reacts to equate the supply to world demand, with a lag. Export supply itself is a function of three factors—capital stock in the export sector (proxied by real GDP), relative prices, and the supply of imported inputs (proxied by import volume). Three points should be noted here. First, growth in industrial countries has an immediate impact on export price, rather than on export supply (although supply is subsequently affected by the corresponding improvement in relative prices). Second, import compression reduces export volume. Thus, if developing countries are forced to reduce imports, this, in turn, reduces exports, creating a vicious circle. Third, export prices for developing countries can and do differ from the domestic prices in both industrial and developing countries, allowing relative prices to play a role in the model.

When the nominal exchange rate is held constant, imports are determined simply as a residual, given the current account balance, available exports, and debt-servicing obligations. In the longer term, of course, such exchange rate rigidity is unrealistic. Therefore, in alternative simulations, a simple import demand equation dependent on real income and relative prices is added, with the exchange rate adjusting to achieve the required trade balance.

3. The Fiscal and Real Sectors in Developing Countries

The specification of the fiscal sector in developing countries is relatively straightforward. Tax revenues are directly dependent on import value (in domestic currency) and GDP. Interest payments on government foreign debt are related both to the exchange rate and to a weighted average of market and concessional interest rates. Interest payments on domestic debt are related to domestic interest rates, which are initially assumed to be fixed in nominal terms.13 Other government expenditures are assumed to be fixed in real terms. Changes in the government deficit in developing countries, induced by external shocks or otherwise, can be financed either by recourse to the sale of domestic debt14 or by borrowing from the domestic banking system. In the latter case, the increase in money stock adds to inflation through a simple price equation similar to that of Khan and Knight (1981). Real GDP in developing countries (Y*) is specified as a function of import volume (MV*).

This formulation ignores many important factors, but it is sufficient to allow us to focus on the effects of import compression on developing country growth. The coefficient α4 is set equal to 0.3. (See the survey by Goldsborough and Zaidi (1986).)

Overall, the model works in broadly the same way as the theoretical model described in Section II. However, in the longer term, the temporary boost to industrial countries’ GDP caused by the fiscal expansion wears off, and—reflecting reduced investment and thus lower capacity output–real GDP falls below the baseline level, lowering developing country exports. In addition, a sustained higher government deficit adds significantly to private sector wealth, and the higher interest payments boost disposable income. If the Ricardian constant is less than 1, both these factors tend to increase consumption, creating an upward pressure on interest rates and adding to the import compression faced by developing countries.

IV. Simulation Results

Before we turn to the simulation results themselves, some brief introductory comments may be helpful. First, the simulation model was calibrated to produce the actual outturn over the period 1976-84 with the given policy stance. The results of changes in policies are therefore all expressed in terms of divergences from this baseline. Second, as noted above, the effects of a fiscal policy change depend critically on policy responses to the change in other areas in both industrial and developing countries. We assume that monetary policy in industrial countries is adjusted to maintain prices at the baseline level,15 whereas in developing countries it adjusts only to take into account any changes resulting from the monetary financing of the government deficit. In the first round of simulations, the exchange rate and the domestic interest rate in developing countries are assumed to be fixed at the baseline levels. In later simulations, we examine the effects of allowing interest and exchange rates in developing countries to adjust, in line with some simple policy rules described below.

1. A Sustained Increase in Expenditure in Industrial Countries

The model was first used to examine the effects of a sustained debt-financed increase in government expenditure, sufficient to increase the fiscal deficit by about 1 percent of GDP in the first year, and maintained in real terms thereafter.16

a. Macroeconomic Effects

The macroeconomic impact of the fiscal expansion can be divided into two stages. In the first year, the fiscal stimulus raises real GDP in industrial countries by about 0.5 percent, and the real interest rate by about 1.4 percentage points (Table 3 and Chart 1). The boost in industrial countries’ demand results in some increase in developing countries’ export prices, but this is more than offset by the effect of higher interest rates on their debt-service payments. Consequently, in the short run, import volume falls—which further weakens exports—and results in a 0.4 percentage point fall in real GDP.

Table 3.Effects of a Sustained Increase in Government Expenditure1
Years After Initial Expansion
(Deviations from baseline values; in percent)
Industrial countriesPercentage Difference
Real GDP0.50.2−0.1−0.2
Real consumption−0.5−0.8−1.1−0.9
Real interest rate1.
Percentage of GDP
Fiscal deficit1.
Government interest payments0.
Percentage Difference
Developing countries2
Real GDP−0.4−0.4−0.5−1.3
Real GNP−0.7−0.6−0.8−1.8
Consumer price index0.
Import volume−2.0−2.0−3.1−7.7
Export volume−0.3−0.5−1.3−3.3
Export prices0.
Percentage of GDP
Government expenditure0.
Interest payments0.
Government revenue−0.1−0.1
Trade taxes−0.1−0.2
Fiscal deficit0.

Assumes a sustained debt-financed increase in government expenditure in industrial countries sufficient to increase the fiscal deficit by 1 percent of GDP in the first year.

Fiscal deficit is assumed to be money-financed.

Assumes a sustained debt-financed increase in government expenditure in industrial countries sufficient to increase the fiscal deficit by 1 percent of GDP in the first year.

Fiscal deficit is assumed to be money-financed.

Chart 1.A Sustained Increase in Government Expenditure

(Divergences from baseline)

In the medium term, the effects of the initial fiscal stimulus on industrial countries’ income die away, as economic activity moves back in line with underlying supply conditions. Government debt and government interest payments in industrial countries continue to mount and exert a significant effect on consumption, reducing the level of savings available to finance the deficit at current interest rates. At the same time, the reduction in capacity output caused by the reduced investment results in a fall in real GDP. Both these factors increase the global real interest rate (2.6 percentage points from the baseline) and weaken the demand for imports from developing countries. Consequent further compression of imports in developing countries reinforces the weakness of exports and significantly reduces the growth of real GDP (by 1.3 percent).

b. Effects on Fiscal Deficit of Developing Countries

In the medium term, the overall deficit of developing countries deteriorates by 0.6-0.8 percent of GDP. Although the ultimate effect on the fiscal balance is broadly similar whether the deficits are financed by additional bank credit or by debt financing, the channels are somewhat different. In the money-financed case, real GDP falls by a large amount; the bulk of the increase in the deficit is due to an increase in expenditure in relation to GDP (as expenditure is assumed to be maintained in real terms). The increase in interest payments and the decrease in revenue contribute moderately toward fiscal deterioration in relation to GDP.

In the case of debt financing, lower recourse to monetary financing results in lower price inflation and increasing export profitability. Given the current account constraint and the assumption of a fixed nominal exchange rate, this implies somewhat higher imports and real GDP and, therefore, higher tax revenues. However, these favorable developments are more than offset by the growth in government debt, which results in rapid growth in interest payments; thus, the medium-term fiscal outturn of a debt-financed deficit is broadly similar to that of a money-financed deficit.

c. Sensitivity to Changes in the Values of Key Parameters

Sensitivity analyses based on changes in six important parameters, and expressed in terms of the divergences from the baseline scenario, show that changes in the values of three key parameters—the interest elasticity of consumption, the Ricardian constant, and the interest elasticity of investment—can have significant quantitative effects on the results (Table 4). The impact of changes in the first two of these parameters increases over time, as the effect of the fiscal expansion on wealth and disposable income increases. If the interest elasticity of consumption is zero, the entire burden of adjustment has to be borne by investment, requiring a very large increase in the interest rate. Changes in the interest elasticity of investment, however, have diminishing effects over time, since neither wealth nor disposable income directly influences investment in the model.

Table 4.Sensitivity Analyses of Changes in Key Parameters for Developing Countries
Short-Term Effect1Medium-Term Effect2




Change long-term interest elasticity of consumption from
−0.8 to 03.2−1.10.615.4−11.24.5
−0.8 to −0.51.8−−2.00.8
−0.8 to −2.5−0.7−−0.40.2
Change interest elasticity of
investment from
−0.8 to −0.41.8−−1.50.6
Reduce Ricardian constant from
0.5 to 0.251.3−−1.80.8
Eliminate import constraint on developing country exports1.4−−1.00.5
Relax current account constraint by 2 percent of imports1.80.7−0.12.3−1.30.6
Increase original debt stock by 25 percent1.4−−1.50.6

In the first year.

By the eighth year.

In the first year.

By the eighth year.

Relaxing the current account constraint would, of course, significantly enhance growth in the short term; in the longer term, however, the favorable effect would be substantially reduced as a result of increased interest payments on the higher level of foreign debt if accommodating finance is not available. Finally, the greater the outstanding foreign debt of developing countries, the larger the reduction in growth caused by fiscal expansion in industrial countries. The impact of higher interest payments on real interest rates is negligible.

Overall, the results are qualitatively, if not always quantitatively, robust with respect to changes in parameter values within empirically plausible ranges.

2. Transmission of Fiscal Shocks with Exchange and Interest Rate “Flexibility” in Developing Countries

A flexible domestic interest rate policy is one that allows for interest rates to be ajusted upward or downward from the baseline level in response to changes in inflation, effectively maintaining the real interest rate in each period. Similarly, a flexible exchange rate policy implies adjustments in the exchange rate to bring about the required strengthening of the trade balance in the face of higher interest payments, given the current account constraint.

Devaluation of the exchange rate in developing countries boosts exports, reduces the degree of import compression, and reduces the fall in real GDP by approximately one half (Chart 2). The effects of flexible interest and exchange rates on the fiscal deficit of developing countries are, however, not completely clear.17 Increases in the deficit in the initial years are due to sharply higher interest payments on foreign and domestic public debt resulting from the exchange rate depreciation and a higher nominal interest rate. In the medium term, in either the money-financed or debt-financed case, the overall fiscal deficit increases marginally less than in the fixed-price scenarios described earlier. Deficit financing through domestic bank borrowing increases inflation and requires a much larger nominal devaluation to achieve the necessary real depreciation of the exchange rate. In the bond-financing case, inflation is lower and, consequently, the domestic interest rates and the fiscal deficit are initially lower. In the medium term, however, government debt and interest payments increase rapidly in relation to GDP, with consequent adverse effects on the fiscal deficit.

Chart 2.A Sustained Increase in Government Expenditure with Flexible Exchange and Interest Rates in Developing Countries

(Divergences from baseline)

A policy of maintaining the domestic interest rate unchanged in nominal terms may improve the fiscal deficit in the short term. However, in the medium term, this policy would reduce the attractiveness of government bonds to the private sector and would eventually imply a financing of the deficit through inflationary means, with all the disadvantages that would entail.

3. Effect of Maintaining Industrial Countries’ Fiscal Deficit at the 1977 Level

Finally, we consider the outcome of maintaining the industrial countries’ fiscal deficit at its 1977 level (approximately 3 percent of GDP), assuming policies in developing countries remain fixed as defined in Subsection IV.1.18 In the first few years, the effect would have been minimal, since the deficit of industrial countries remained stable in relation to GDP (Chart 3). From 1980 onward, however, containment of the fiscal deficit in industrial countries would have significantly lowered the global interest rate as well as the fiscal deficit of developing countries.

Chart 3.Maintaining Industrial Countries’ Fiscal Deficit as a Percentage of GDP at Its 1977 Level

(Divergences from baseline)

The simulations suggest that real GDP in the industrial countries would have fallen by an average of about ⅓ of 1 percent during 1980-83; real interest rates would have been 1 percentage point lower in 1981, and nearly 3 percentage points lower in 1983 (when the actual deficit peaked). Developing country interest payments would have fallen sharply, allowing imports to increase by 4 percent in real terms by 1984, and real GDP could have been about 0.8 percent higher than in the baseline.

The fiscal situation of developing countries also would have improved by about 0.4 percent of GDP by the end of the period. In all, about one fifth of the increase in the overall fiscal deficit of developing countries between 1977 and 1984 could be ascribed to the fiscal expansion in industrial countries.

V. Conclusions

Both the analytical and empirical analyses presented in this paper suggest that, for plausible values of key parameters, an increase in the fiscal deficit of industrial countries causes higher real interest rates, leading to an increase in external debt-service payments for developing countries. Where developing countries face a current account constraint, these higher debt-service payments generally more than offset the boost to exports caused by high demand in industrial countries, and result in import compression and slower growth. The fiscal deficit of developing countries also tends to increase, owing to lower revenue (from import duties and domestic-based taxes) and higher outlays (in relation to GDP).

The quantitative effects, of course, vary considerably among different economies, depending, inter alia, on the size of their external debt, the dependence of their outputs on imports, and the share of taxes coming from international trade. In the short run, the heavily indebted countries are likely to be far more adversely affected than those countries whose commercial borrowing from external sources is limited and whose external financing comes mostly from official or multilateral sources. In the aggregate, however, our simulations suggest that if industrial countries had held their fiscal deficits at the level of 1977, the fiscal deficit in developing countries would have been almost ½ of 1 percent of GDP lower by 1983-84. The simulations also indicate that if the exchange rate is allowed to adjust in response to external developments, the adverse effects on both output and the budget deficit would be sharply lower in the medium term, compared with the fixed-price scenario.

However, it should be emphasized that the higher fiscal deficit in industrial countries can explain only about one fifth of the fiscal deterioration in developing countries in recent years. In many developing countries, the increase in public spending was due to investments in inappropriate or unsuccessful projects, the continued operation of inefficient projects and enterprises originating from past investments, or a general increase in public sector operations with social progress without a commensurate increase in the revenue base. Other external shocks—such as terms of trade deterioration (apart from the fiscal deficit-induced terms of trade effect discussed in this paper), economic recession in the industrial countries, and trade restrictions—also led to slower growth and a higher fiscal deficit during the same period. Inadequate adjustments to external shocks or antirecessionary policies in many developing countries might have contributed further to the higher fiscal deficit.

A number of policy recommendations follow from our analysis. First, reductions in the fiscal deficits of industrial countries are likely to improve the growth prospects and fiscal position of developing countries, because lower global interest rates will ease the availability of imports. This conclusion differs somewhat from the standard analysis, which suggests that higher fiscal deficits in industrial countries stimulate growth in developing countries; although the standard open-economy models emphasize the transmission of output effect, this paper also highlights the role of the interest rate, the burden of external debt, and the current account constraint. We are aware that this paper does not exhaustively cover all the channels through which international transmission may take place; however, we have focused our analysis on the important ones based on stylized facts.

Second, when developing countries face a rigid current account constraint, as we have assumed in this paper, increases in interest rates are particularly important in transmitting shocks from industrial to developing countries. The latter would benefit substantially from policies aimed at lowering the global interest rate and from measures to increase capital flows to them. Third, the impact of higher fiscal deficits in industrial countries on developing countries is partly mitigated if the latter follow a flexible exchange rate policy.


I. An Analytical Exposition Of The Model

The total differential of a simplified version of the system of equations (1)-(4) can be expressed in matrix form as

where a subscript (i) to a function (F) denotes differentiation of that function with respect to that variable (i.e., Fi = ∂F/δi).

Given the simplified structure of the model, the first two equations determining output and interest rate in industrial countries are simultaneous. Once these two equations are solved, the equilibrium values for Y and R can be substituted in equations (3) and (4), along with the current account constraint, to solve for the relative price and output level in developing countries.

The total differential of the subsystem represented by equations (1) and (2) is

The determinant of the coefficient matrix, Δ, is

which, given the normal assumptions about partial derivatives, is of an indeterminate sign. If the sensitivity of absorption with respect to interest is small, Δ would be positive. A change in the fiscal deficit of the industrial countries has the following effects on the interest rate and output in the industrial countries:

As expected, both the interest rate and output effects of an increase in the fiscal deficit are ambiguous, even if we assume Δ > 0. However, output in industrial countries would expand if the sensitivity of absorption with respect to fiscal deficit is high and the interest sensitivity of domestic absorption is relatively small. Moreover, the output effect would be larger, the greater the Barro-Ricardo effect, which would tend to dampen the effect on the interest rate. If dR/dD > 0, the effect of fiscal expansion on the interest rate would be higher, the smaller is θ and the less the sensitivity of absorption to fiscal deficit. Thus, the conditions under which an expansionary fiscal policy leads to a smaller increase in the interest rate would also lead to a greater expansionary effect on output.

The effect of a change in the fiscal deficit on the relative price (of developing countries’ output in terms of industrial countries’ output) may be observed by taking the total differential of equation (3)

If we assume that dY/dD and dR/dD are both positive, then the relative price would decrease or increase, depending on the relative strength of the negative effect from the interest rate increase and the positive effect from output expansion in industrial countries. On the one hand, an increase in the interest rate would cause higher interest payments for developing countries, and lead to a compression of imports through a reduction in the relative price of their output in terms of industrial countries’ output. On the other hand, an increase in output in industrial countries would allow for more exports from developing countries, leading to an easing of import availability through an increase in the relative price, given the current account constraint. If the interest rate effect dominates, dp/dD < 0—that is, the relative price or the terms of trade would deteriorate for developing countries; the reverse would happen if the output effect dominates.

The effect of a change in the fiscal deficit on output in developing countries can similarly be expressed as

Once again, the direction of change in output in developing countries is ambiguous, and would remain so even if we assumed (as is likely) that both dY/dD and dR/dD > 0, and dp/dD < 0. An expansion of output in industrial countries resulting from a higher fiscal deficit would increase the exports of developing countries and, through the current account constraint, also increase their imports and output. An increase in the interest rate and a deterioration in the terms of trade would both reduce the imports of developing countries and contribute to a reduction in their output. Thus, an expansion in the fiscal deficit may easily contribute to a reduction in output in developing countries through a higher interest rate.

The effects of an increase in industrial countries’ fiscal deficit on the global real interest rate, on relative price, and on the levels of economic activity in the two groups of countries may be illustrated in terms of a four-quadrant diagram (Figure 1). In quadrant I (the upper right-hand panel), we show the locus of combinations of real interest rates and output of the industrial countries (SI) which, for a given public sector fiscal position ID), equate the ex ante private savings-investment balance with the ex ante current account balance for industrial countries. The SI locus slopes downward because of our assumption that a rise in the interest rate causes the desired investment to fall relative to intended savings, leading to an improvement in the current account balance in real terms; given the constraint on the current account, output must decline to ensure equality in the new desired pattern of savings and investment.

Figure 1.Effects of a Change in Fiscal Deficit on Output, Interest Rate, and Exchange Rate

The locus of the constrained current account balance, for a given level of the relative price, CA¯(p0) slopes upward in (Y, R) space. The SI locus for industrial countries and the locus of current account constraint for developing countries, given the initial fiscal balance (D0), interest rate (R0), and relative price (p0), are shown in (Y, Y*) space in quadrant IV. Analogously, the saving-investment locus for a given budget deficit and output in industrial countries and the constrained current account balance for developing countries (given industrial countries’ output level) are shown in quadrant II (the lower right-hand panel). The initial equilibrium interest rate (R0), the relative price (p0), and the output levels in the two groups of countries (Y0 and Y0*) are characterized by the intersection of the two loci in each quadrant (at points shown as A0) for a given level of the fiscal deficit in industrial countries.

This diagrammatic presentation, notwithstanding the underlying simplifications, enables us to capture simultaneously the effects of a change in fiscal deficit or of shifts in other exogenous variables in industrial countries on the real interest rate, relative price, and output levels. Suppose that the industrial countries experience an expansionary fiscal policy or an autonomous increase in private sector investment, or some combination of the two. The effects of this increase are illustrated by shifts in the SI and CA¯ curves in all three quadrants. Although the shift in each quadrant could be described in several stages, we show only the short-run final positions to reveal the analytical conclusions.

Starting from the first quadrant, the expansionary fiscal policy shifts the SI(D0) locus to SI’(D1) showing an increase in real interest rates to R1 and an expansion of output of industrial countries to Y1 from R0 and Y0, respectively. In the fourth quadrant, the SI(D0, Y0) locus would correspondingly shift to S/’(D1, Y1), and the locus of the current account constraint would shift to CA¯(Y1) the equilibrium relative price corresponding to the interest rate R1 would decline to p1. The new equilibrium levels of income would be determined in quadrant IV at A1 or A2 the intersection of SI’(D1, p1,R1) with CA1¯(D1,p1,R1) or with CA2¯(D1,p1,R1). Thus, under this scenario, the new short-run equilibrium would involve a higher world interest rate, a deterioration of developing countries’ terms of trade, and an expansion of output in industrial countries. The new short-run equilibrium, however, would not necessarily imply an increase in output for developing countries. If the strength of the output expansion in industrial countries is offset by a higher interest rate, the secondary effects of the output expansion on the developing countries would easily be outweighed by the negative effects of the higher interest rate; the CA¯(D0,p0,R0) curve in quadrant IV may shift rightward to, say, CA1¯(D1,p1,R1). For heavily indebted developing countries, such an adverse shift is certainly possible, and would result in a decline in their output, as shown at the intersection of CA1¯(D1,p1,R1) and SI’(D1, p1,R1) in quadrant IV.

II. The Simulation Model

The basic model consists of the following equations:19

Industrial Countries

Fiscal Policy

Real Sector

Financial Sector

Real-External Identity

External Sector

Developing Countries

Real Sector and Prices

Fiscal Policy

Monetary Policy

External Sector

In addition, the following equation for developing country import demand was used in the simulations in which the exchange rate was allowed to float:

Rest of the World

External Sector

A variable coefficient was added into each equation to ensure that it replicated the baseline data.

Definition of Variables and Coefficients

The following key defines the variables used in the model. Unless otherwise indicated, the variables with asterisks refer to developing countries; the suffix R indicates that the variable is in real terms; the suffix L indicates that the variable is in the currency of the developing country; and the suffix OTR indicates that the variable refers to the rest of the world.

CA= current account deficit
CAOTR= current account deficit of the rest of the world (including the global discrepancy)
CP= private consumption
D= government deficit
E= exchange rate: units of developing countries’ currency per unit of industrial countries’ currency
F= proportion of foreign debt at fixed interest rates
FA= foreign assets
FINT= foreign interest payments
G= government expenditure and net lending
GDEBT= government debt
GDEBTD= government domestic debt
GDEBTF= government foreign debt
GDEFR= real government deficit (increase in real government debt)
GEXPO= government expenditure on goods and services (excluding interest payments)
GGRANT= government grants
GINT= government interest payments
GREV= government revenue
GREVO= government revenue from nontrade taxes
GREVT= government revenue from trade taxes
I= investment
IMP= imports
IMP*= total exports to developing countries
K= capital stock
M= imports of industrial countries from developing countries
MON= broad money stock
MOND= money demand
MONFIN= financing of the deficit through bank borrowing
MP= import prices
MV= import volume
P= consumer price index
R= real interest rate, defined as nominal rate less actual inflation
RF= average interest rate on fixed-rate foreign debt
RN= nominal interest rate
RNC= concessional interest rate on developing country borrowing
UT= unrequited transfers
W= wealth
X= exports of industrial countries to developing countries
Y= gross domestic product (real)
YC= capacity output (nominal)
YDIS= disposable income (nominal)
YGNP= gross national product
YN= gross domestic product (nominal)
θ= Ricardian coefficient
θ1= proportion of foreign debt held on concessional terms
θ2= rate of depreciation

III. Values of Parameters Used in Simulations and Data Sources

The behavioral equations and estimates for their parameters are derived from various sources, sometimes with minor adjustments. The parameters in equations (9)-(10), (14), and (32)-(33) are estimated directly from the data based on Masson and Knight (1986). The parameters in equation (14) are approximate averages of those derived in the paper for the United States, the Federal Republic of Germany, and Japan.

The relationship determining output in the industrial countries (equation (12)), as noted in the text, has essentially been improvised by the authors. It assumes that half of the difference between real and capacity output is eliminated within a year, but the speed of adjustment can be affected by changes in real government expenditure. (See Laidler and O’Shea (1980) for a similar fiscal specification; the value of A4 was taken from their estimates.)

Private sector consumption behavior (equation (15)), is based on the “no surprise” consumption function estimated for the United States by Blinder and Deaton (1985). Variables that are not relevant for the analysis in this paper were omitted.

The behavioral relationships determining developing countries’ export price (equation (24)), export volume (equation (25)), and imports (equation (46)) are based on Khan and Knight (1981).

The equation determining real output in developing countries (equation (28)) is based on Goldsborough and Zaidi (1986); their survey of the literature concludes that the value of the parameter A25 was between 0.14 and 0.28 for countries subject to foreign exchange rationing. The demand-for-money function (equation (40)) and the equation determining the price level (equation (41)) in developing countries are taken from Khan and Knight (1981).

The data used were largely taken from the following Fund publications: International Financial Statistics (various issues), Government Finance Statistics Yearbook (various issues), and World Economic Outlook (April 1987). Some stock figures were estimated by accumulating the relevant flows. The exchange rate E was derived as the ratio of an aggregate GDP deflator for developing countries and an estimated deflator in U.S. dollars (GDP in U.S. dollars divided by real GDP).


    BlanchardOlivierJ.“Debt, Deficits, and Finite Horizons,”Journal of Political Economy (Chicago) Vol. 93 (April1985) pp. 223-47.

    • Search Google Scholar
    • Export Citation

    BlinderAlanS. and AngusDeaton“The Time Series Consumption Function Revisited,”Brookings Papers on Economic Activity: 2 (1985) The Brookings Institution (Washington) pp. 465-521.

    • Search Google Scholar
    • Export Citation

    BovenbergA.Lans“Long-Term Interest Rates in the United States: An Empirical Analysis,”Staff PapersInternational Monetary Fund (Washington) Vol. 35 (June1988) pp. 382-90.

    • Search Google Scholar
    • Export Citation

    BuiterWillemH. and JamesTobin“Debt Neutrality: A Brief Review of Doctrine and Evidence,” in Social Security versus Private Savingsed. by GeorgeM. von Furstenberg (Cambridge, Massachussetts: Ballinger1979) pp. 39-63.

    • Search Google Scholar
    • Export Citation

    de LeeuwFrank and ThomasM. Holloway“The Measurement and Significance of the Cyclically Adjusted Federal Budget and Debt,”Journal of Money Credit and Banking (Columbus, Ohio) Vol. 17 (May1985) pp. 232-42.

    • Search Google Scholar
    • Export Citation

    GoldsboroughDavid and IqbalZaidi“Transmission of Economic Influences from Industrial to Developing Countries,” in International Monetary Fund Staff Studies for the World Economic Outlook (Washington1986) pp. 150-95.

    • Search Google Scholar
    • Export Citation

    KhanMohsinS. and MalcolmD. Knight“Stabilization Programs in Developing Countries: A Formal Framework,”Staff PapersInternational Monetary Fund (Washington) Vol. 28 (March1981) pp. 1-53.

    • Search Google Scholar
    • Export Citation

    KochinLewisA.“Are Future Taxes Anticipated by Consumers? Comment,”Journal of Money Credit and Banking (Columbus, Ohio) Vol. 6 (August1974) pp. 385-94.

    • Search Google Scholar
    • Export Citation

    LaidlerDavid and PatrickO’Shea“An Empirical Macro-model of an Open Economy under Fixed Exchange Rates: The United Kingdom, 1954-70,”Economica (London) Vol. 47 (May1980) pp. 141-58.

    • Search Google Scholar
    • Export Citation

    MassonPaulR. and MalcolmKnight“International Transmission of Fiscal Policies in Major Industrial Countries,”Staff PapersInternational Monetary Fund (Washington) Vol. 33 (September1986) pp. 387-438.

    • Search Google Scholar
    • Export Citation

    MullerPatrice and RobertPrice“Public Sector Indebtedness and Long-Term Interest Rates,”paper presented at the World Bank-Brookings Workshop (unpublished; WashingtonSeptember1984).

    • Search Google Scholar
    • Export Citation

    TannerJ.Ernest“An Empirical Investigation of Tax Discounting: A Comment,”Journal of Money Credit and Banking (Columbus, Ohio) Vol. 11 (May1979) pp. 214-18.

    • Search Google Scholar
    • Export Citation
1Large external borrowings to finance public sector development programs during the second half of the 1970s, and to pursue countercyclical fiscal policies in the early 1980s, led to a rapid accumulation of external debt by the developing countries. Total outstanding debt of the non-oil developing countries (excluding debt owed to the Fund) increased from less than $100 billion at the beginning of 1974 to more than $700 billion by the end of 1982, and during the same period, long-term debt in relation to gross domestic product (GDP), and to exports of goods and services, was doubled.
2Throughout this paper, the variables without asterisks refer to industrial countries, and those with asterisks refer to developing countries.
3Based on a simple model of aggregate consumption behavior with expected taxes and interest rates assumed to remain constant in the future, Blanchard (1985) and Masson and Knight (1986) have shown that θ should equal unity minus the ratio of the government’s discount rate to that of the private sector.
4The assumption that (SR - IR) > 0 is weaker than the assumptions SR> 0 and IR < 0. A higher real interest rate reduces consumption, given the rate of time preference and expected future wage income; however, since current income is increased for households holding positive net claims, the sign of SR may be ambiguous. Here, the weaker restriction implies that, if intended savings decline with higher interest rates, such savings fall by less than the intended investment.
5In line with the stylized facts of the early 1980s, this assumption reflects the severely limited availability of external financing from industrial countries.
6Empirical observations generally suggest that neither complete debt neutrality nor the full inclusion of government bonds in private net wealth is supported on the basis of the data; for more on empirical observations, see Kochin (1974), Tanner (1979), Buiter and Tobin (1979), and Masson and Knight (1986).
7The relationship between the fiscal deficit and interest rates was found to be significant for the United States by Muller and Price (1984), de Leeuw and Holloway (1985), and Bovenberg (1988).
8In the following years the debt-service ratio for the heavily indebted countries declined to around 40 percent, owing to rescheduling agreements with the creditors and to increases in exports resulting from the world economic recovery.
9Except that the empirical model has an equation for consumption rather than savings, since consumption functions have been more commonly estimated in econometric work.
10See Blinder and Deaton (1985) for a discussion of why the nominal, rather than the real, interest rate appears to matter. In our model, the two are effectively the same. Other variables that Blinder and Deaton find significant (for example, the relative price of consumer durables) have been omitted.
11Note that the income effect of higher interest rates on wealth holders is taken into account in the definition of disposable income, YDIS.
12See Masson and Knight (1986) for a detailed derivation of this formulation.
13The capital market in developing countries is assumed to be completely insulated from that of industrial countries, so developing countries’ interest rates are independent of those in industrial countries. In the alternative simulation, however, the domestic interest rate is assumed to change in line with domestic inflation.
14Not, however, by borrowing abroad, since we assume that foreign borrowing by developing countries is fixed.
15This implies that changes in nominal and real interest rates in industrial countries are identical.
16The actual increase in expenditure applied was about 0.8 percent of GDP. However, the ensuing rise in interest rates increased the ratio of interest payments to GDP, accounting for the remaining 0.2 percent.
17The quantitative results of the following simulations are subject to considerable margins of error, owing to a lack of information on the structure of government debt in developing countries. In the simulations, it is assumed that 50 percent of domestic government debt would be subject to the flexible interest rate policy.
18A reduction in industrial countries’ fiscal deficit would, under flexible policies, imply an appreciation of the exchange rate and lower domestic interest rates in developing countries. Since the exchange rate is likely to have been overvalued, and domestic interest rates too low, in developing countries, this would seem a perverse reaction, and the assumption of fixed policies seems more reasonable.
19The figures directly below the coefficients (within parentheses) indicate the values of the parameters used in the simulation exercises in the baseline scenario.

    Other Resources Citing This Publication