Abstract
This paper is an attempt to assess empirically the role of monetary factors in the process of short-run determination of output growth, inflation, and the trade balance for Japan during the period extending from the first quarter of 1965 to the fourth quarter of 1976. The theoretical framework is highly aggregated and focuses on the explanation of the behavior of the key economic variables. Despite its apparent simplicity, however, the model tracks the short-run movements of output, prices, and the trade balance well. The framework is essentially monetarist, in that the excess supply of real cash balances plays the leading role in the short-run adjustment process of the economy. As a first theoretical approximation, we envision a money-goods economy in which an excess supply of money, opposed by an excess demand for goods, results in higher output, higher prices, or deterioration of the trade balance. The adjustment of output, prices, and the trade balance is assumed to depend on the excess supply of money, the degree of openness of the economy, and the existing level of excess capacity.
This paper is an attempt to assess empirically the role of monetary factors in the process of short-run determination of output growth, inflation, and the trade balance for Japan during the period extending from the first quarter of 1965 to the fourth quarter of 1976. The theoretical framework is highly aggregated and focuses on the explanation of the behavior of the key economic variables. Despite its apparent simplicity, however, the model tracks the short-run movements of output, prices, and the trade balance well. The framework is essentially monetarist, in that the excess supply of real cash balances plays the leading role in the short-run adjustment process of the economy. As a first theoretical approximation, we envision a money-goods economy in which an excess supply of money, opposed by an excess demand for goods, results in higher output, higher prices, or deterioration of the trade balance. The adjustment of output, prices, and the trade balance is assumed to depend on the excess supply of money, the degree of openness of the economy, and the existing level of excess capacity.
The idea that an excess supply of money affects output and prices has received considerable attention in the literature, notably by Milton Friedman (1970). Friedman views the excess supply of money as one of the major factors determining changes in nominal spending, which, in turn, are divided between output and price changes, depending on the state of inflationary expectations and the current deviations of real output from its full-employment level. Friedman’s model, however, assumes a closed economy and therefore does not allow for the possible inflationary effects of foreign price changes or for the excess demand for goods to be partially channeled abroad through a balance of trade deficit. This last mechanism is the focus of the “monetary approach to the balance of payments,” 1 which views the trade balance as one of the main endogenous channels through which the public adjusts its actual cash balances in an open economy. 2 To date, however, most empirical works on the monetary approach have been conducted under the “full-employment” and “small country” 3 assumptions, which imply that output and prices are exogenously given and thus independent of domestic monetary developments. In this paper, we drop the full-employment assumption and recognize that at least some prices in the economy may not be immediately responsive to world prices. An excess supply of money thus has three “escape valves”—namely, raising domestic prices, increasing output, and diminishing the trade balance. Since monetary factors alone cannot explain the long-run real growth of an economy, we incorporate the concept of potential output, which is assumed to be independent of monetary factors, and concern ourselves with explaining the short-run behavior of the gap between actual and potential output.
The structure of the paper is as follows: Section I presents the basic structure of the model and estimating equations; Section II presents and discusses the results; Section III contains simulations under alternative assumptions regarding the conduct of monetary policy; and Section IV provides some concluding remarks.
I. Basic Structure of the Model
The basic theoretical premise of our model is that an excess supply of real cash balances will tend to increase prices, reduce the gap between actual and potential output, and reduce the trade surplus. The price and output effects of an excess supply of real cash balances are self-evident and have been widely dealt with in the closed-economy theoretical literature. In an open economy, part of the excess supply of money is likely to be channeled into additional expenditure on traded goods, diminishing the trade surplus. We now discuss in greater detail the derivation of the basic estimating equations.
dynamic adjustment of prices
We assume that Japan is a price taker in world markets for its imports, but it can affect the world prices of its exports. 4 We also assume the existence of an import-competing sector and a nontraded sector. Since the prices of the export sector and the nontraded sector are domestically determined, both sectors are aggregated into a “home sector” in which the nominal price is denoted by ph while the price in the import-competing sector is denoted by pm and varies directly with the world price of Japanese imports converted by the current exchange rate.
The aggregate price index can then be written as follows:
The parameter ϵ denotes the output or consumption share of home goods, depending on whether p is the gross national product (GNP) deflator or the consumer price index. From equation (1), the inflation rate, Δln p, is a weighted average of inflation in home goods, Δln ph and in imported goods prices, Δln pm:
Home goods prices are assumed to respond to the real excess supply of money and to the gap between the actual and potential output. To the extent that increased expenditure on goods owing to an excess supply of real balances m falls on home goods, their price tends to rise. Moreover, a larger excess of actual over potential output is expected to increase inflationary pressures on home goods in accordance with the standard “Phillips curve” arguments.
In order to keep the model simple, we specify the desired demand for real balances m * as the following function of income, y, where the income elasticity η is expected to be close to unity.
We can then postulate the following adjustment mechanism for home goods prices, when g t denotes the ratio of actual to potential output 5 and the subscript t refers to the time period.
The parameters α1 and α2 measure the speed of adjustment of home goods prices with respect to the excess supply of money and the output gap. Using equation (1), it follows that
Substituting m* and (ph/pm) from equations (3) and (5) into equation (4) and then substituting the result for Δln ph in equation (2), we get the following equation for the inflation rate:
where α0 = – ϵα1 ln k.
It could be argued, on theoretical grounds, that equation (6) should include a separate term to incorporate the effects of expected inflation on actual price changes, both directly through price markups and indirectly through its effects on money demand. Results from experiments incorporating an expected inflation term constructed according to an adaptive expectations scheme, however, did not prove satisfactory, and the expected inflation variable was therefore excluded. It should be noted that, for most of the estimation period, inflation fluctuated with no apparent trend and remained below 10 per cent, except for the unexpected price increases associated with the commodities boom and oil crisis of 1973–74 (See Chart 1.). Given the fluctuating behavior of the inflation rate, it is likely that expected inflation was approximately constant during the period, providing a possible explanation for the failure of the expected inflation variable constructed according to the adaptive scheme to capture the relevant inflationary expectations of the public. In this case, the effects of expected inflation on actual price changes, both through markups and through changes in the demand for money, would be incorporated into the constant term of the equation.
Japan: Actual and Predicted Inflation Rate, 1965–76, and Simulated Inflation Rate, 1 1977–82
Per cent
Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A002
1 SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation. See Section III of the text for further details.Equation (6) incorporates all three major sources of inflation discussed in the literature—namely, monetary, real, and external sources. The impact of a rise in import prices on domestic inflation is proportional to the share of the import-competing goods in domestic output (1–ϵ). Import prices are assumed to be exogenously determined, based on the assumption of an exogenous foreign currency price for imports at a fixed exchange rate up to 1971 and an exogenously-determined exchange rate after 1971 under a regime of controlled floating.
dynamic adjustment of output
The gap between actual and potential output is assumed to respond positively to the level of excess demand in the goods market, which is generated by the excess supply of money. In addition, for any given excess supply of money, the rate of adjustment of actual output toward its potential level is assumed to depend on the excess capacity present in the economy. 6
The above self-adjusting mechanism of output toward its potential level is more in line with the monetarist formulation, as opposed to the Keynesian view, which holds that output is demand-determined and is prevented from returning to its potential level owing to structural rigidities in the economy. Thus, the estimated value of the parameter β2 will be a suggestive empirical test for the presence of the self-correcting forces in the process of output determination, which are implied by the monetarist view. Moreover, the estimated size of the coefficient β1 will indicate to what extent an excess supply of money induces an increase in the rate of growth of output in the short run.
Using the transformation ln y = ln g + ln y*, equation (7) can be transformed into the following relationship determining the size of the gap between the actual and potential output y*.7
where
The above equation is quite similar to the one postulated by Friedman (1970), and the speed-of-adjustment coefficients, β1 and β2, indicate the relative importance of the excess supply of money and the self-correcting mechanism of real output in bringing the economy to its potential level of output.
dynamic adjustment of the trade balance
As mentioned in the introduction, an excess supply of money results in an excess demand for traded goods and, therefore, in a deterioration in the trade balance. However, the closer the economy is to full capacity, the higher is the trade deficit associated with a given excess supply of money, since more of the required adjustment in the economy takes place through the domestic inflation (as implied by equation (4)) and the external sector. Consequently, the trade balance is a negative function of the excess supply of money and the output gap. The trade balance is expressed in units of domestic goods and formulated in levels rather than in logarithms. 8 In addition to the excess supply of money and the output gap, we have also included the terms of trade in the trade balance equation, which is specified as follows:
where γ1 and γ2 capture the effects of the excess supply of money and the degree of excess capacity of the economy on the trade balance, respectively. Similarly, γ3 captures the effect of the current value of the terms of trade, and its sign is not determined on theoretical grounds. 9
II. Description of Data and Estimation Results
The model has been estimated using quarterly data for Japan for the period extending from the first quarter of 1965 to the fourth quarter of 1976. We have chosen 1965 as the starting year for the estimation because Japanese monetary data were revised in 1964, and the earlier data are not comparable with the post-1964 period. The money supply is defined as total liquidity (currency plus demand deposits and quasi-money). Because the main purpose of this paper is to analyze the division of nominal income changes into price and output changes, the GNP deflator is used to deflate all nominal variables, and the inflation rate is defined in terms of this index. Potential output is estimated based on indices of actual and potential manufacturing output that have been estimated by Artus (1977) utilizing the production function approach. A more detailed description of the data may be found in the Appendix.
The complete model is composed of equations (6), (8), and (9), which determine the rate of inflation, the output gap, and the trade balance, respectively. For purposes of exposition, these equations are reproduced here.
Output gap
Inflation rate
Trade balance
The above system of equations is nonlinear in variables and is therefore difficult to estimate simultaneously. 10 However, in the above specification, the trade balance does not appear in the first two equations, and the rate of inflation does not appear in the output gap equation. The system is therefore recursive. The ordinary least-squares (OLS) method of estimation can be used to provide unbiased estimates for the parameters of the system. All equations are corrected for autocorrelation using the Cochrane-Orcutt iteration method, and the t-values (in parentheses) are given underneath the coefficients.
Output gap
Inflation rate
Trade balance
As the estimation results indicate, the model explains the movements of inflation, the output gap, and the trade balance quite well. It is particularly interesting to note that the values of R2 coefficients are quite high, even though all of the endogenous variables exhibit a large degree of fluctuation with no apparent trend in their movements (See Charts 1, 2, and 3.). All the estimated coefficients have the correct sign, and most are significant at the 0.025 confidence level. 11 The D-W statistics, after correction for first-order autocorrelation, still indicate the presence of higher-order serial correlation in the output gap and trade balance equations; correcting for second-order autocorrelation, however, does not change the result appreciably. 12 The assumption of unit income elasticity of the desired demand for real balances (i.e., η=1) implies that the first two estimated coefficients in the output gap and the inflation equations should have opposite signs and be equal in absolute value. The estimated coefficients of real balances and output in the inflation equation are practically identical. While the other set of coefficients in the output gap equation are not as close to each other, they are not significantly different from each other. 13 Consequently, in order to estimate the speed-of-adjustment parameters of the model, the above constraints were imposed on the system, and each equation was estimated nonlinearly. The results are summarized in Table 1, and, as the t-values (in parentheses) underneath the coefficients indicate, all parameter estimates have the correct sign and are significant at the 0.05 confidence level.
Japan: Actual and Predicted Output Gap, 1965–76, and Simulated Output Gap, 1 1977–82
Per cent
Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A002
1 SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation. See Section III of the text for further details.Japan: Actual and Predicted Trade Balance, 1965–76, and Simulated Trade Balance, 1 1977–82
Billions of 1975 yen
Citation: IMF Staff Papers 1979, 001; 10.5089/9781451956528.024.A002
1 SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation. See Section III of the text for further details.Estimates of the Speed-of-Adjustment Parameters of the Model 1
Figures in parentheses are t-ratios.
Estimates of the Speed-of-Adjustment Parameters of the Model 1
| Equations | ||
|---|---|---|
| Inflation rate | α1 = 0.33 (5.0) |
α2 = 0.23 (2.0) |
| Output gap | β1 = 0.31 (4.0) |
β2 = 0.32 (2.6) |
| Trade balance | γ1 = 0.14 (3.2) |
γ2 = 165.00 (3.3) |
Figures in parentheses are t-ratios.
Estimates of the Speed-of-Adjustment Parameters of the Model 1
| Equations | ||
|---|---|---|
| Inflation rate | α1 = 0.33 (5.0) |
α2 = 0.23 (2.0) |
| Output gap | β1 = 0.31 (4.0) |
β2 = 0.32 (2.6) |
| Trade balance | γ1 = 0.14 (3.2) |
γ2 = 165.00 (3.3) |
Figures in parentheses are t-ratios.
It is important to analyze the implications of these results for stabilization policy. The estimated values of coefficients α1, β1, and γ1 indicate that an excess supply of money has a pronounced impact on the inflation rate, the output gap, and the trade balance in the short run. Specifically, consider the impact of a 10 per cent excess supply of money over desired real balances. The coefficient α1 indicates that initially the prices of home goods will rise by approximately 3.3 percentage points. Given the share of home goods in the price index, ϵ, which is 0.84, the rise in the GNP deflator within the first quarter is about 2.4 percentage points. On the output side, the 10 per cent excess supply of money will increase the gap between actual output and potential output by 3.1 percentage points, as indicated by the estimated coefficient β1. Moreover, the coefficient γ1 indicates that 14 per cent of the excess supply of money will flow into the trade balance. 14
The estimated value of 0.23 for α2, the coefficient of the output gap in the inflation rate equation, shows that there is a strong “Phillips curve” relationship present, in the sense that inflation tends to rise faster whenever the economy is close to full capacity. The value of 0.32 for β2, the coefficient of the output gap in the output gap equation, shows that there is also a strong tendency for the output gap to close with a lag.
III. Dynamic Simulation of the Model
In order to illustrate the short-run, as well as the long-run, impact of the money supply on the rate of inflation, the output gap, and the trade balance, the complete system was dynamically simulated for the period 1977–82 under three different monetary rules of 10, 15, and 20 per cent rates of growth of the money supply. These three simulations are referred to as SIM 10, SIM 15, and SIM 20, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation. The only other exogenous variables in the model are the rates of growth of potential output and of import prices. We have assumed that potential output will grow at 5 per cent per annum. As regards the rate of growth of import prices, we have assumed that future variations in the exchange rate will exactly match the differential between the inflation rates of Japan and that of its trading partners. Thus, the rate of increase of imported goods prices (in yen) will be exactly equal to the rate of increase of home goods prices, and the terms of trade will not change. In the absence of any information regarding the future movements of Japan’s import prices and the exchange rate, these rather arbitrary assumptions are probably the most reasonable ones. It should be emphasized that, while these assumptions regarding the rate of growth of potential output and import prices have some impact on the long-term levels of the inflation rate, the output gap, and the trade balance, they do not alter the basic qualitative properties of the dynamic paths of adjustment of these variables, which are of more importance for our purposes. 15
The results of the three sets of policy simulations are presented in Table 2 and are plotted in Charts 1, 2, and 3. These policy simulations lead to some interesting results. First, the rate of inflation rises gradually toward its long-term value, which is given by the difference between the rate of monetary expansion and the growth of output. The adjustment of the inflation rate to the excess supply of money, however, is fairly sluggish, and it takes about three years before the inflation rate is close to its “steady state” value. The path of adjustment is direct, and there is no evidence of “overshooting.”
Japan: Simulation Results, 1977–82 1
SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation.
Japan: Simulation Results, 1977–82 1
| Year | Output Gap | Inflation Rate | Trade Balance | ||||||
|---|---|---|---|---|---|---|---|---|---|
| SIM 10 | SIM 15 | SIM 20 | SIM 10 | SIM 15 | SIM 20 | SIM 10 | SIM 15 | SIM 20 | |
per cent
|
billions of 1970 yen
|
||||||||
| 1977 | −6.7 | −5.9 | −5.2 | 6.0 | 6.6 | 7.2 | 1,630 | 1,339 | 1,041 |
| 1978 | −6.9 | −4.2 | −1.5 | 5.7 | 7.5 | 9.3 | 1,743 | 821 | −153 |
| 1979 | −7.3 | −3.0 | 0.7 | 5.5 | 8.2 | 10.5 | 1,865 | 421 | −1,136 |
| 1980 | −7.5 | −2.1 | 3.0 | 5.4 | 8.7 | 12.0 | 1,964 | 107 | 1,929 |
| 1981 | −7.7 | −1.4 | 4.9 | 5.3 | 9.0 | 12.8 | 2,045 | −143 | −2,572 |
| 1982 | −7.9 | −0.9 | 6.0 | 5.2 | 9.3 | 13.4 | 2,114 | −344 | −3,103 |
SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation.
Japan: Simulation Results, 1977–82 1
| Year | Output Gap | Inflation Rate | Trade Balance | ||||||
|---|---|---|---|---|---|---|---|---|---|
| SIM 10 | SIM 15 | SIM 20 | SIM 10 | SIM 15 | SIM 20 | SIM 10 | SIM 15 | SIM 20 | |
|
per cent
|
billions of 1970 yen
|
||||||||
| 1977 | −6.7 | −5.9 | −5.2 | 6.0 | 6.6 | 7.2 | 1,630 | 1,339 | 1,041 |
| 1978 | −6.9 | −4.2 | −1.5 | 5.7 | 7.5 | 9.3 | 1,743 | 821 | −153 |
| 1979 | −7.3 | −3.0 | 0.7 | 5.5 | 8.2 | 10.5 | 1,865 | 421 | −1,136 |
| 1980 | −7.5 | −2.1 | 3.0 | 5.4 | 8.7 | 12.0 | 1,964 | 107 | 1,929 |
| 1981 | −7.7 | −1.4 | 4.9 | 5.3 | 9.0 | 12.8 | 2,045 | −143 | −2,572 |
| 1982 | −7.9 | −0.9 | 6.0 | 5.2 | 9.3 | 13.4 | 2,114 | −344 | −3,103 |
SIM 10, SIM 15, and SIM 20 refer to dynamic simulations of the complete system for the period 1977–82, with the last two digits denoting the annual percentage rate of monetary expansion associated with the simulation.
The simulation results for the output gap confirm the importance of monetary policy in closing the gap in the short run. The results of SIM 10 indicate that a 10 per cent monetary expansion keeps the inflation rate at a relatively low 5 per cent, but the actual output growth rate is somewhat slower than the growth rate of potential output, leading to a marginal increase in the level of excess capacity. A 15 per cent rate of monetary expansion, however, reduces the output gap (its absolute value) gradually; although it is not eliminated completely by 1982, it is reduced to less than one per cent. In order to eliminate the gap altogether, the rate of monetary expansion must be set at 20 per cent; even so, it takes close to three years before output reaches its full-capacity level. A 20 per cent rate of monetary expansion, however, induces a 15 per cent inflation rate. As far as the trade balance is concerned, a 15 per cent rate of monetary expansion gradually eliminates the trade balance surpluses, while the 10 per cent and the 20 per cent rates lead to relatively large surpluses and deficits, respectively.
It should be noted that, in this paper, we have ignored the impact of the balance of payments on reserve money and the money supply. Insofar as the authorities control capital flows and sterilize a large portion of them, 16 this effect is negligible. Moreover, in Japan, the relationship between reserve money and the money supply is a tenuous one, since the Bank of Japan controls bank credit directly through “window guidance” and other controls. The money supply can therefore be treated as being more or less exogenously determined by the authorities.
While the above results indicate the short-run effects of monetary expansion on certain key economic variables, the longer-term implications should be considered more cautiously. Our framework is basically a short-term one, and it is not proper to draw strong conclusions based on simulations that are carried far into the future. The simulations for the five years 1977–82 do, however, demonstrate the basic dynamic properties of our model. Another important point is that the parameter estimates are based on an estimation period when monetary shocks were induced fairly randomly. It is not clear that the same parameter estimates can be used for a longer-term model when the rate of monetary expansion is set at a constant rate. Presumably, the public would learn to discount any set policy in the longer term and, as a result, the output gap could in the long run become independent of the rate of monetary expansion.
IV. Conclusion
The purpose of this paper has been to examine the effects of monetary changes in Japan on the inflation rate, the output gap, and the trade balance. The theoretical framework is an extension of Friedman’s (1970) analysis to an open economy. In Friedman’s formulation, any excess supply of money is reflected in an increase in inflation and output growth. In this paper, the excess supply of money may be reflected in the deterioration of the trade balance as well. The dynamic model is estimated for Japan for the period extending from the first quarter of 1965 to the fourth quarter of 1976. The model explains the movements of endogenous variables—namely, the inflation rate, the output gap, and the trade balance—quite well, despite the large disturbances associated with the world monetary system in the early seventies and the commodity boom and the “oil crisis” of the mid-seventies.
The estimated coefficients indicate that variations in the money supply have major stabilization effects. In addition, the level of excess capacity in the economy acts as an important self-equilibrating variable; a larger level of excess capacity is shown to reduce inflation and to stimulate the growth of output. Starting with the fairly large level of excess capacity in 1977, the model is simulated to show the reaction of output, prices, and the trade balance to various rates of monetary expansion. At one extreme, a rate of monetary expansion close to 20 per cent is required to close the output gap completely by 1980. Such an early achievement of full employment, however, would result in a substantial increase in the rate of inflation and a turnaround in the trade balance, which would move from the present large surplus to a deficit within a couple of years.
APPENDIX
All data used in this study were taken from various issues of the Fund’s monthly publication, International Financial Statistics (IFS). Variables and the corresponding line numbers in IFS, which refer to data on the pages of that publication that cover Japan, are as follows:
M = total liquidity = Line 31n + Line 32
y = output (GNP, 1970 prices) = Line 99a.r
p = GNP deflator (1970 = 100) = Line 99a divided by Line 99a.r
= value of exports = Line 70 = value of imports = Line 71 pm = unit value of imports (1970 = 100) = Line 75
B = trade balance in real terms = (
– )/p
All variables are seasonally adjusted, and all “real” variables are deflated by the GNP price deflator. Potential output y* was constructed based on the indices of actual industrial output I and potential industrial output I* provided by Artus (1977). Assuming that aggregate output is a log-linear function of I and the ratio of (I/I*), the following regression was carried out for the period extending from the first quarter of 1965 to the fourth quarter of 1976:
The second term in the regression indicates that aggregate output reacts to the movements of industrial output in a nonlinear fashion. Thus, a fall in industrial output will have relatively less impact on aggregate output when there are higher levels of excess capacity in the manufacturing sector. Using the relationship (10), potential output y* can be constructed by substituting I* for I,
BIBLIOGRAPHY
Aghevli, Bijan B., and Mohsin S. Khan, “Credit Policy and the Balance of Payments in Developing Countries”, paper presented at the SSRC-Ford Foundation Conference on “Macroeconomic Policy and Adjustment in Open Economies,” at Ware, England, April 28-May 1, 1976 (unpublished, International Monetary Fund, December 27, 1976).
Artus, Jacques R., “Measures of Potential Output in Manufacturing for Eight Industrial Countries, 1955–78,” Staff Papers, Vol. 24 (March 1977), pp. 1–35.
Blejer, M.I., “The Short-run Dynamics of Prices and the Balance of Payments,” American Economic Review, Vol. 67 (June 1977), pp. 419–28.
Borts, G.H., and J.A. Hanson, “The Government Budget and the Balance of Payments” (unpublished, Brown University, 1977).
Frenkel, Jacob A., and Harry G. Johnson, eds., The Monetary Approach to the Balance of Payments (London, 1976).
Friedman, Milton, “A Theoretical Framework for Monetary Analysis,” Journal of Political Economy, Vol. 78 (March/April 1970), pp. 193–238.
International Monetary Fund, The Monetary Approach to the Balance of Payments (Washington, 1977).
Mathieson, Donald J., “The Effects of Eliminating Capital Controls on Japanese Monetary Policy” (unpublished, International Monetary Fund, July 1, 1977).
Sargent, T.J., “A Classical Macroeconometric Model for the United States,” Journal of Political Economy, Vol. 84 (April 1976), pp. 207–37.
Stein, Jerome L., “Inflation, Employment and Stagflation,” Journal of Monetary Economics, Vol. 4 (April 1978), pp. 193–228.
Mr. Aghevli, Senior Economist in the Asian Department, received his doctorate from Brown University.
Mr. Rodriguez, Associate Professor at Columbia University, was on leave as a visiting scholar in the Asian Department of the Fund when this paper was prepared. He received his doctorate from the University of Chicago.
The authors are grateful to Larry Sjaastad for comments and suggestions and to Sharon Foley for efficient research assistance.
Some of the recent works in this area are collected in Frenkel and Johnson’s The Monetary Approach to the Balance of Payments (1976) and the Fund’s The Monetary Approach to the Balance of Payments (1977). See also Borts and Hanson (1977).
The other channel in this approach is the capital account of the balance of payments; throughout this paper, however, we are only concerned with explaining the trade balance. In a more general framework, one might also wish to explain the flow of international capital. In the case of Japan, however, there were strict controls on such flows of capital during our sample period (See Mathieson (1977).).
For an empirical evaluation of the impact of monetary policy on the trade balance in the developing countries, see Aghevli and Khan (1976).
While these assumptions do not strictly hold for every commodity, they have been utilized to keep the framework simple and manageable.
It should be noted that ln gt corresponds to the excess of actual over potential output in percentage terms.
For an alternative view of the determination of the output gap that utilizes the rational expectations framework, see Sargent (1976). See also Stein (1978) for an empirical test of Keynesian versus monetarist determination of inflation and output in the context of a closed economy.
This transformation is used in order to eliminate the output term, which is the endogenous variable from the right-hand side of the equation.
Since the trade balance often becomes negative, we could not use the log-linear formulation. The use of a linear term for output in equation (9) implies that we are constraining the income elasticity of money demand to be unity; such a constraint is warranted, however, since our results from the other two estimated equations strongly indicate an income elasticity not different from unity.
It should be noted that, in this context, a deterioration in the terms of trade is likely to reduce the volume of imports as well as that of exports, because the exportable surplus will be smaller. Moreover, the value of imports is likely to rise, at least in the short run (J-curve effect).
The first two equations are log-linear in variables, while the last equation is linear. There are some nonlinear estimation techniques available, but the statistical properties of the nonlinear estimators remain largely unknown.
The only exception is the coefficient of the output gap in the inflation equation, which is significant at the 0.05 level.
Correcting the equations for inflation and the trade balance for second-order autocorrelation gives the following results:
In order to determine whether the coefficient a is significantly different from β, the following expression, which is distributed as a t, is constructed:
It should be noted that since the trade balance equation is linear, the coefficient γ1 is not an elasticity and cannot, therefore, be compared directly to the coefficients α1 and α2, which are elasticities.
Additional simulations involving the assumption of 7 per cent growth in potential output and constant nominal import prices were also conducted, and no qualitative changes in the time paths of the variables were found. Quantitatively, however, the results are sensitive to alternative assumptions, and we emphasize that these simulations should not be interpreted as “forecasts” for the Japanese economy.
The financial developments that took place in 1972, when the authorities were not fully successful in controlling and neutralizing short-term speculative capital inflows, were an exception.
