This Selected Issues paper analyzes Japan’s medium- and long-term fiscal challenges. It discusses the initiatives that will be necessary to cope with Japan’s medium- and long-term fiscal challenges. It describes medium- and long-term projections of the fiscal position, and assesses the size of the consolidation measures needed to restore long-term fiscal sustainability. The paper explores possible options for consolidation and offers a representative package of such measures. The paper also describes the projected baseline path for the long-term fiscal balance, incorporating the long-term pension reform plan formulated in 1994.

Abstract

This Selected Issues paper analyzes Japan’s medium- and long-term fiscal challenges. It discusses the initiatives that will be necessary to cope with Japan’s medium- and long-term fiscal challenges. It describes medium- and long-term projections of the fiscal position, and assesses the size of the consolidation measures needed to restore long-term fiscal sustainability. The paper explores possible options for consolidation and offers a representative package of such measures. The paper also describes the projected baseline path for the long-term fiscal balance, incorporating the long-term pension reform plan formulated in 1994.

VI. External Adjustment in Japan: Recent Developments and the Medium-Term Outlook 1/

Japan’s persistent current account surpluses since the 1980s have been a source of considerable controversy. After eventually declining in the late 1980s following the substantial appreciation of the yen in 1985-1986, the surplus began to widen again in 1991. As the surplus continued to increase through 1993 while the yen rose, skepticism again arose about the role of traditional economic factors in explaining the surplus and, in particular, of the role of relative prices in affecting external adjustment. The rapid declines of the surplus during 1995 and so far in 1996, in spite of a substantial depreciation of the yen, has led various observers to forecast substantial continuing reductions in the surplus. The staff, on the other hand, is projecting that the surplus will soon bottom out and then increase.

This chapter reviews the historical experience of external adjustment in Japan since the early 1980s and explains the staff’s forecast of the external surplus. It is argued that the major swings in Japan’s current account can be identified with movements in the yen, relative activity, and movements in world oil and commodity prices, suggesting that traditional factors have played an important role in affecting the surplus. A traditional econometric model of the current account is shown to track well the major swings in the surplus over the historical period. Structural changes have, however, also played an important role in the most recent period. The prospects for continued structural changes, the projected recovery in domestic activity, and the lagged and projected effects of relative prices suggest, on net, that after declining through the first half of 1996 the surplus will increase modestly as a proportion of GDP over the medium term.

The chapter is organized as follows. Section 1 reviews historical developments in Japan’s current account surplus. Section 2 describes the staff’s current account model and discusses empirical estimates of the behavioral relationships in the model. Section 3 examines the performance of the model in explaining movements in trade flows and prices over the historical period, and decomposes prediction errors in the overall surplus into merchandise trade and services components, and exports and imports. Section 4 decomposes Japan’s historical external adjustment into that due to relative prices and demand. Section 5 explains the assumptions underlying the staff’s forecast of the current account and decomposes it into that due to relative price changes, the recovery in domestic demand, and structural changes in trade flows. Section 6 offers some concluding remarks.

1. Historical developments, 1981-95

The 1980s and early 1990s witnessed large swings in Japan’s external balance (Chart VI.1). While there is considerable controversy as to the quantitative importance of traditional channels of external adjustment in Japan, the major swings in Japan’s current account have, at least in their timing, been closely related to movements in the yen, relative demand, and movements in oil and world commodity prices.

Chart VI.1
Chart VI.1

Japan: Current Account Balance, 1981-2001: Actual, Predicted, and Forecast.

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

Note: A (four quarter) moving average of the current account balance is plotted to smooth out seasonally in the services components.

The current account rose from approximate balance in the early 1980s, in the wake of the 1979-80 oil-price shock, to a surplus of over 4 percent of GDP by the mid-1980s. This increase took place against the background of a depreciating yen and a sharp pick-up in partner country demand growth that far outpaced that in Japanese domestic demand (Chart VI.2). Following the sharp rise of the yen during 1985-86, by 1987 the surplus had begun to decline. The decline was given added impetus in the ensuing “bubble” years, as extraordinary growth in domestic demand fuelled a rapid increase in the demand for imports. Consequently, the surplus continued to decline steadily during the latter half of the 1980s, falling to 1 percent of GDP in 1990.

Chart VI.2
Chart VI.2

Japan: Relative Prices and Demand, 1981-2001

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

Starting in 1991, as the bubble burst, domestic activity slowed. This slowdown coincided with a pick-up in foreign activity, and the surplus began to rise again, exceeding 3 percent of GDP by 1993. During this period the yen appreciated persistently. While the appreciation checked export volume growth and encouraged imports, lowering the surplus, valuation (“J-curve”) effects increased dollar export prices before trade volumes responded, increasing the value of exports and the surplus.

As domestic demand growth bottomed out in 1993, while the yen remained at a relatively appreciated level, the surplus began to decline. The decline accelerated during the second half of 1995, and by the first quarter of 1996 the surplus had dropped to 1 1/4 percent of GDP. The sharp fall in the surplus since mid-1995 is attributable to a variety of factors that acted in concert: (i) the lagged effects of the yen’s appreciation through mid-1995 on trade volumes; (ii) J-curve effects from the yen’s depreciation since mid-1995; (iii) the further recovery of domestic demand; (iv) a slowing of foreign demand; (v) increases in oil and world commodity prices; and (vi) an acceleration of “structural changes” in trade flows.

While historical movements in Japan’s current account surplus have been closely linked to movements in relative prices and activity, structural changes in Japanese trade flows appear to have played an important role in recent years. These changes encompass a variety of factors including: the continued relocation of Japanese manufacturing facilities abroad, which have reduced exports from Japan and increased imports from Japanese enterprises abroad; inroads made by foreign manufacturers into the Japanese distribution system, spurred by the large appreciation of the yen through mid-1995, which widened price differentials between Japan and abroad, and increased the incentives for foreign manufacturers to incur the substantial fixed costs of investing in the Japanese distribution system; and shifts in Japanese consumer preferences, partly as a result of the recession, toward lower-priced (imported) goods.

2. The model

The staff’s model of Japan’s current account is based on the “elasticities approach.” 1/ Merchandise export and import volumes, and services receipts and payments are modelled as functions of relative prices and activity variables. Merchandise exports, comprised almost entirely of manufactures, are treated as an aggregate, while merchandise imports are disaggregated into: manufactures; food; raw materials; and mineral fuels. Services flows are disaggregated into transportation, travel, and ‘other’ services. Export prices (in dollars) are determined as a markup over input costs, and a time trend is used to capture the productivity growth differential between the Japanese exportables sector and the rest of the manufacturing sector. 2/ Import prices (in dollars) are modelled as functions of partner country export prices and world oil and commodity prices. Net investment income is modeled as the (exogenous) return on the net stock of foreign assets, which accumulates endogenously with the current account.

Table VI.1 reports the results of ordinary least squares estimates of a general error-correction form for each equation along with the implied long-run elasticities. 3/ The first panel of Table VI.1 shows that both relative prices and income have been statistically significant determinants—and in most cases strongly so—of merchandise trade volumes over the long run. The long-run relative price elasticity of Japan’s export volumes, estimated at 0.79, is consistent with, though slightly lower than, the average of several other studies of 0.92. 4/ The estimated elasticity of export volumes with respect to foreign activity, of 0.67, is substantially lower than that of other studies, which have tended to range between 1.5 and 2. This is, however, a direct consequence of the use of an index of world non-oil imports as the foreign activity variable as opposed to the use of a weighted average of partner country real GDP in other studies. 5/

Table VI.1.

Japan: Estimates of Trade Flows and Prices

(t-statistics in parentheses)

Merchandise Trade Volumes 1/

article image

Estimates were obtained over the sample period 1974Q1-1995Q2.

Table VI.1.

Japan: Estimates of Trade Flows and Prices

(t-statistics in parentheses)

Merchandise Trade Prices 1/

article image

Estimates were obtained over the sample period 1974Q1-1995Q2.

Table VI.1.

Japan: Estimates of Trade Flows and Prices

(t-statistics in parentheses)

Nonfactor Services 1/

article image

Sample was 1979Q1-1995Q2. All equations include a constant and seasonal dummies which are not reported.

Table VI.1.

Japan: Estimates of Trade Flows and Prices

Variable and data definitions

(natural logarithms)

article image

The estimates of relative price elasticities for Japan’s imports differ substantially among the components. While the price elasticity of manufactures slightly exceeds unity at 1.04, the other components are, as would be expected from the nature of the goods, considerably less responsive to relative price changes, with estimated elasticities of 0.35 for food, 0.22 for raw materials, and 0.15 for mineral fuels. Since most other studies have focussed on aggregate imports, these estimates are not directly comparable. Weighting the estimated elasticities of the components by their shares in Japan’s trade yields, however, an overall elasticity close to the average of 0.7 found in other studies. 1/ The elasticity of import volumes to domestic activity also varies widely across the components, ranging from 2.16 for manufactures to 0.23 for mineral fuels. Weighting the estimated elasticities of the components by their shares in Japan’s trade yields an overall elasticity of 1.7, which is somewhat higher than the average value of 1.1 found in other studies.

The second panel of Table VI.1 reports the estimates of the trade price equations. The deviation of export prices from domestic costs depends positively in the short run on competitor’s prices, implying an element of pricing-to-market. Over the long run, export prices diverge systematically from domestic costs in the manufacturing sector, as indicated by the significant coefficient on the time trend. The estimate implies that export prices have fallen relative to domestic costs at an annual average rate of 2 percent. All the import price equations have a significant negative coefficient on the long-run error correction terms which impose homogeneity of domestic prices with respect to foreign prices.

The third panel of Table VI.1 reports estimates for the services flows. These data suffer from a number of shortcomings and, in the event, a somewhat eclectic approach to modelling these flows was adopted. 2/ In the final specifications reported, long-run relative price effects were found for all the flows except transportation receipts, though they were statistically significant only for payments of transportation and travel. The coefficients representing the long-run effects of activity all have the expected sign and are statistically significant for four of the six equations—receipts from transportation and ‘other’ services, and for payments for travel and ‘other’ services.

3. Predictive performance of model

To examine the ability of the estimated relationships in explaining trade flows and prices over the historical period, Chart VI.3 compares dynamic forecasts from each of the estimated merchandise trade volume and price equations and an aggregate of services over the historical sample period with the actual data. 1/ The charts suggest that, in general, over most of the sample period, the estimated equations explain the levels of trade flows and prices reasonably well. An important exception is the behavior of import volumes of manufactures over the recent past, which have been systematically underpredicted over the last year and a half. 2/ The strong growth of import volumes of manufactures is associated with the process of structural changes in Japanese trade flows discussed above. It is, of course, straightforward to mechanically account for the process of structural change over the historical period by incorporating a dummy variable (or time trend) into the estimated equation. Since the objective is to use the model for forecasting, however, and since the process of structural change is still ongoing, this was not done. Instead, as discussed below, the “addfactors” in the import volume equation are used to introduce judgementally a forecast of structural changes to the import volumes of manufactures.

Chart VI.3
Chart VI.3

Japan: Actual and Predicted Export Volumes and Prices, 1974-2001

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

1/ Bank of Japan price indices are employed.
Chart VI.3
Chart VI.3

Japan: Actual and Predicted Import Volumes, 1974-2001

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

1/ Customs unit values are employed.
Chart VI.3
Chart VI.3

Japan: Actual and Predicted Import Prices, 1974-2001

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

1/ Customs unit values are employed.
Chart VI.3
Chart VI.3

Japan: Actual and Predicted Aggregated Service Flows, 1979-2001

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

The actual overall current account is compared with a dynamic forecast generated from simulating the full model over the historical sample period in Chart VI.1. The model tracks the actual current account balance reasonably well. It is notable that the model captures all the major swings in the current account during 1981-95, and that the turning points in the surplus in 1987 and 1991 predicted by the model are quite closely aligned with the turning points in the actual surplus. The persistence of the deviations of predicted values of the surplus from actual values reflects in part the fact that in making a dynamic forecast, errors in prediction are carried forward and does not necessarily represent any systematic bias in the model. Chart VI.4 presents the prediction errors in the current account and breaks them down into broad components. Since the largest contributor to the overall current account surplus is the trade surplus, it is not surprising that the prediction errors are closely aligned. A breakdown of the prediction errors in the merchandise trade balance into those due to exports and imports reveals no systematic relationship between the two over the historical period. Chart VI.4 reveals that the overprediction of the current account surplus since 1995 largely reflects an underprediction of imports.

Chart VI.4
Chart VI.4

Japan: Prediction Errors in Current Account Balance, 1981-1996Q1

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

Note: Data are not smoothed.

4. Decomposing Japanese external adjustment, 1981-96

To delineate the effects of exchange rate changes and the associated change in relative prices on the external surplus, the model was simulated from the first quarter of 1981 through the first quarter of 1996 with the nominal exchange rate maintained constant at its initial level. 1/ The top panel of Chart VI.5 compares the simulated current account with the actual historical data, while the lower panel reports the difference of the two as the contribution of exchange rate changes (since 1981) to the level of the external surplus. 2/ The second panel shows that the sharp appreciation of the yen during 1985-86 raised the surplus during 1986-1988, as J-curve effects outweighed the effects on trade volumes, which responded only gradually to changes in relative prices. By 1988, however, the response of trade volumes offset these effects and the net effect was a lower current account surplus. The lower panel also reveals that a substantial amount of the decline in the surplus since mid-1995 can be attributed to the sharp depreciation of the yen, again due to the J-curve effect.

Chart VI.5
Chart VI.5

Japan: Decomposition of Current Account Balance, 1981-1996Q1

Citation: IMF Staff Country Reports 1996, 114; 10.5089/9781451820522.002.A006

Note: A (four quarter) moving average of the current account is plotted to smooth out the seasonality in the services components.

To delineate the role of cyclical movements in domestic and foreign activity in affecting the external surplus, the model was simulated over the historical period with each growing at their trend levels. Specifically, domestic demand was assumed to grow at 3.1 percent and growth in world activity, as measured by non-oil imports, at 8 percent. Chart VI.4 suggests that up until 1989, relative cyclical positions had little net impact on the surplus. While foreign demand growth rose substantially in 1984, the decline was short-lived. The sharp increase in foreign demand in 1987-88 coincided with a sharp increase in domestic demand and again, relative cyclical positions appear to have played little role. Since 1992, however, as the decline in domestic demand growth coincided with a pickup in foreign demand growth, relative cyclical positions acted to significantly raise the surplus. Relative cyclical positions continued to raise the surplus steadily through the last quarter of 1995, and then levelled off in the first quarter of 1996.

5. The medium-term forecast

a. Assumptions and inputs

To forecast the current account using the model described above requires three types of inputs. First, projections and assumptions are required about the external environment. These are taken from the World Economic Outlook (WEO). These include projections of partner country growth, inflation, export prices, world commodity and oil prices, and an assumption about the exchange rate. A critical assumption in the WEO forecast is the behavior of the exchange rate. The WEO assumes unchanged real exchange rates over the projection period, where real exchange rates are defined in terms of relative unit labor costs in manufacturing. Second, projections for domestic variables such as GDP growth, domestic demand, and inflation are required. These are taken from the most recent real sector forecast, which is produced separately. 1/ Third, judgement is used to adjust the mechanical “pure” model forecast for special factors by adjusting the residuals in the equations, that is the “addfactors”, over the forecast horizon.

The implications of the most recent assumptions in the WEO for the key relative prices affecting import and export volumes over the forecast horizon can be seen in the top panel of Chart VI.2. Since unit labor cost growth in Japanese manufacturing has in the past exceeded that in partner countries, and is projected to continue to do so over the forecast horizon, the assumption of constant relative unit labor costs implies that the nominal value of the yen appreciates at approximately 1 1/2 percent per year over the medium term. This rate of appreciation also approximately equals the differential between projected Japanese dollar import (foreign export) price inflation and the projected rate of Japanese domestic wholesale price inflation. Relative import prices are, therefore, projected to be flat over the forecast horizon. Relative export prices, on the other hand, are projected to decline over the medium term. This reflects, in the short run, the lagged effects of the yen’s depreciation since mid-1995 and, over the longer run, the trend decline in Japanese export prices relative to domestic costs in manufacturing. The projected decline in export prices relative to domestic costs implies a decline in Japanese export prices relative to those of partner countries over the forecast horizon.

Over the forecast horizon the growth of foreign activity, as measured by world non-oil imports, is projected to continue to slow through 1997 before settling down at its long-run growth rate. Domestic demand growth, after picking up sharply in late 1995, is expected to fall temporarily in late 1996 as the effects of the September 1995 fiscal package wear off, before settling down at its long-run growth rate.

There are substantial uncertainties in predicting the pace of continued structural changes and the impact of these effects on Japanese trade flows. There are factors that suggest that the process will slow down or even stop, and others that suggest it will continue, perhaps at the same rapid pace as in recent years. The investment decisions that caused many of these structural changes were spurred to a considerable extent by the massive appreciation of the yen between 1993 and mid-1995. On the one hand, with the yen having weakened substantially below the level at which many of these decisions were made, an acceleration of structural changes would appear unlikely, and it would be reasonable to conjecture a slowing. On the other hand, while there has been a notable relocation of Japanese manufacturing facilities abroad, the overseas production ratio of Japanese firms, that is the ratio of production abroad to total production, at slightly below 10 percent, remains well below that for other major industrial countries—about 20 percent for the United States, for example. Some observers have argued that an increase of the relatively low overseas production ratio of Japanese firms to levels of other countries should be expected as a natural consequence of the maturing of the Japanese economy. This suggests that structural changes could continue at a rapid pace.

Structural changes in imports, particularly of manufactured consumer goods, have been tied to changes in the Japanese distribution system. Anecdotal evidence suggests that domestic import prices have risen by less than the magnitude of exchange rate changes during the latest round of yen depreciation, as importers cut profit margins, and this has continued to spur imports. There are obviously limits to how far importers can lower profit margins, however, suggesting a deceleration in imports. The outlook for structural changes in trade flows is also inextricably tied to the pace of structural reforms, further increasing the uncertainty in projecting such changes.

The following assumption is made regarding the effects of structural changes on trade flows over the forecast horizon. These changes are assumed to continue to raise the growth rate of the import volume of manufactures during 1996 and 1997. The addfactors are then maintained as a relatively constant proportion of import volumes, so that while the structural changes are assumed to have a permanent effect on the level of trade flows, they do not have any effect on growth rates of import volumes after 1997. In addition it is assumed that structural changes depress export volume growth during 1996 and 1997, and then have a permanent depressing effect on the level of exports.

In summary, while relative import prices are projected to remain unchanged, relative export prices are projected to decline over the forecast horizon. Relative price movements are, therefore, projected to raise the surplus. The slowdown in foreign activity and the pickup in domestic activity will, on the other hand, act to lower the surplus. Finally, the assumed continuation of structural changes that boost import volumes of manufactures and depress export volumes will lower the surplus.

b. The forecast

The paths projected for trade flows and prices over the forecast horizon are presented in Chart VI.3, while that projected for the overall current account balance is presented in Chart VI.1. In nominal terms, after continuing to decline through the early part of 1996, the current account surplus turns around and rises steadily over the medium term. As a proportion of GDP, however, the turnaround in the surplus is modest, and it then remains relatively steady slightly below 2 percent of GDP over the medium term.

To quantify the roles of relative prices, relative activity, and assumed structural changes in affecting the surplus over the medium term, three separate forecasts were constructed and compared with the baseline forecast. First, a forecast was constructed where the exchange rate was assumed to appreciate at a faster rate over the forecast horizon than in the baseline forecast so as to maintain relative export prices constant over the medium term. 1/ Second, a forecast was constructed where domestic and foreign demand were assumed to grow over the forecast horizon at their average rates during 1995. Third, a forecast was constructed where the acceleration in the addfactors for import and export volumes attributed to further structural changes during the course of 1996 and 1997 did not occur. The results of these alternative forecasts imply that the “change” in the current account surplus as a proportion of GDP from 1995 to 2001 can be decomposed into those due to exchange rate effects, the dissipation of the cyclical component, and further structural changes and other factors. The breakdown, presented in the following tabulation, shows that over the medium term, relative price effects, which tend to raise the surplus, are more than offset by the return of activity to long-run levels and the further effects of structural changes on trade flows:

article image

c. Risks to the outlook

There are numerous sources of risk and uncertainty in the outlook. Among these, three are worth special mention. First, the volatility of the yen since the advent of floating rates suggests that there is always the possibility of further sharp movements in the yen which, the model suggests, would significantly impact the external forecast. Second, while the productivity growth differential between Japan’s export industries relative to the remainder of manufacturing has been a stable feature of the postwar period, this has no doubt been due in part to the “catch-up” in productivity of Japan’s export industries with those in the rest of the world. There is evidence that this “catch-up” has now been completed, and prospects for a continuation of the trend productivity differential are unclear. A break in the trend and a decline in this differential in the future would imply that the baseline forecast would, all else equal, overestimate the decline in the relative prices of Japanese exports and overestimate the surplus. Finally, as already emphasized earlier, there are considerable uncertainties in predicting the pace of continued structural changes in Japanese trade flows. Empirical efforts to model structural changes in Japanese trade flows by linking them, for example, to foreign direct investment flows have so far been unsuccessful, leaving a forecast of such changes as largely a judgemental exercise.

6. Conclusion

This chapter has shown that historical movements in Japan’s current account surplus are well explained by traditional factors—movements in the yen, cyclical factors affecting domestic and external activity, and changes in world oil and commodity prices. Over the more recent past, however, structural changes to trade flows resulting from a variety of forces have played an important role in reducing the surplus. While there are a number of uncertainties in the outlook, the prospects for continued structural changes, the projected recovery in domestic activity, and the lagged and projected effects of relative prices suggest that, on net, after declining through the first half of 1996, the surplus will increase modestly as a proportion of GDP over the medium term.

References

  • Corker, Robert,External Adjustment and the Strong Yen: Recent Japanese Experience”, IMF Staff Papers. Volume 36, No. 2, pages 464493, (June 1989).

    • Search Google Scholar
    • Export Citation
  • Goldstein, Maurice and Mohsin Khan,Income and Price effects in Foreign Trade”, in Handbook of International Economics, Volume 2, edited by Ronald Jones and Peter Kenen, North Holland, Amsterdam, pp. 10411105, (1985).

    • Search Google Scholar
    • Export Citation
  • “Japan—Background Papers”, Chapter I, “The Yen from a Long-Run Perspective,” SM/95/163, Supplement 1, July 1995.

  • “Japan—Background Papers”, Chapter VIII, “Exchange Rate Movements: The Role of Openness, Trade Responsiveness, and Pass-Through”, SM/95/163, Supplement 1, July 1995.

    • Search Google Scholar
    • Export Citation
  • Meredith, Guy,Revisiting Japan’s External Adjustment Since 1985,IMF Working Paper WP/93/52, Washington, D.C., (1993).

1/

This chapter was prepared by Bankim Chadha.

1/

The present model is based on Corker (1989). For a discussion of the conventional determinants of trade flows see Goldstein and Khan (1985).

2/

Long-term movements in the yen and the role of sectoral productivity differentials in generating these movements are discussed in Chapter I, and the pass-through of exchange rate changes to export prices in Chapter VIII, of last year’s background papers (SM/95/163, Supplement 1).

3/

In cases where the estimates implied perverse short-run dynamics, these terms were dropped. These cases—where terms have been dropped—are self-evident in Table 1. In almost all these cases, the coefficients were also statistically insignificant.

4/

See Meredith (1993) and Chapter VIII in last year’s background papers (SM/95/163, Supplement 1) for comparisons of Japan’s trade elasticities with other studies and countries.

5/

Since world trade has grown faster than incomes, use of the world nonoil imports series has the advantage of incorporating the upward trend in world trade into the activity series itself rather than ascribing it to a higher elasticity.

1/

The composition of Japan’s imports has changed over the years. Particularly notable has been the steady increase in the share of manufactures, which has risen from 23 percent in 1980 to 60 percent in 1995.

2/

Deflators for services flows are not available, making construction of volume and relative price series difficult. In the event, nominal flows were deflated by broad indices of domestic or foreign prices. For transportation flows, use of various alternative price series to deflate the nominal flows produced unsatisfactory results, and the equation was estimated in nominal terms.

1/

Each equation in Chart 3 is forecast separately. This implies that predictions of trade volumes, for example, are based on actual export and import prices, as opposed to the fitted values from the price equations. The shocks or prediction errors, therefore, represent disturbances only to that specific equation.

2/

In interpreting the charts, note that all scales are logarithmic, but the span for each component varies. For example, the scale for the import volume of manufactures equation spans a much larger range of movements than the panel for the raw materials equation.

1/

Since relative export and import price changes reflect changes in either the exchange rate, domestic prices, or foreign prices, there are a variety of ways in which the model could be used to examine the effects of relative price changes. Examining alternative paths of the nominal exchange rate allows changes in Japanese export and import prices to be determined endogenously.

2/

It is worth emphasizing the partial equilibrium nature of the model and, therefore, the caution with which these counterfactual simulations should be interpreted. If the nominal value of the yen had indeed remained unchanged during 1981-95, the economic environment, that is growth and inflation, both at home and abroad would have been different.

1/

In practice, since the real sector forecast also depends on the external forecast, iterations are carried out to make the two consistent.

1/

This also implies, of course, that relative import prices rise over the forecast horizon, as compared to the baseline forecast in which they are flat.

Japan: Selected Issues
Author: International Monetary Fund