A Test of the Efficacy of Exchange Rate Adjustments in Indonesia

International Monetary Fund. Research Dept.
Published Date:
January 1984
  • ShareShare
Show Summary Details

This paper will analyze the impact of relative prices on the non-oil trade account. The analysis will primarily focus on the effectiveness of the exchange rate adjustment of November 1978 in promoting non-oil exports and restraining imports. The importance of fiscal and monetary policies in supporting an exchange rate action is indicated by Indonesia’s experience. On November 15, 1978, the rupiah/dollar exchange rate, which had been fixed since 1971, was depreciated by 33 percent. After this devaluation, a stable link of the rupiah to the dollar re-emerged. During the first year after the devaluation, tight fiscal and monetary policies were successful in retaining the improvement in relative prices stemming from the devaluation. Domestic inflation was reduced to international levels within six months and remained at that level for the next six months.

However, with the additional foreign resources made available by the second round of oil price increases, domestic expenditure began to outstrip domestic productive capacity and to generate pressures on prices and imports. By mid-1981, the relative price of domestic to foreign goods in domestic currency had returned to the level existing prior to the 1978 devaluation; export volume began to stagnate, and import volume expanded rapidly. The rupiah’s close link to the dollar was broken in 1982, when the currency was depreciated by almost 8 percent.

Nevertheless, the competitive position of the non-oil sector deteriorated because of the sharp appreciation of the dollar against major currencies and the relatively high domestic inflation. Conditions in the international oil market progressively weakened during 1982 and early 1983 and strained the balance of payments. Following the Organization of Petroleum Exporting Countries (OPEC) agreement in early 1983, Indonesia’s oil prices were reduced and a production ceiling imposed. In response to lower oil revenue and in order to improve the competitive position of non-oil exports, the rupiah was depreciated by 28 percent on March 30, 1983; this devaluation returned the real exchange rate to about the level it had reached immediately following the November 1978 devaluation.

The paper is organized as follows: the first section reviews the background to the 1978 devaluation, the developments that immediately followed the devaluation, and the events leading to the exchange rate action of March 1983—in particular, the consequences of the second round of oil price increases for the competitiveness of the non-oil sector; the next section attempts to quantify the impact of the exchange rate adjustment on the non-oil trade account by estimating import-demand and export-supply equations and then simulating these equations under alternative exchange rate policies; and the final section summarizes the findings.

I. Review of Developments, 1971–82

background to 1978 depreciation

Indonesian exports are composed almost totally of primary commodities, with crude petroleum being the most important export; net oil and liquefied natural gas (LNG) exports accounted for about 60 percent of total merchandise exports during 1973–78. Exports of other primary commodities (i.e., timber, rubber, copper, tin, palm oil, and coffee) averaged more than 90 percent of non-oil exports. The immediate consequence of this concentration in commodity exports has been a close association between export prices and economic developments in industrial countries; changes in primary commodity prices can be explained to a substantial extent by fluctuations in economic activity of industrial countries.1 Moreover, the volume of commodity exports is also responsive to economic conditions in industrial countries. By combining these two observations, one reaches the conclusion that export earnings—and, therefore, incomes of Indonesians employed in producing primary commodities—would be subject to cyclical fluctuations.

In addition to the cyclical nature of primary commodity exports, markets for these exports have generally grown more slowly than markets for manufactured goods. Industrial countries’ imports of primary commodities grew about 2 percent, on average, in volume terms during 1973–78, while their imports of manufactured goods grew by more than 6 percent per annum. For Indonesia, export volume of important commodities such as timber, rubber, and tin virtually stagnated from 1973 to 1977. Thus, shifting the export structure toward manufactures and away from primary commodities appeared to be a strategy that would increase the growth of non-oil exports and stabilize both earnings and income.

With the fixed exchange rate against the dollar, domestic inflation had a major impact on the competitiveness of non-oil tradables (i.e., export- and import-competing goods). The sixfold increase in petroleum prices during 1971–78 made the containment of domestic inflation a difficult task. Primarily because of rising oil earnings, real budget revenues increased more than twofold during this period and reached nearly 18 percent of gross domestic product (GDP). These greater revenues provided the financial resources needed to intensify the development effort and to expand governmental services. Accordingly, real government expenditures doubled from 1971 to 1978 and increased its share in GDP to 18 percent. While the bulk of the revenues was derived from external sources, expenditures were mainly on domestic goods and services. As a result, budgetary operations had an expansionary impact on domestic credit. At the same time, private-sector expenditures were also stimulated by rapid expansion in domestic credit. Credit to the private sector and state enterprises rose at an annual average rate of 43 percent from 1972 to 1978; liquidity grew at an average rate of 33 percent per annum during this period and created strong inflationary pressures.

Greater expenditures by both private and public sectors were not, however, matched by expansion of domestic productive capacity. Thus, heightened demand pressures strained domestic resources and resulted in higher prices and more imports; in particular, between 1972 and 1975, the price level more than doubled and imports by the private sector nearly tripled. Over 1971–78, domestic inflation averaged 19 percent, compared with an annual average increase in export unit values of industrial countries of 12 percent; real imports rose by nearly 13 percent per annum. The sharp rise in Indonesian consumer prices relative to foreign consumer prices is shown in Chart 1. Indonesian producers of import- competing goods during this period were squeezed, since domestic prices rose 60 percent faster than domestic-currency prices of imports (Chart 2). Notwithstanding an improvement in the non- oil terms of trade during this period, the profitability of the export sector declined by 40 percent. This deterioration in the competitive position of the traded-goods sector was considered to be a major factor underlying the sluggish development of the sector—non-oil exports’ share of GDP declined from 8 percent to 7 percent during 1971–78. Moreover, aside from restraining growth in products already exported, the high relative cost-price structure also deterred development of a competitive manufacturing sector. Consequently, manufacturing production was concentrated in heavily protected industries.

Chart 1.Indonesia: Indices of Nominal and Real Effective Exchange Rates,1 First Quarter 1971-Fourth Quarter 1981

(1971–73 = 100)

Source: International Monetary Fund, International Financial Statistics, various issues.

1 An increase in the index represents an appreciation.

2Nominal effective exchange rate index adjusted for movements in relative consumer price indices.

3Trade weighted.

Chart 2.Indonesia: Ratios of Consumer Price Index (CPI)1to Indices of Export and Import Prices2 First Quarter 1971-First Quarter 1981

(1971–73 = 100)

Source: International Monetary Fund, International Financial Statistics, various issues.

1 An increase in the index represents an appreciation.

2 Export and import prices are measured in domestic-currency terms.

An important consideration in the decision to stimulate the non-oil sector was the outlook for the oil sector. Over the four-year period 1971–74, net oil sector exports increased twelvefold to $2.6 billion. However, in the subsequent four-year period, net oil sector exports rose by only 68 percent to $4.4 billion. Unlike the earlier period, when price increases were responsible for most of the increased value of exports, this later expansion stemmed primarily from a volume increase of 42 percent; prices rose by only 18 percent. By 1978, the outlook for further large export volume increases for petroleum was dim because rapidly rising domestic consumption was expected to cut into a stagnating or slightly declining production. In the absence of large price increases for petroleum, the oil sector did not appear to offer the same potential for financing development and stimulating growth that it had in the past. Thus, more rapid development of the non-oil sector, especially tradable goods, was needed to supplement the growth stimulus previously provided by the oil sector and to support the balance of payments.

immediate post-devaluation developments

The adjustment of the rupiah/dollar exchange rate in November 1978 raised the rupiah equivalent of a unit of foreign exchange by 51 percent and, thereby, made production of export- or importcompeting goods substantially more profitable. However, by raising the domestic-currency price of imports, the devaluation also caused the consumer price index to rise. As seen in Chart 3, this resulted in a burst of inflation following the devaluation (based on a three-month moving average of annualized monthly inflation rates). With the relatively constant rupiah/dollar exchange prevailing after the devaluation, if prices for nontradable goods were to increase pari passu with prices for tradable goods, then a more favorable relative price relationship for tradable goods would not be established. To deter nontradable goods prices from increasing and thus support the expenditure-switching objective of exchange rate policy, tighter demand management policies were pursued. Although the overall budgetary deficit as a percentage of GDP in fiscal year 1978/79 remained virtually unchanged from the previous year, the domestic resource imbalance emanating from the budget was greatly reduced (Table 1). The domestic budgetary deficit—the difference between domestic expenditures and domestic revenues—was lowered from 6.9 percent of GDP in 1977/78 to 5.2 percent in 1978/79 and 1979/80. Monetary policy during 1979/80 contributed to reducing inflationary pressures by reducing the expansion in domestic credit to the private sector and state enterprises to only 14 percent in 1979/80; with the 23 percent inflation, domestic credit declined in real terms by 9 percent.

Chart 3.Indonesia: Three-Month Moving Average of Annualized Monthly Inflation Rates, January 1978–December 1982

(In percent)

Source: International Monetary Fund, International Financial Statistics, various issues.

Table 1.Indonesia: Government Budget, 1977/78–1982/83
Trillions of rupiahs
Foreign sources1(1.9)(2.3)(4.3)(7.0)(8.6)(8.2)
Domestic sources2(1.6)(2.0)(2.4)(3.2)(3.6)(4.2)
Foreign expenditures3(0.9)(1.4)(2.3)(2.9)(3.4)(3.7)
Domestic expenditures(2.9)(3.3)(4.3)(6.7)(9.9)(11.1)
Overall balance–0.4––1.1–2.4
Foreign balance(1.0)(0.9)(2.0)(4.1)(5.2)(4.5)
Domestic balance(1.4)(–1.3)(–1.9)(–3.5) ((–6.3)(–6.9)
Percent of GDP
Overall balance–1.9––2.1–3.9
Foreign balance(5.0)(3.4)(5.6)(8.6)(9.4)(7.3)
Domestic balance(–6.9)(–5.2)(–5.3)(–7.3)(–11.5)(–11.2)
Source: Data supplied by the Indonesian authorities.

As a result of tight demand management, inflation quickly abated after March 1979, and by October 1979 it had dropped to less than 10 percent. During the second semester of 1979/80 (October-March), inflation averaged 11 percent, which was 3 percentage points below foreign inflation as measured by export unit values of industrial countries. Thus, demand-management policies succeeded in reducing domestic inflation to less than the international level within a short time (six months) after the devaluation and maintained inflation at that level for six months.

Earnings from non-oil exports rose by more than 75 percent during 1977–79; export volume increased by 35 percent. The rise in nontraditional exports—nonprimary commodity exports such as handicrafts, rattan, and electrical appliances—was even greater; earnings from these exports more than tripled. In 1979, export volume of these items expanded by 86 percent compared with an average increase of 20 percent in each of the preceding years (Table 2). Consequently, their share in non-oil export earnings was increased from 7 percent to 10 percent, even though traditional export earnings were also expanding rapidly. Thus, progress was made in shifting the composition of exports toward manufactures and away from primary commodities.2 Imports, which had been growing at a 22 percent annual rate during 1976–78, slowed to a 12 percent growth rate in 1979, while the increase in import volume decelerated from an annual average of 10 percent in 1976–78 to 2 percent in 1979.

Table 2.Indonesia: Selected Non-Oil Export Volume, 1976–81(1978 = 100)
Electrical appliances66.065.8100.0147.2107.5116.6
Source: Data provided by Bank Indonesia.
impact of second round of oil price increases

With the more than twofold increase in petroleum prices during 1979–80, Indonesia’s external terms of trade improved sharply. Consequently, real gross national income grew twice as fast as real GDP during 1979–81. The greater imbalance between purchasing power and domestically produced goods fueled inflation and an import boom. Underlying this expansion in expenditures was the stance of fiscal and monetary policy during 1980/81 and 1981/82. The budgetary domestic deficit more than doubled, rising from 5 percent of GDP in 1979/80 to an average of 11 percent during 1981/82 and 1982/83. The higher domestic budgetary deficits placed greater pressure on domestic resources.3 In addition, private sector demand was stimulated by a higher growth rate for credit to the private sector and state enterprises that nearly doubled to an annual average of 27 percent for 1980/81–1982/83. Liquidity expansion also increased, averaging 40 percent per annum during 1980/81 and 1981/82, before slowing to 10 percent in 1982/83.

With the increase in budgetary and aggregate demand pressures, inflation accelerated to an average of about 18 percent in 1980. Foreign inflation slowed from 12 percent during 1980 to 2 percent during 1981 and contributed to the reduction in domestic inflation to about 8 percent during 1981. The higher average inflation in Indonesia than in industrial countries during 1980–82, combined with a decline in foreign export prices, produced a deterioration in the cost-price structure for the traded goods sector. By late 1980, higher domestic inflation raised the relative price of Indonesian to foreign goods to 20 percent above the level prevailing after the devaluation; by late 1981, the relative price structure deteriorated further, and the ratio of domestic to foreign prices was 30 percent higher. Thus, by mid-1981, the relative price of domestic goods to imported goods or exports had returned to the relationship existing prior to the 1978 devaluation. Associated with this erosion in external competitiveness was first stagnation and then a decline in exports. In particular, export volume of nontraditional exports was unchanged in 1980 and fell by 18 percent in 1981.

II. Quantitative Impact of Exchange Rate Adjustment on the Non-Oil Trade Account

The competitive position of tradable goods was improved for a relatively short time by the exchange rate adjustment of November 1978. Moreover, other developments make it difficult to distinguish the impact of the change in relative prices on non-oil trade flows from the impacts of other factors. To disentangle the impact of these various factors on the non-oil trade account, the import demand and export supply functions were econometrically estimated and employed to simulate imports and exports under alternative exchange rate scenarios.4 The technical issues and estimation procedures are described in the next subsection. However, the reader may wish to proceed immediately to the results presented in the following subsection.

specification and estimation of import and export equations5

Demand for imports

In Indonesia, as with most developing countries, the foreign price of imports can be treated as exogenous because the country is small relative to the global market. Indonesian imports accounted for only about 0.6 percent of global imports during 1971–81. It can, therefore, be assumed that imports are demand-determined. Demand for real imports Md is a function of the domestic income RI, the relative price of foreign to domestic goods RP, and the excess supply of liquidity Ls. This function was specified in a log-linear form to permit direct estimation of elasticities

As is standard for demand curves, the income variable has a positive effect and the price term a negative effect.6 An excess supply of money increases domestic prices relative to foreign prices for a given exchange rate and thereby shifts demand to foreign goods. The relative-price term incorporates this indirect impact of excess liquidity on imports. However, excess liquidity also directly increases imports by raising the demand for all goods—foreign and domestic. This mechanism is emphasized by the monetary approach to the balance of payments. Until recently, import-demand functions have not attempted to measure this direct impact. The positive impact of excess liquidity on imports is measured by actual liquidity, less the demand for liquidity, which is a function of real income.

To account for importers’ behavior when they are off their long-run demand curve, a partial-adjustment mechanism was introduced relating the change in imports at time t to the difference between import demand in that period and actual imports in the previous period

where Δ log Mt, = log Mt – log Mt – 1and k denotes the coefficient of adjustment (0 ≤ k ≤ 1). A theoretical rationale for equation (2) is provided by the existence of adjustment costs, delivery delays, recognition lags, and contracts extending beyond the frequency of the data (i.e., three months). All these factors would reduce, in the short run, the speed at which importers adjust to changes in relative price or real income. This framework permits estimation of the mean time lag in the adjustment of actual imports to import demand.

Substituting equation (1) into equation (2) and solving for imports in period t results in

where the coefficients of real income, relative prices, and liquidity represent short-run, or impact, elasticities. The coefficient of real income is indeterminate because higher real income increases money demand, and thereby reduces pressure on import demand.

Data on non-oil imports is available for three categories: imports under government programs, which are primarily food items; imports of capital goods, which are largely used in the implementation of government projects; and imports by the private sector—nonprogram imports. Because government policy determines imports in the first two categories,7 the import- demand equation was estimated only for nonprogram imports. Real imports were obtained by deflating nonprogram imports by a foreign price deflator calculated as a weighted average of trading partners’ export prices. equation (3) was estimated using ordinary least squares; the period of estimation was the first quarter of 1971 through the fourth quarter of 1981, and the data were seasonally adjusted. In addition, the more traditional specification of an import-demand equation—without the direct effect of excess liquidity (a4 = 0)—was also estimated. The results are shown in Table 3, with the t-values appearing in parentheses below the estimated coefficients.

The coefficient for relative prices has the expected sign and is significantly different from zero at the 5 percent level for both equations. A dummy variable, which was introduced to measure the quantum effect of the 1978 devaluation on relative prices, also has the anticipated sign and is statistically different from zero at the 10 percent level of confidence. These coefficients are virtually identical for the two equations, indicating that these values are fairly robust to this change in specification. As expected, the coefficient on real income is positive and smaller in the equation with real money balances than in the equation without real money balances; both coefficients are significant at the 1 percent level.

The coefficient on real money balances has the anticipated sign and is statistically different from zero at the 10 percent confidence level. The coefficient on lagged imports is significant at the 1 percent level. The fit of these two equations, as evidenced by the R2 and Charts 4 and 5, is very good. The hypothesis of serially correlated error terms, which also is an indication of a misspecified equation, can be rejected by the near-zero value of the H-statistic.

Chart 4.Indonesia: Real Nonprogram Imports—Actual and Predicted, First Quarter 1971-Fourth Quarter 19811

(Logarithmic value in millions of 1975 dollars)

1 Based on an equation with real money balances.

Chart 5.Indonesia: Real Nonprogram Imports—Actual and Predicted, First Quarter 1971-Fourth Quarter 19811

(Logarithmic value in millions of 1975 dollars)

1 Based on an equation without real money balances.

The estimation results do not permit one to determine which of the two specifications is better. However, econometric theory indicates that omitting a relevant variable, such as real balances, would bias the coefficients of the remaining variables. In this case the bias on the income coefficient would be upward, but it is not clear in what direction the coefficient on relative prices would be biased. In this connection, it is interesting to recall that the income elasticity in the equation without real money balances is larger than the one in the equation with real money balances. Moreover, if the proper specification is the equation with real money balances, then the income coefficient in the equation without real balances would be larger by the income elasticity of real money. The income elasticity for real money derived from these two equations is 1.23, which is consistent with direct estimates of the income elasticity for real balances.8 This is indirect support for the specification with real balances.9

Supply of exports

Like importers, exporters in Indonesia are also assumed to be price takers on the world market. For most commodities, this assumption presents no problem, since Indonesia has a negligible share of the world market. Even in those commodities where Indonesian exports are not trivial, they were a small share of the relevant market during this period.10 Export supply is specified as a function of profitability and domestic capacity. Export profitability, and thus export supply, increase as the relative price of exports (in domestic currency terms) to domestic price increases; the relative-price term (RPD) was measured as the ratio of the weighted average of Indonesia’s major export commodities to its domestic consumer price index (CPI). The ability of exporters to supply the foreign market is constrained by the country’s productive capacity. For Indonesia, real gross domestic product (RGDP) serves as a proxy for capacity. As with the previous equation, the export supply equation is specified as a log-linear function

This equilibrium supply equation can be transformed to accommodate sluggish supply response by employing a partial- adjustment mechanism similar to equation (2)

Substituting equation (4) into equation (5) and solving for exports in period t yields

Timber exports, which averaged 30 percent of non-oil exports during the 1970s, were restricted in 1980 and 1981 by the Government to promote conservation and the domestic processing of logs. Existing capacity for processing timber limited exports of plywood and other sawn-wood products; consequently, earnings from timber exports declined by 43 percent, from its peak of $1,917 million in 1979 to $1,095 million in 1981, as export volume fell by 45 percent. To segregate the impact of this policy from the impact of relative prices, non-oil exports were estimated excluding timber. Seasonally adjusted data for the period extending from the second quarter of 1971 to the fourth quarter of 1981 were employed.

The estimation results for equation (6) are provided in Table 3. The coefficient on relative prices and the coefficient on lagged exports have the expected positive sign and are significantly different from zero at the 1 percent level. The coefficient of the capacity variable RGDP is positive, as was anticipated, and significant at the 5 percent level. The coefficient on the dummy variable, which was introduced to measure the quantum effect of the devaluation on relative prices, is not significantly different from zero. Although this equation does not fit as well as the other equations in Table 3, it nevertheless performs reasonably well, as indicated by the high R2 in Chart 6. The H-statistic is substantially less than the critical value; thus, serial correlation is not present. Greater than usual caution should be exercised in interpreting these estimation results because the coefficient of lagged exports is so near its upper limit, which implies relatively sluggish adjustment. In addition, small changes in this coefficient have a large impact on the derived long-run elasticities.

Table 3.Indonesia: Import and Export Equations1

Real Gross






Real nonprogram imports–3.74






Real nonprogram imports–5.18





Real non-oil exports, excluding timber–5.51






Chart 6.Indonesia: Supply of Real Non-Oil Exports, Excluding Timber—Actual and Predicted, Second Quarter 1971– Third Quarter 1981

(Logarithmic value of export volume)
estimation results

The econometric results of the previous subsection indicate that the estimated equations for import demand and export supply are well specified according to standard measures. The estimated elasticities and the adjustment lags obtained from these regressions are presented in Table 4. The mean adjustment lag for import demand is shorter than for export supply. It is not surprising that production lags constrain short-run increases in export supply more than delivery lags constrain imports. These adjustment lags also imply that short-run elasticities for exports are substantially smaller than the long-run elasticities.

Table 4.Indonesia: Short-Run and Long-run Elasticities and Mean Adjustment Lags
Price ElasticitiesIncome ElasticitiesExcess LiquidityMean Adjustment Lag
EquationShort-runLong-runShort-runLong-runShort-runLong-run(In months)
Import demand
Without real money balances–0.201–0.310.831.302
With real money balances–0.201–0.300.610.930.1960.2992
Export supply
Real non-oil exports excluding timber0.616.000.5325.19226

The short- and long-run elasticities for import demand are within the range reported for other developing countries. The relative price elasticities are small, both in the short run and in the long run, and are on the inelastic portion of the demand curve. These elasticities are, nevertheless, significantly different from zero, indicating that relative prices do have an influence on import demand, even in the short run. The short lag in the adjustment of import demand to changes in income and relative price movements is to be expected, given the relatively free exchange system in Indonesia.

Turning to the export-supply equation, the results indicate that the supply-price elasticity is moderate in the short run and rises to a long-run value of about 6. (Price elasticities of export supply have generally been estimated for industrial countries and have ranged between 1 and 7, with values as high as 15 reported.) A high supply-price elasticity could reasonably be expected in countries with relatively small export sectors, since it is easier to increase exports when there is large domestic productive capacity to draw upon; during the 1970s, non-oil exports excluding timber averaged about 6 percent of GDP. The small share of the non-oil export sector in total GDP would also explain the high coefficient for the capacity variable (i.e., real GDP). The mean lag time for the adjustment of export supply is nine quarters, which is not surprising given that production of the major export commodities (tin, rubber, coffee) is characterized by a long lead time between investment and incremental output.11

It should be recalled that the mean adjustment lag and the estimated long-run elasticities (i.e., the estimated coefficient divided by the adjustment coefficient) are only known with limited precision—defined by the standard error of the coefficient. This caveat becomes especially important as the adjustment coefficient nears its upper bound, because the same imprecision is introduced into the derived long-run elasticities; for example, if the adjustment coefficient were smaller by one standard deviation, then the long-run elasticities would be only half their reported values.12

simulation exercises

Using the specification of import-demand and export-supply functions, the impact of the 1978 devaluation on non-oil trade account was analyzed by simulating hypothetical exchange rate scenarios. Real imports were simulated under two alternative scenarios. Scenario I assumes that the structure of relative prices prevailing prior to the depreciation (i.e., prior to the third quarter of 1978) continued unchanged for the remainder of the simulation period. This scenario is termed no real devaluation. Scenario II, on the other hand, assumes that the more favorable relative prices created after the exchange rate adjustment were maintained—specifically those established in the first quarter of 1979. This scenario is referred to as the maintained real devaluation.

The impact of the actual path of relative price on import volume is compared with the paths under the two assumed exchange rate assumptions in Chart 7. As can be seen, the actual relative price developments kept import volume below the level simulated by the no real devaluation scenario until 1981. However, as the relative price structures under Scenario I converged with actual developments, the import volumes would have been less disparate; and by 1981, when the relative prices had become virtually identical, import volumes would have been the same. Under Scenario II, import volume would have grown much more slowly and would have been 20 percent lower than the actual volume in 1981. The full implication of these alternative scenarios on foreign resources are summarized in Chart 8, where the cumulative differences between simulated imports and actual imports are presented. According to the simulation, if no real devaluation had taken place, higher imports would have resulted in international reserves decreasing by about $1 billion. Even more interesting is the simulation result under a maintained real devaluation; imports over the three post-devaluation years would have been $3.5 billion less than actual imports. Thus, the total savings from reduced imports under the two alternative scenarios would have been over $4.5 billion, or 5.3 percent of GDP in 1981.

Chart 7.Indonesia: Simulation of Import Demand, 1978–811

(Logarithmic value in millions of 1975 dollars)

1 Volume of nonprogram imports.

Chart 8.Indonesia: Cumulative Difference in Import Payments Under Exchange Rate Scenarios, Fourth Quarter 1978-Fourth Quarter 19811

(In billions of dollars)

1 Predicted imports under alternative scenarios less actual imports.

As with imports, the impact of the actual path of relative prices on real export supply was compared with the paths under two alternative relative price assumptions (Chart 9). Scenario I has no exchange rate adjustment in November 1978, and consequently relative prices would have been less favorable for exporters. In Scenario II, the improved cost-price structure existing after the depreciation was maintained for the remainder of the period—specifically, the relative prices in the second quarter of 1979 were used.13 Export volumes, under the alternative exchange rate scenarios and the actual exchange rate policy, were multiplied by actual export prices to obtain export earnings. As previously noted, export prices may be affected by changes in export volume, but the impact is likely to be small; consequently, this feedback in prices was ignored in these calculations. The cumulative differences between simulated export earnings and actual earnings are presented in Chart 9. According to these calculations, reduced exports under Scenario I would have resulted in international reserves decreasing by about $0.8 billion in late 1981. However, under Scenario II, the simulation results indicate that international reserves would have been almost $0.9 billion above their actual level. Thus, the difference in export earnings under the two alternative scenarios would have been about $1.7 billion, or 2.0 percent of GDP in 1981. The gains from exports were less than the savings reported from imports, because nonprogram imports were nearly triple non-oil exports (excluding timber).

Chart 9.Indonesia: Simulation of Export Supply, Fourth Quarter 1978-Third Quarter 19811

(Logarithmic value of export volume)

1 Volume of non-oil exports, excluding timber.

The use of import and export equations to simulate alternative scenarios is subject to a number of caveats. First, in simulating the scenarios, the point value of the estimated coefficients was employed. The results are, therefore, subject to a margin of error because the coefficients are only known with limited precision. Second, even though relative prices are substantially different under the alternative scenarios, this is assumed not to have any impact on the performance of the economy other than on import and export volume. Thus, the actual data and the estimated coefficients for real national income, real balances, and real capacity were utilized in the simulations. Moreover, the additional export volume was assumed not to change world prices for Indonesia’s exports.14

Chart 10.Indonesia: Cumulative Difference in Export Earnings Under Alternative Exchange Rate Scenarios, Fourth Quarter 1978-Third Quarter 19811

(In millions of dollars)

1 Predicted exports under alternative scenarios less actual exports.

III. Conclusions

An exchange rate adjustment can successfully redirect resources into export and import-competing activities only if the more favorable cost-price relationship established after the adjustment is not subsequently reversed by domestic price increases. The exchange rate adjustment of November 1978 raised the prices of tradable goods by 50 percent and resulted in a burst of inflation in early 1979. Inflation was quickly reduced to international levels by appropriate demand-management policies. The competitiveness of the traded goods sector was, thus, significantly improved. In 1979, private sector import growth slowed, while earnings from non-oil exports grew more rapidly. These developments reflected adjustment policies as well as a better external environment.

With the sharp increase in oil revenues in 1979–80, the balance of payments constraint, which had prompted the exchange rate adjustment, was removed. The subsequent expansion of government expenditure and domestic liquidity led to a rapid growth in aggregate demand and, thereby, to greater inflation. As a result of the resurgence in inflation, the improvement in the competitiveness of the non-oil trade sector was reversed. Given the short duration of the enhanced competitiveness, the favorable effect of relative prices on the non-oil trade account was quickly dissipated. To isolate the impact of the exchange rate on the non-oil trade account, import-demand and export-supply equations were specified and estimated; these estimates were then employed to simulate the impact on the non-oil trade account under alternative exchange rate scenarios. While the results of both these exercises should be interpreted with due caution, they nevertheless indicate that the non-oil trade account was strongly influenced by the relationships of domestic prices to domestic currency prices of exports and imports.


Data Definitions and Sources

All data are quarterly, seasonally adjusted, for 1971–81 and are taken from four sources:

  • (1) International Monetary Fund, International Financial Statistics (Washington), various issues.

  • (2) Indonesia, Central Bureau of Statistics, Monthly Statistical Bulletin (Jakarta), various issues.

  • (3) Bank Indonesia, Weekly Report (Jakarta), various issues.

  • (4) World Bank, Indonesia: Financial Resources and Human Development in the Eighties (Washington, May 3, 1982).

NM=value of nonprogram imports in dollars, Source (3).
M=volume of nonprogram imports obtained by deflating nominal imports NM by foreign prices FP.
X=volume of exports, Source (2).
PX=unit value of exports, derived from value and volume data, Source (2).
FD=foreign demand, expressed as an index with 1975 = 100. This series is constructed as the export-weighted sum of trading partners’ real gross national product. Source (1).
FP=foreign price, expressed as an index with 1975 = 100. This series was constructed as the import-weighted sum of trading partners’ export unit value or export prices. Source (1).
RI=real gross national income defined as real gross national product plus terms-of-trade effect. This is an annual time series that was transformed into a quarterly basis using an interpolation method. Source (4).
CPI=Indonesia’s consumer price index, 1975 = 100, Source (1).
L=real liquidity defined as money plus quasi-money divided by CPI, Source (1).
RP=relative price foreign price FP divided by consumer price index CPI.

    Aghevli, Bijan B., Mohsin S.Khan, P.R.Narvekar, and Brock K.Short,“Monetary Policy in Selected Asian Countries,”Staff Papers, International Monetary Fund (Washington), Vol. 26 (December1979), pp. 775824.

    Artus, Jacques R., and Susana C.Sosa,“Relative Price Effects on Export Performance: The Case of Nonelectrical Machinery,”Staff Papers, International Monetary Fund (Washington), Vol. 25 (March1978), pp. 2547.

    Deppler, Michael C., and Duncan M.Ripley,“The World Trade Model: Merchandise Trade,”Staff Papers, International Monetary Fund (Washington), (March1978), pp. 147206.

    Dornbusch, Rudiger, and Dwight J.Jaffe,“Purchasing Power Parity and Exchange Rate Problems: Introduction,”Journal of International Economics (Amsterdam), Vol. 8 (May1978), pp. 15762.

    Goldstein, Morris, and Mohsin S.Khan,“Large versus Small Changes and the Demand for Imports,”Staff Papers, International Monetary Fund (Washington), Vol. 23 (March1976), pp. 20025.

    Goldstein, Morris, and Mohsin S.Khan,“The Supply and Demand for Exports: A Simultaneous Approach,”Review of Economics and Statistics (Cambridge, Massachusetts), Vol. 60 (May1978), pp. 27586.

    Goldstein, Morris, and Mohsin S.Khan,Effects of Slowdown in Industrial Countries on Growth in Non-Oil Developing Countries, IMF Occasional Paper Series, No. 12 (Washington: International Monetary Fund, August1982).

    Goldstein, Morris, and Mohsin S.Khan,“Income and Price Effects in Foreign Trade,” Chapter 20 in Handbook of International Economicsed. byRonald W.Jones and Peter B.Kenen (Amsterdam: North-Holland, forthcoming in1984).

    Goreux, Louis M.,Compensatory Financing Facility, IMF Pamphlet Series, No. 34 (Washington: International Monetary Fund, 1980).

    Hooper, Peter, and StevenKohlhagen,“The Effect of Exchange Rate Uncertainty on the Prices and Volume of International Trade,”Journal of International Economics (Amsterdam), Vol. 8 (November1978), pp. 438511.

    Houthakker, Hendrik S., and Stephen P.Magee,“Income and Price Elasticities in World Trade,”Review of Economics and Statistics (Cambridge, Massachusetts), Vol. 51 (May1969), pp. 11125.

    Isard, Peter,“How Far Can We Push the ‘Law of One Price’?”American Economic Review (Nashville), Vol. 67 (December1977), pp. 94248.

    Junz, Helen B., and Rudolf R.Rhomberg,“Prices and Export Performance of Industrial Countries, 1953–63,”Staff Papers, International Monetary Fund (Washington), Vol. 12 (July1965), pp. 22469.

    Junz, Helen B., and Rudolf R.Rhomberg,“Price Competitiveness in Export Trade Among Industrial Countries,”American Economic Review (Nashville), Vol. 63 (May1973), pp. 41218.

    Khan, Mohsin S.,“Import and Export Demand in Developing Countries,”Staff Papers, International Monetary Fund (Washington), Vol. 26 (November1974), pp. 67893.

    Khan, Mohsin S., and Knud Z.Ross,“The Functional Form of the Aggregate Import Demand Equation,”Journal of International Economics (Amsterdam), Vol. 7 (May1977), pp. 14960.

    Kmenta, Jan,Elements of Econometrics (New York: Macmillan, 1971).

    Learner, Edward E., and Robert M.Stern,Quantitative International Economics (Boston: Allyn and Bacon, 1970).

    Liu, Ta-Chung,“The Elasticity of U.S. Import Demand: A Theoretical and Empirical Reappraisal,”Staff Papers, International Monetary Fund (Washington), Vol. 3 (February1954), pp. 41641.

    Magee, Stephen P.,“Prices, Incomes, and Foreign Trade,” in International Trade and Finance: Frontiers for Research, ed. byPeter B.Kenen (Cambridge University Press, 1975), pp. 175252

    Magee, Stephen P., and Ramesh K.S.Rao,“Vehicle and Nonvehicle Currencies in International Trade,”American Economic Review: Papers and Proceedings of the Ninety-Second Meeting of the American Economic Association (Nashville), Vol. 70 (May1980), pp. 36873.

    Miles, Marc A.,“The Effects of Devaluation on the Trade Balance and the Balance of Payments: Some New Results,”Journal of Political Economy (Chicago), Vol. 87 (June1979), pp. 60020.

    Murray, Tracy, and Peter J.Ginman,“An Empirical Examination of the Traditional Aggregate Import Demand Model,”Review of Economics and Statistics (Cambridge, Massachusetts), Vol. 58 (February1976), pp. 7580.

    Mutti, John H.,“The Specification of Demand Equations for Imports and Domestic Substitutes,”Southern Economic Journal (Chapel Hill, North Carolina), Vol. 44 (July1977), pp. 6873.

    Orcutt, Guy H.,“Measurement of Price Elasticities in International Trade,”Review of Economics and Statistics (Cambridge, Massachusetts), Vol. 32 (May1950), pp. 11732.

    Ramsey, J.B.,“Tests for Specification Errors in Classical Least-Squares Regression Analysis,”Journal of the Royal Statistical Society (London), Series B, Vol. 31 (No. 2, 1969), pp. 35071.

    Spitäller, Erich,“Short-Run Effects of Exchange Rate Changes on Terms of Trade and Trade Balance,”Staff Papers, International Monetary Fund (Washington), Vol. 27 (June1980), pp. 32048.

    Stern, Robert M., and others,Price Elasticities in International Trade: A Compilation and Annotated Bibliography (London: Macmillan, 1976).

    Theil, Henri,Principles of Econometrics (New York: John Wiley & Sons, 1971).

Mr. Kincaid, Senior Economist in the Asian Department, holds degrees from the University of California at Los Angeles and Columbia University.

The author has benefited from the comments of Fund colleagues but is naturally responsible for any remaining errors.

See, for instance, Louis Goreux (1980).

The profitability of exporting manufactured items was also enhanced after the devaluation by an export certificate scheme under which eligible exporters received a rebate of import duties.

In 1980/81, the greater domestic budgetary deficit was more than offset by the increase in the foreign budgetary surplus; thus, a larger overall budget surplus resulted. In 1981/82, however, the increase in the foreign budgetary surplus fell substantially short of the expansion in the domestic deficit and the overall fiscal position turned to deficit. The swing of 3.4 percentage points in the overall fiscal position was equivalent to about half of the change in the external current account between 1980/81 and 1981/82. The decline in the foreign surplus in 1982/83 was not offset by a similar contraction in the domestic deficit; consequently, the overall deficit expanded by almost 2 percent of GDP, which, once again, was equivalent to about half of the deterioration in the external current account.

Due to data availability, the import and export equations were estimated only through 1981.

For an extremely useful presentation of the empirical issues related to estimating export and import functions, see Morris Goldstein and Mohsin S. Khan (1984).

A variable representing abrupt changes in relative prices was also included when the equation was estimated. This variable attempts to measure the impact of large changes in relative prices; some economists have argued that the relative-price elasticity would be larger for large price changes than for small price changes.

Capital imports associated with government projects average about 75 percent of capital-goods imports. Moreover, through the Board of Investment, the Government regulates foreign and domestic investment, which account for the remaining 25 percent.

One example is Bijan B. Aghevli, Mohsin S. Khan, P.R. Narvekar, and Brock K. Short (1979). In that paper, real income was defined as real gross national product, whereas in this paper it is defined as real gross national income (GNY), which includes the gains from the improvement in the terms of trade. During 1971–81, gross national income grew, on average, 50 percent faster than real GDP. Consequently, it would be expected that the long-run income elasticity for real money balances reported using real GDP would be 50 percent higher than the one using real GNY. In fact, the reported elasticity in their paper was 1.85, or 50 percent higher than the value given in the text.

In the simulation exercise, only the results using the real-balance equation are presented; however, both equations were simulated, and the results were nearly identical. This is not surprising, given that the relative-price coefficients are the same and that there is a close relationship (described in the text) between income and real balances.

Indonesia’s shares in world production of coffee and tin were only 5 percent and 10 percent, respectively. For natural rubber, Indonesia’s share of the market was more substantial—averaging 25 percent—during 1971–81. However, the share of natural rubber in the market for both natural and synthetic rubbers was only about 32 percent; therefore, Indonesia’s share of the larger market was limited to 8 percent.

The World Bank has estimated that the adjustment of tin, rubber, and coffee supply to higher prices requires 3, 7, and 7 years, respectively, which accords with this result.

A reduction in the adjustment coefficient for the import equation of the same absolute amount would only change the long-run elasticities by 15 percent, because this adjustment coefficient is much further from the upper bound.

The simulation for the maintained real devaluation scenario begins in the third quarter of 1979, because until then the relative price of exports was improving. Consequently, export supply under Scenario II and under actual relative price developments are the same from the fourth quarter of 1978 to the third quarter of 1979. This also explains the absence in Chart 10 of any additional export earnings until the fourth quarter of 1979.

To quantify the impact of Indonesia’s export supply on its export prices, a price elasticity for export demand was calculated as a weighted average of price elasticities for each commodity. This elasticity indicates that a 10 percent increase in export volume would reduce export prices by between ½ and 1 percent. Thus, as a first approximation, maintaining export prices constant appears workable for simulation purposes, although the results obtained would be somewhat overstated.

Other Resources Citing This Publication