Information about Asia and the Pacific Asia y el Pacífico
Chapter

5 Indonesia’s Fiscal Position: Sustainability Issues

Editor(s):
John Hicklin, David Robinson, and Anoop Singh
Published Date:
July 1997
Share
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details
Author(s)
Geoffrey Bascand and Assaf Razin 

During the late 1980s, the Indonesian authorities implemented a strong fiscal adjustment program to strengthen the fiscal position and contain Indonesia’s public debt, which had risen sharply in the early 1980s. Despite these measures—and sustained improvement in fiscal balances in the 1990s—the level of public debt remains relatively high. Meanwhile, the volume of a major government asset, oil reserves, has declined significantly, reflecting lower oil prices and depletion of the resource. In view of the recent widening of the current account deficit and the volatility of global capital flows, this paper examines whether further fiscal consolidation would reduce the vulnerability of the economy to unfavorable external conditions and help sustain savings, investment, and growth.

To address the merits of fiscal consolidation, the paper develops a framework for assessing the sustainability of the current fiscal position in Indonesia. As a starting point, we accept that fiscal solvency—defined in the narrow sense that current policies can be maintained without raising taxes, cutting expenditures, or resorting to monetization or debt repudiation—is a minimum condition for sustainability. We then incorporate into this framework a number of specific features of the Indonesian economy to provide for a more comprehensive analysis of fiscal sustainability. First, we take into account possible unfunded expenditure liabilities and the prospective further decline in oil and gas revenues over the long term. Second, we consider the impact of lower inflation on central government finances. Third, we examine the sustainability of the external current account, which, in principle, could be closely related to the fiscal position in Indonesia because the government is not permitted to borrow domestically, but may incur external debt obligations.1

In assessing sustainability of the fiscal balance, we have considered, in particular, its resilience to abrupt and discrete change when the economy is hit by a negative shock. The rationale for this approach is that, to minimize risks, the fiscal position should provide a buffer to deal with unexpected shocks; the size of this buffer will be related to the probability and severity of potential shocks.2 The need for a buffer—positive net worth—may be particularly relevant in Indonesia because of the balanced budget rule, which limits recourse to debt finance.

For Indonesia, fiscal sustainability means that3

  • sharp fiscal adjustment is not likely to be required to meet public debt obligations, given reasonable expectations for expenditure and revenue developments, including the exhaustibility of oil and gas reserves, and allowing for potential risks to the fiscal position, including interest rate and growth shocks;

  • the fiscal position is not reliant upon inflation, such that a lower rate of inflation would not lead to a marked deterioration in the government’s finances;

  • the fiscal position is consistent with a sustainable external position, which in turn is defined as a situation where the risks of large and sudden reversals of capital flows are relatively low.

Fiscal Policy: A Long-Term Perspective

Indonesia, like most other member countries of the Association of South East Asian Nations (ASEAN), has grown rapidly for over twenty years. In the 1990s, average GDP growth has exceeded 8 percent, although inflation has remained higher than in Malaysia, Singapore, and Thailand (Table 1).4 Investment and saving rates have risen steadily, especially in the private sector, and are similar to those in other ASEAN countries. Long term, the current account deficit has averaged about 2 percent of GDP, but has recently widened sharply, largely reflecting domestic demand pressures. Export growth and the share of exports in GDP are relatively low by ASEAN standards and need to be taken into account when assessing debt-servicing potential.5 Capital movements have been free since the late 1960s, and the capital account has recorded substantial inflows, especially over the recent period. These large capital flows have reflected investor confidence in Indonesia’s strong performance and growth prospects, but they have also been associated with a widening current account deficit, potentially leaving the country vulnerable to shifts in foreign demand and market sentiment.

Table 1.Selected Economic Indicators: ASEAN, 1991-95
Economic IndicatorIndonesiaMalaysiaPhilippinesSingaporeThailand
GDP growth7.88.72.28.88.4
Inflation8.94.310.42.64.8
Exports-GDP26.188.032.1149.829.3
Savings31.530.419.047.634.8
Current account-2.4-6.3-3.414.4-6.4
Fiscal balance-0.20.2-1.213.22.8
Public debt137.422.243.00.011.4
Debt-exports2176.78.8150.645.359.5
Non-oil tax revenue—GDP10.318.715.218.716.6
Source: Data provided by country authorities.Note: Five-year average, in percent.

External public debt in percent of GDP.

Gross external debt less foreign exchange reserves in percent of exports of goods and nonfactor services.

Source: Data provided by country authorities.Note: Five-year average, in percent.

External public debt in percent of GDP.

Gross external debt less foreign exchange reserves in percent of exports of goods and nonfactor services.

Indonesia has achieved this impressive economic performance through generally sound macroeconomic policies and increasingly open trade, financial, and investment arrangements. In particular, the balanced budget rule has contributed to sound fiscal policy. Adopted in the Guidelines for State Policy in 1968, the rule requires that in each fiscal year total budgeted expenditures be equal to budgeted revenues, defined to include foreign financing. The policy also requires that domestic revenues be large enough to cover routine expenditures (which include amortization payments in the government budget) and a portion of development expenditure. The rule largely prevents domestic private bank and nonbank financing. To the extent that development expenditure exceeds public saving, the gap can normally be filled only through foreign borrowing.

In practice, however, there is some flexibility in the balanced budget rule, beyond foreign borrowing. In particular, the rule is confined to of ficial budgetary accounts, and government transactions may be conducted through extrabudgetary accounts, mainly for investment or net lending transactions. Net spending from these accounts has been limited in the last several years, but, on occasion, movements in them have been large relative to the official fiscal balance. Overall fiscal balance figures reported in this paper include the net balance of these extra-budgetary transactions. However, the financial position of the wider public sector may show significantly different figures for revenue, expenditure, and the fiscal balance, as could accrual rather than cash accounts, but we do not have enough data to adjust for these outcomes. Finally, seigniorage revenue may have been used for financing, and budgetary operations financed through government borrowing abroad may still have monetary consequences unless monetary policy actively offsets them. Nevertheless, the balanced budget rule has played an important role in shaping fiscal settings and in bringing about significant fiscal policy adjustment in the presence of declining oil revenues.

Notwithstanding generally prudent fiscal policy, the fiscal deficit has averaged about 1½ percent of GDP since the early 1970s (Figure 1).6 Oil revenues increased rapidly from 4 percent of GDP in 1972/73 to 12 percent of GDP in 1981/82, peaking at 70 percent of total revenues (Figure 2).7 Through this period, oil and gas revenues financed higher public expenditure, particularly investment and net lending transactions with state enterprises, and the fiscal deficit was small. However, when oil prices fell in the early 1980s, oil revenues declined sharply. Expenditure was initially slow to adjust to the lower revenues, and budget deficits averaged 3-4 percent of GDP from 1981/82 through 1986/87. Confronted by rising public debt levels—public external debt was over 50 percent of GDP in 1987/88—the government developed a concerted strategy to raise non-oil revenues and restrain spending, especially public investment. Expenditure declined from a peak of 23 percent of GDP in 1985/86 to about 17 percent of GDP in 1990/91, with two-thirds of the cut falling on investment expenditures. The overall fiscal balance improved in 1990/91 (buoyed by cyclical factors), registering a surplus of 2 percent of GDP, and has been close to balance on average since then. The ratio of public debt to GDP has declined gradually since the late 1980s and currently amounts to just over 30 percent of GDP.

Figure 1.Fiscal Balance and Public Debt

(In percent of GDP)

Figure 2.Revenue and Expenditure Trends

(In percent of GDP)

The above trends were amplified in the non-oil fiscal balance, which deteriorated sharply in the late 1970s and early 1980s, when expenditure increases were funded mostly through oil revenues. Recognizing the need to provide for future declines in oil revenues, Indonesia began strengthening the non-oil tax base in the mid-1980s, with the restructuring of income taxation in 1984, the introduction of the value-added tax in 1985, and reform of property taxation in 1986.8 Non-oil (and gas) tax revenues increased from only 5 percent of GDP in 1984/85 to over 10 percent of GDP by 1990/91, although they have since shown little further increase. Consequently, the non-oil balance has improved over the past decade, but remains in small deficit. The non-oil current primary balance—which excludes interest payments and investment expenditures—has been in surplus since 1988/89 and amounted to 3 percent of GDP in 1995/96. Since the current primary balance is a key determinant of fiscal solvency, the government’s debt-servicing capacity would, on this measure, appear to be sound.

Fiscal Solvency and Treatment of Nonrenewable Oil and Gas Reserves

In this section, we first outline a conceptual framework to assess fiscal solvency9 and then apply it to Indonesia by comparing the government’s existing liabilities with the debt level that can be sustained given current trends in government expenditure and revenue flows.10 Although this conceptual framework is based on the standard approach to determine solvency, it differs from existing studies in an important way in that we allow for the fact that, in Indonesia, oil and gas revenues cannot be assumed to continue forever.

The analysis in this section is confined to the central government plus the central bank (the results would be only marginally altered by consolidation of central and provincial government accounts). We include the central bank’s profits (seigniorage) to determine fiscal solvency because they enhance the government’s fiscal surplus and its ability to meet future debt obligations.11

Conceptual Framework

The consolidated balance sheets of the central bank and the non-oil central government can be used to derive an accounting identity for the accumulation of public debt. (Appendix II provides a more detailed treatment.) The change in debt between one period and the next is equal to the sum of interest on last period’s outstanding debt plus the non-oil current primary deficit less seigniorage revenue, which is effectively given by the change in nominal reserve money balances. The level of debt can therefore be written as

where Bt, is public external debt in dollars; St is the rupiah-U.S. dollar exchange rate; it is the nominal domestic rate of interest; Pt is the GDP deflator; Dt is the overall non-oil current primary deficit, defined as total current expenditure excluding interest payments less total non-oil tax revenue (in constant prices); and Mt is the nominal base or reserve money. Stock variables are valued at the beginning of the period and flows are assumed to occur during the period.

The sustainable level of public debt is the level of debt that can be sustained in perpetuity for a given primary balance and seigniorage revenue. When an economy is in a steady state, in which all the fiscal magnitudes are constant as a fraction of GDP (and are expected to remain constant) and where the nominal domestic interest rate equals the nominal world rate of interest plus expected exchange rate depreciation, the sustainable level of debt (omitting time subscripts) is equal to

where b, d, and h are public debt, the non-oil current primary deficit, and seigniorage revenue, respectively, all expressed in terms of units of GDP; g is the real growth rate of the economy, i* is the world rate of interest in dollars, and ε is the real rate of exchange rate depreciation.12

Solvency requires that the current level of net government liabilities as a fraction of GDP be less than or equal to the maximum debt level (b) that can be sustained by ongoing revenue and expenditure flows, which we calculate using estimates for the right-hand-side variables. We can express the difference between these values as government net worth,13 which is positive when the present value of future fiscal flows exceeds current net liabilities: NWt = b - lt.

Observe from equation (2) that the larger the non-oil current primary fiscal surpluses (including seigniorage), the higher the growth rate, and that the lower the real interest rate, the higher the level of debt that can be sustained. By the same token, the level of assets required for solvency in the presence of fiscal deficits increases as the interest rate declines or the growth rate increases because the financial return necessary to balance the budget is diminished in the first instance, and the stock of assets is smaller in relation to future income in the latter.

The government’s net financial position is obviously enhanced by the future stream of oil revenues, but in contrast to tax revenues that can be assumed constant in relation to GDP (holding policy constant), it is important to allow for the future decline in oil revenues in line with the depletion of oil reserves. The government’s liabilities (in terms of GDP units PtYt) should be adjusted for the net value of the stock of oil reserves, et, so that the net liabilities position measured at any time reflects the market value of the remaining reserves. We therefore compare the actual net liabilities position, lt = bt - et, with the sustainable debt level, b, where et is defined as follows:

where Et represents known oil reserves, in millions of barrels; and PtE is the price of oil, in dollars.

We have defined the deficit as current noninterest expenditure less non-oil tax revenues, on the assumption that investment income matches the cost of capital for public investment, leaving solvency calculations unaltered. Correspondingly, investments in, and profit receipts from, public enterprises have been omitted.14

Application to Indonesia

The government’s actual net liabilities position (taking account of declining oil reserves) is contrasted in Table 2 with the sustainable debt level—based on the framework outlined above—for the period 1972/73 to 1995/96, on the assumption that Indonesia now has 15 years of oil and gas reserves remaining.15 Actual net liabilities comprise public debt, the market value of the oil reserves, and the government’s foreign currency reserves. The table (column 1) shows that the government’s net assets position (negative net liabilities) declined markedly during the 1980s, partly as a result of higher public debt, but mainly reflecting the declining value of oil reserves.16 Extended to include unfunded post-retirement pension and health care liabilities of about 12 percent of GDP, the government is estimated to have held net liabilities of about 7 percent of GDP in 1995/96 (see column 2).17

Table 2.Fiscal Solvency(In percent of GDP)
Actual Net Public Liabilities1Sustainable Public Liabilities2Sustainable Public Liabilities Under Alternative Assumptions2
Excluding pensionsIncluding pensionsWith zero inflationWith 2 percent lower growth3
1972/73-197.543.6-20.6-13.8
1973/74-242.635.7-25.5-17.1
1974/75-365.925.7-10.9-27.4
1975/76-292.511.1-39.6-26.6
1976/77-257.71.5-39.4-26.5
1977/78-245.1-20.3-57.1-38.3
1978/79-331.0-19.5-46.4-31.1
1979/80-420.5-22.4-69.1-46.4
1980/81-345.4-80.7-121.8-81.8
1981/82-265.9-121.0-140.0-94.0
1982/83-235.3-73.3-88.1-59.1
1983/84-219.5-47.4-75.4-50.6
1984/85-194.4-65.6-83.8-56.3
1985/86-151.8-44.8-70.8-47.5
1986/87-94.5-11.9-35.4-23.8
1987/88-79.7-8.0-28.4-19.1
1988/89-49.112.116.811.3
1989/90-55.144.222.615.2
1990/91-52.955.248.932.8
1991/92-33.794.960.140.4
1992/93-20.578.769.646.8
1993/94-9.4101.277.452.0
1994/95-2.79.8101.475.350.6
1995/96-4.87.3105.763.742.8
Projections4
1996/97-3.48.597.270.247.1
1997/98-2.09.498.270.945.9
1998/99-0.910.199.571.944.9
1999/20000.611.3100.772.745.5
2000/20011.611.9102.173.746.1
Sources: Authors’ calculations from data provided by Indonesian authorities; and World Bank for pension liabilities.

Net public liabilities equal public debt less foreign reserves less oil reserves plus, in column 2, unfunded pension liabilities. Negative values imply a net asset position.

Using non-oil current primary balance. Negative values imply that a net asset position is required for sustainability.

With zero seigniorage. Growth minus 2 percent is approximately equal to a real interest rate that is 2 percentage points higher.

See Appendix III for assumptions.

Sources: Authors’ calculations from data provided by Indonesian authorities; and World Bank for pension liabilities.

Net public liabilities equal public debt less foreign reserves less oil reserves plus, in column 2, unfunded pension liabilities. Negative values imply a net asset position.

Using non-oil current primary balance. Negative values imply that a net asset position is required for sustainability.

With zero seigniorage. Growth minus 2 percent is approximately equal to a real interest rate that is 2 percentage points higher.

See Appendix III for assumptions.

To evaluate fiscal solvency, several estimates for the sustainable level of net liabilities are presented in Table 2 and compared with the actual net liabilities position. The first measure, reported in column 3, is the precise application of equation (2); it is the product of the non-oil current primary balance plus seigniorage revenue and the discount factor.18 It assumes that inflation and, hence, seigniorage revenue are maintained perpetually at the rate recorded that year and that the return on public capital expenditure is equal to its cost to the government.

By this measure, the sustainable level of debt declined until the early 1980s, when substantial net assets were required, and has increased since then in response to the fiscal consolidation of the late 1980s. In recent years, debt of about 100 percent of GDP could be sustained if current expenditure, tax, and inflation revenues were maintained indefinitely. While the sustainable debt level has increased over time, the actual net asset position has declined markedly. Nonetheless, by this measure, actual net assets exceeded the required asset position at all times, indicating fiscal solvency. These trends in net assets and sustainable liabilities are illustrated in Figure 3 (which excludes pension liabilities).

Figure 3.Actual Net Assets and Sustainable Liabilities

(In percent of GDP)

To illustrate the sensitivity of the solvency calculation to particular assumptions about future developments, Table 2 also presents estimates of sustainable public liabilities under two alternative scenarios—zero inflation and lower growth. The assumption of zero inflation would imply a lower level of sustainable liabilities because no seigniorage revenue would accrue to the government. However, although the level of sustainable liabilities is markedly lower under zero inflation than under baseline assumptions, especially over the last three years (column 4), fiscal solvency is ensured by a relatively wide margin. The last column in Table 2 shows that if, in addition, 2 percent lower medium-term growth is assumed (approximately equivalent to raising real interest rates by 2 percentage points), sustainable liabilities would be limited to less than 50 percent of GDP.

The analysis in this section provides a useful framework for assessing sustainability issues under alternative scenarios, but the approach is based on the assumption that existing policies will be continued indefinitely into the future. If the recent improvement in the fiscal position turns out to be temporary, for example, because of cyclical factors, or if the projected expenditure and revenue trends are somewhat optimistic, then questions of sustainability could reemerge. Appendix III discusses this issue. It shows that while the fiscal solvency results appear robust to correction for cyclical factors, a permanent increase in expenditure by 1 percent of GDP—without a corresponding revenue increase—would make solvency more marginal. The current fiscal position in Indonesia appears to be solvent according to our measure and is expected to remain so, although the sensitivity analysis suggests limited capacity to withstand adverse interest rate or growth shocks. Moreover, given the decline in actual assets (with, now, an actual net liabilities position), solvency has become more reliant on sustaining projected fiscal surpluses and relatively high growth.

Fiscal Position and Inflation

In this section, we examine whether the fiscal position relies on inflation to meet debt obligations. If so, the authorities’ incentive to reduce inflation may be weakened. In principle, given that most government debt in Indonesia is denominated in foreign currency, there should not be any strong incentive to maintain relatively high inflation in order to lower the real value of the debt. However, other forms of public debt, including nonindexed pension obligations, can be reduced through inflation. Furthermore, inflation can help improve the fiscal balance in the short run through its effects on both revenues and government expenditures.19 To evaluate the effect of inflation on government finances, we first outline a conceptual framework to help identify the channels through which inflation affects the fiscal position and then assess the extent to which inflation has contributed to the public finances in Indonesia.

Conceptual Framework

One can assess the fiscal gains of inflation by determining the effect on the fiscal position of a lower rate of inflation. Following Persson, Persson, and Svensson (1996), we calculate a summary measure of the effect of a switch from high inflation (superscript H) to low inflation (superscript L) on the government budget, as follows:

where Zt, the nominal cash flow of the government budget, can be defined as

where dt is the discount rate, Tt represents the tax obligations in nominal terms, Pt is the price level, Gt is government spending commitments in nominal terms, and Bt is the sum of interest payments and maturing principal of the nominal public debt.

The budgetary gains from inflation consist of the present value of seigniorage, Mt+1 - Mt; the price-level effect on nominal tax and transfers through bracket creep (mitigated through the Tanzi inflation effect), Tt; nominally committed spending, Gt; and debt service, Bt. Measured appropriately, these present value gains can be compared with the government’s actual and sustainable liabilities specified earlier.

Application to Indonesia

In this section, we first assess whether fiscal solvency would be maintained with zero inflation. Then, posing this question the other way around, we attempt to quantify the fiscal effects for the Indonesian government, through seigniorage and through the erosion of nonindexed public obligations (retirement benefits), of maintaining moderately high inflation for five years.

Seigniorage

Table 3 reports estimates of seigniorage, as well as its decomposition into inflationary money creation and money growth as a result of real growth-induced changes in the demand for money (see Appendix IV for details).20 Seigniorage revenue in Indonesia is smaller than in high-inflation countries, declining from 2½-3 percent of GDP in the early 1970s to 1 percent of GDP in the 1990s. Nevertheless, at about 10 percent of revenue in 1995/96, it is not a trivial element of budgetary financing (e.g., it is higher than the share of excise revenue). Revisiting the fiscal solvency calculations in Table 2, column 4, demonstrates that inflation was not imperative for debt sustainability at any time. Even with seigniorage revenue assumed to be zero, sustainable liabilities exceeded the actual level. On this basis, net liabilities of about 64 percent of GDP could be sustained in 1995/96.

Table 3.Seigniorage(In percent of GDP)
YearSeignioragePure Inflation ComponentVelocity-Growth Component1Seigniorage (in percent of revenue)2
1972/732.61.41.218.4
1973/742.52.20.316.5
1974/752.72.10.614.6
1975/762.01.01.010.9
1976/771.71.10.68.6
1977/781.51.00.57.8
1978/791.11.2-0.16.0
1979/801.91.80.19.0
1980/811.71.50.27.3
1981/820.80.80.13.6
1982/830.60.50.13.1
1983/841.10.90.25.9
1984/850.70.40.34.1
1985/861.10.30.85.1
1986/871.00.30.75.9
1987/880.80.9-0.14.9
1988/89-0.20.4-0.6-1.3
1989/900.90.50.45.3
1990/910.30.4-0.11.3
1991/921.40.41.07.8
1992/930.40.30.02.1
1993/941.00.40.55.5
1994/951.10.40.66.5
1995/961.70.51.210.1
Source: Authors’ calculations from data provided by Indonesian authorities.

The income elasticity of demand is varied to ensure that the seigniorage components add up. Implicitly, velocity is variable, which explains the variation and the occasional negative values for the velocity-growth component. The average implicit income elasticity equals 0.9.

In Percent of revenue, including seigniorage.

Source: Authors’ calculations from data provided by Indonesian authorities.

The income elasticity of demand is varied to ensure that the seigniorage components add up. Implicitly, velocity is variable, which explains the variation and the occasional negative values for the velocity-growth component. The average implicit income elasticity equals 0.9.

In Percent of revenue, including seigniorage.

We also consider the present value gains from seigniorage when disinflation is delayed by five years. These are given by

where DL - DH equals the difference in the nominal discount rate between low and high inflation environments. We assume that inflation is maintained at 10 percent over the five-year period (above the long-run, low inflation rate assumption of zero), and project money growth is maintained at a constant rate equal to the average over the last five years, when the inflation rate was close to 10 percent a year. Assuming the real interest rate of 7.5 percent is unaffected by the change in inflation, the nominal interest rate over the next five years is assumed to be 10 percentage points above that which would prevail in a low inflation environment. We assume zero seigniorage revenue beyond the five-year period. On this basis, the present value of seigniorage revenue as a result of delayed disinflation amounts to about 2 percent of GDP, which, by itself, has a very small effect on the fiscal position.

Nonindexed Government Obligations

To further illustrate the effects of inflation, we also investigate the revenue gains from delayed disinflation through diminution of the real value of public liabilities. Because the public debt is almost entirely held in foreign currency, the effects are confined to projected annual civil service pension liabilities (and post-retirement health benefits), which, in present value terms, are estimated to amount to about 25 percent of GDP in 1995/96. Civil service pension benefits are adjusted on a discretionary basis by the government, generally in line with wage movements, but there is no formal requirement for these benefits to be indexed to inflation. The gains from maintaining high inflation are given by

where B95t is the period t nominal cash-flow pension obligations in 1995. With B95t equal to approximately 0.6 percent of GDP, rising to almost 1 percent of GDP by about 2020, the gain resulting from inflation erosion is estimated to be nearly 10 percent of 1995/96 GDP. The zero-indexation assumption is, however, intended only to provide an illustrative scenario—there is no presumption here that such a policy is either feasible or desirable.

The analysis in this section strongly suggests that inflation is not required for fiscal solvency in Indonesia. Although there may be some disincentive to reduce inflation when debt levels are high, the maximum potential present value revenue loss as a result of moving from 10 percent inflation to zero inflation is only just over 10 percent of GDP. This result must be weighed against the longer-term costs of inflation, in terms of both the welfare costs and the negative effect on the economy’s longer-term growth potential. Fiscal consolidation would, in any case, eliminate any disincentive to reduce inflation.

Fiscal Position and External Sustainability

We address, in this section, the issue of external current account sustainability for Indonesia and its relationship to the fiscal position. We do so, first, by examining the trade surplus necessary to service Indonesia’s foreign liabilities,21 and second (drawing on Milesi-Ferretti and Razin, 1996), by reviewing a number of indicators that have a bearing on current account sustainability.

One argument for linking external sustainability with fiscal sustainability is demonstrated in Figure 4, which charts Indonesia’s external current account balance against the fiscal balance. Fiscal consolidation has generally been associated with an improvement in the external balance, and fiscal loosening with current account deterioration. Until 1990/91, the two indicators moved closely together. In large part, this correspondence reflected the impact of oil price movements on the two balances—when oil was more important to each indicator than it is now—but it also reflects the effects of discretionary policy changes; the ratio of government expenditure to GDP follows a similar pattern.

Figure 4.External Current Account and Fiscal Balances

(In percent of GDP)

The apparent weakening of the relationship since 1990/91 can be attributed to a number of factors. First, fiscal policy has become more actively countercyclical; deterioration in the external current account balance in 1990/91 and 1995/96 spurred a fiscal policy surplus, while the economic slowdown in 1992/93 and 1993/94 led to fiscal easing. Second, against the background of high debt levels, structural improvement in the fiscal position in the late 1980s appears to have offset private sector savings substantially, thereby limiting the rise in national saving. The increase in confidence, however, crowded in private sector investment, with the result that the combined effect of these two factors led to a smaller improvement in the external balance than in the fiscal accounts. Third, monetary policy, which at times moved in an opposite direction, may have offset some of the contractionary effects of fiscal policy (e.g., money and credit growth were very rapid in 1990/91 and 1995/96). Nevertheless, evidence over the longer term and across countries suggests that tighter fiscal policy would boost national savings and reduce external current account imbalances.22

Conceptual Framework

External solvency can be evaluated in a way that is analogous to the assessment of fiscal solvency. In this case, solvency requires that the present value of trade surpluses (in future periods) be high enough to repay the country’s net external liabilities (similar to the correspondence between fiscal surpluses and public debt shown in the fiscal solvency condition earlier). The sustainable level of net foreign liabilities (external debt less foreign reserves) will depend on the trade balance (defined as net exports of goods and services) and the discount factor defined earlier, incorporating the (foreign) interest rate, the rate of depreciation of the real exchange rate, and the growth rate. (See Appendix V for a detailed derivation of the solvency condition.) The level of foreign liabilities that can be sustained rises with increases in the trade balance and the growth rate and with decreases in the real interest rate. The calculation of external solvency is, however, extremely sensitive to the components of the discount factor. In particular, it hinges critically on the relationship between the projected real interest rate and the projected growth rate. In general, with net liabilities and trade deficits, external solvency requires that the real interest rate be lower than the growth rate—effectively implying that countries will “grow out of debt.” Conversely, if the projected real interest rate exceeds the growth rate, external solvency requires trade surpluses when existing foreign liabilities are high. This issue is discussed further in the next section, where this framework is applied to Indonesia.

The approach also requires projections for annual trade balances indefinitely into the future, but because these are typically subject to relatively large fluctuations from year to year, we apply a three-year moving average of the trade balance to dampen cyclical factors and to help identify trends. In the general formulation of this framework, it is not possible to distinguish between trade deficits that reflect primarily imports of consumer goods and those that reflect mainly capital goods imports. The latter may well foreshadow higher national income and exports in future periods, thereby increasing the sustainable level of debt. Nevertheless, this notion of solvency provides some guidance on the required (or expected) adjustment in the trade balance and on the effects of sharp movements in the exchange rate or in the real interest rate in the event of adverse external developments or domestic disturbances.

Application to Indonesia

The evaluation of external sustainability in Indonesia depends on assumptions about the future profile of trade balances, the growth rate, world interest rates, and the real interest rate. However, the effective real interest rate on Indonesia’s foreign liabilities does not depend only on the nominal rate of interest on external debt. It also depends on the rate of appreciation or depreciation of the real exchange rate because the effective cost of debt service—measured in units of GDP—will be determined by the value of domestic output relative to that of foreign output. As noted earlier, calculations of external solvency are extremely sensitive to assumptions about future real interest rates and, therefore, also to assumptions about the future path of the real exchange rate. In the present analysis, we have assumed that the real exchange rate will remain constant. As with the analysis of fiscal solvency, the present value of oil reserves is included in calculations of Indonesia’s net foreign assets, which implies that sustainability should be assessed with regard to the trade balance on products other than oil and gas.

Actual net foreign assets including oil reserves were nearly 500 percent of GDP in 1979/80. However, with the declining value of oil reserves, net foreign assets diminished sharply and are estimated to have become negative (a net liabilities position) in 1993/94. Actual net foreign liabilities were estimated to be 13 percent of GDP in 1995/96. We can compare this actual foreign liabilities position with the sustainable level of foreign liabilities. If the real interest rate is lower than the growth rate, a high level of foreign liabilities is sustainable, provided that the trade deficit is not larger than the increase in debt service capacity resulting from economic growth. To illustrate, we assume that the real interest rate is 5.9 percent (the average rate recorded over 1990/91-1995/96, assuming a constant real exchange rate) and that the growth rate is 7.1 percent. Then, if the non-oil trade balance is equal to 1.8 percent of GDP (its average for the last three years), sustainable liabilities equal 160 percent of GDP.

Sustainability indicators

The analysis in the previous section provides only one yardstick, albeit an important one, for assessing sustainability on the basis of projections for the key determinants of the level of external indebtedness. In practice, however, policymakers need to focus on a wider range of indicators to assess sustainability and to ensure that policy responds to emerging external imbalances long before the external position becomes insolvent. Both favorable and unfavorable experiences illustrate that persistent current account deficits and high debt levels leave countries more vulnerable to terms of trade and real interest rate shocks.

Indonesia’s debt-GDP and debt-servicing ratios are relatively high by ASEAN standards (Table 1), In the event of adverse external developments, sustaining the external position may require strong domestic macroeconomic adjustment measures. Indonesia’s real interest rates have proved volatile—at least in an absolute sense—as have its terms of trade.23 Since 1971, the terms of trade have fluctuated by an average of 11.5 percent a year (in absolute values), more than in other ASEAN countries (Table 4), in part reflecting the predominance of agricultural, mineral, and forestry commodities in exports, whose prices have been more volatile than the prices of manufactures. Although terms of trade and export price volatility have been lower since the 1980s, domestic income and, hence, the fiscal position have still been subject to substantial swings resulting from the terms of trade movements. This effect is partly mitigated by the relatively low share of trade in economic activity in Indonesia. Fortuitously, more of the recent swings have been positive on average (Indonesia’s terms of trade have remained 20 percent above their long-term historical average).24

Table 4.External Trade Price Volatility
1971-95
Includes oil and gasExcludes oil and gas1971-801981-95
Includes oil and gas
Export prices
Indonesia0.3090.2390.4130.142
Malaysia0.1510.2020.151
Philippines0.1820.2700.182
Singapore0.1280.1620.128
Thailand0.1650.2230.165
Developing countries0.2140.2820.214
Industrial countries0.0820.0700.082
Import prices
Indonesia0.1180.1600.072
Malaysia0.1150.1430.042
Philippines0.1460.1980.067
Singapore0.1170.1420.095
Thailand0.1080.1200.070
Developing countries0.0930.1090.041
Industrial countries0.1070.1150.059
Terms of trade
Indonesia0.1740.1290.2140.106
Malaysia0.1280.1620.093
Philippines0.0940.1150.076
Singapore0.0360.0510.020
Thailand0.1020.1410.070
Developing countries0.1010.1360.041
Industrial countries0.0300.0380.019
Source: IMF, International Financial Statistics Yearbook (various years).Note: Standard deviation of percent change in trade prices.
Source: IMF, International Financial Statistics Yearbook (various years).Note: Standard deviation of percent change in trade prices.

As argued in Milesi-Ferretti and Razin (1996), an assessment of current account sustainability should take into account the willingness of capital markets to lend, which depends on a number of structural features of the economy and on prudent economic policies. They found, in particular, that sustainability is influenced by

  • high national savings and investment and the rate of GDP growth, which can signal debt-servicing capacity. Indonesia compares relatively favorably on these indicators, with savings and investment rates above 30 percent of GDP and with real GDP growth averaging 8 percent.

  • the composition of external liabilities. In principle, equity financing allows asset price adjustments to absorb at least a part of negative shocks so that foreign investors bear some of the risk. In contrast, with debt financing, the country bears most of the burden. About one-third of Indonesia’s net capital flows in the 1990s have been foreign direct investment. Consequently, its net external liabilities—measured by cumulative current account deficits—are substantially above its net external debt. Indonesia has relatively low receipts of portfolio (mostly debt) inflows (1.7 percent of GDP), and short-term debt is low. Thus, the composition of the external liabilities is sufficiently diversified.

  • large movements in the real exchange rate. Among the countries studied by Milesi-Ferretti and Razin (1996), current account deficits often exceeded 5 percent of GDP, but external problems were typically associated with large real exchange rate appreciations. The ASEAN countries have generally registered real exchange rate depreciations of over 10 percent on average in the 1990s and this has helped sustain external positions. In Indonesia, the real exchange rate depreciated very sharply in the mid-1980s, with large nominal depreciations following declines in oil prices, and was sustained by Indonesia’s practice of depreciating the nominal rate in line with domestic-world inflation differentials thereafter (see Figure 5). The real exchange rate in 1995 was low (depreciated) relative to Indonesia’s historical average, and there are no indications that the rate is overvalued, which bodes well for external sustainability.25

  • the fiscal stance. Although the fiscal balance is not an unambiguous predictor of external debt problems, Milesi-Ferretti and Razin found that the turnaround from current account deficits toward surpluses was driven in most cases by fiscal consolidation. Against the countries they examined, Indonesia compares well in terms of its average fiscal position and the current account deficit.

Figure 5.Value of Oil Reserves and the Real Exchange Rate

Taken together, structural features of the Indonesian economy and its policy characteristics would seem to invite confidence in the resilience and sustainability of the debt outlook, should adverse shocks eventuate.26 Nevertheless, given Indonesia’s net foreign liabilities position, further fiscal consolidation would strengthen sustainability and support a turnaround in the trade balance. A lower level of debt would also provide the government with more scope to accommodate shocks through increased external borrowing, rather than obliging the government to impose tax increases or to force expenditure reductions, which may compound the real shock.

Conclusion and Policy Implications

This paper has examined a number of aspects of the sustainability of the Indonesian fiscal position. We considered both public debt and external sustainability and took account specifically of the exhaustibility of oil and gas reserves. There is no issue of fiscal solvency at present, with government net worth positive on all measures. There is, however, a potential vulnerability to interest rate and growth rate shocks, under the assumption of zero seigniorage and allowing for public pension liabilities.

We considered the impact of achieving low rates of inflation and showed that the potential loss of seigniorage revenue and the higher real burden of nonindexed public liabilities do not endanger solvency and, therefore, should not deter disinflation. However, fiscal consolidation would remove any perceived obstacle, as well as support disinflation through demand-reducing effects.

Indonesia’s share of trade in GDP is lower than that in other ASEAN countries, and debt-servicing ratios are relatively high, implying a potential vulnerability to declines in market demand or in investor confidence. Nevertheless, sustainability of external debt would seem to be ensured, given such structural features of the Indonesian economy as its high rate of investment and output growth. In light of the evidence that the private sector offset to higher public savings is less than complete, further fiscal consolidation would strengthen the non-oil trade balance and help reinforce external sustainability. Strong crowding in of private investment may reduce the measured improvement in the external current account, but it would probably still improve external sustainability through higher prospective trade surpluses in future years.

Appendix I. Data Sources and Description

Indonesian fiscal and balance of payments data are reported on a fiscal year (April 1-March 31) basis. Calendar year national accounts and inflation data are transformed into fiscal years in this paper. Fiscal data were supplied by the Indonesian authorities, except for provincial government data from 1975/76 to 1979/80, which are drawn from the IMF’s Government Finance Statistics.

The overall fiscal balance is derived as follows:

-BALTF = DF + FFE - R - G,

where BAL is the overall fiscal balance, total financing (TF) is the sum of domestic financing (DF) and foreign financing (FF), E is total expenditure plus net lending, R is total revenue, and G is grants. Then,

E = CE + DNLER + G +TF,

and

DNLER + G - CE + TF,

where CE is current expenditure, and DNLE is development and net lending.

The overall fiscal balance is derived from data on foreign financing in the balance of payments and domestic financing given by changes in net government deposits with the banking system (at Bank Indonesia and deposit money banks). The calculation of expenditure follows from the assumption that revenues are reasonably thoroughly reported. Therefore (utilizing the identity that E = TF + R + G), financing movements are allocated to development expenditure and net lending, because inadequate information is available to allocate them among current, investment, net lending, and revenue categories. Consequently, the measure of capital expenditure utilized for the purposes of fiscal solvency calculations includes net lending transactions. Current expenditure is defined to include regional transfers (although this is partly for capital investment) and spending on military personnel (treated as development expenditures in the Indonesian budget system). Current expenditure may be understated to the extent that substantial expenditures on education and health are allocated to development expenditure in the Indonesian budget, although these may not be totally capital items.

Other fiscal variables are defined as follows: current fiscal balances are revenue plus grants less current expenditure; primary balances are fiscal balances plus interest expenditure (which is included in current expenditure), and current primary balances exclude capital expenditure. For the purpose of the solvency analysis, the non-oil current primary tax balance is calculated excluding grants (on the grounds that grants are not an enduring revenue stream) and excluding nontax revenues (i.e., tax revenues other than those from gas and oil less noninterest current expenditure).

The statistical analyses in the paper are confined to central government aggregates. Provincial and central government accounts are consolidated into general government accounts from 1975/76 by eliminating provincial government transfer receipts from central government. Consolidated expenditure and revenue aggregates are generally about 1 percent of GDP higher than central government estimates, but the fiscal balance is generally little affected. All fiscal figures exclude social security transactions.

All ratios to GDP are defined relative to current-period GDP, except for those variables in Table 2 for the fiscal solvency analysis, which are defined as in Appendix II.

Nominal national accounts data (consumption, investment, imports and exports, GDP, GNP, and national income) from 1971 to 1987 are from the IMF’s International Financial Statistics Yearbook, and are spliced with new national accounts data from 1988 to 1995 on the basis of information supplied by the authorities for Article IV consultations. Constant-price figures are derived by deflating the nominal series by the GDP deflator, and the real GDP components therefore differ from published national accounts statistics. The implicit GDP deflator movement is derived from the 1983-based GDP series until 1988, and spliced with the percentage change in the deflator from the 1993-based GDP series from 1989 to 1995. All projection values for 1996/97-2001/2002 are drawn from the projections contained in the IMF’s World Economic Outlook.

The current account balance is drawn from data supplied by the authorities. The trade balance for the current account sustainability analysis is calculated as net exports of goods and nonfactor services from the national accounts (which is equal to Y - C - I).

Consumer price index (CPI) inflation data are from International Financial Statistics. External debt data for 1972-79 are from International Financial Statistics and, for 1980-95, are from data supplied by the authorities.

The value of oil and gas reserves to the government budget is calculated by deflating the revenue stream (from the fiscal data) by the oil price (from International Financial Statistics), and totaling the derived volumes, which are then valued at the current market price, according to Appendix II. The volumes of production and prices through to 2010/2011 are projected to equal the average annual volumes and prices over the last five years. The value of oil reserves to the economy (for the purposes of calculating national wealth) is the stream of oil revenues grossed up by 1/to where to is assumed to be 0.85, as the government receives 85 percent of the net operating income of the private oil-producing companies plus transfers from Pertamina.

Monetary and exchange rate data were taken from International Financial Statistics until 1987, and from data supplied by the authorities thereafter. The real exchange rate is from the IMF’s Information Notice System beginning in 1979. Before 1979, it is calculated from relative nominal exchange rate and inflation rates versus the U.S. dollar, and the series is spliced.

Nominal interest rates are from International Financial Statistics. Interest payments on external debt are from the World Economic Outlook. Real interest rates are calculated as follows: (1) the interest rate on U.S. dollar external debt is deflated by the average of the percent change in Indonesian export unit values and world export unit values; and alternatively (2) the U.S. ten-year bond rate plus the nominal rate of exchange depreciation divided by the GDP deflator.

Actual and projected pension payments were calculated from data supplied by the World Bank.

Appendix II. Fiscal Solvency Derivation

Consolidating the central bank’s budget constraint with the (non-oil) budget constraint of the government, the change in debt between one period and the next is derived from the government’s flow budget constraint, as follows:

where Bt is the public external debt in dollars; St is the rupiah-U.S. dollar exchange rate; it is the nominal domestic rate of interest; Pt is the GDP deflator; Dt is the overall non-oil current primary deficit, defined as total current expenditure excluding interest payments less total non-oil tax revenue (in constant prices); and Mt is nominal base or reserve money. Stock variables are valued at the beginning of the period, and flows are assumed to occur during the period.

Assuming interest parity and expressing this flow constraint in terms of units of GDP, we derive

Then, we can show that

where lowercase letters convey variables in terms of units of GDP, the numeraire, and

and

Consider now the economy in a steady state, in which all the fiscal magnitudes are constant as a fraction of GDP. Steady-state bt+1 is equal to bt; thus, henceforth, we omit time subscripts for steady-state variables. We observe that the level of debt that can be sustained in the steady state is equal to

where ε is the real rate of depreciation, and we ignore the term ε·i*. We also provide estimates where capital expenditure is assumed to earn economic return, but possibly below the market rate of return. Equation (2) is modified as follows:

where the stock of accumulated public capital expenditure, defined in proportion to GDP, is denoted as k, with rate of return ρ, and the deficit d′ is now defined to incorporate capital expenditure.

Derivation of Value of Oil and Gas Reserves

The value of oil and gas reserves, et, is defined as follows:

where Et represents known oil reserves, in millions of barrels, and PtE is the price of oil, in dollars.

We estimate the value of the reserve, Et, from the annual oil revenue flows, as follows. The value of gross reserves can be determined according to the relation Pt(Et+1 - Et) = - PtRt which, for N periods, yields

where δt is the long-run rate of increase in the price of oil, and Rt represents annual oil production. If we assume that the price of oil rises at a rate equal to the real rate of interest (the Hotelling-Solow rule), equation (5) simplifies to

Because we are interested in the economic value (or rent) of the oil reserves, the value of the reserves (and the annual production stream) should be reduced by the costs of extraction. Our methodology for calculating the value of the reserves accomplishes this, because we derive values for these variables from government revenues from oil and gas production. Because oil tax revenues represent a fixed proportion (85 percent) of the net operating income of the oil production companies, we can deflate the revenue stream by petroleum prices to obtain the real resource rental or “net production” stream, and proceed accordingly.

Appendix III. Fiscal Solvency—Cyclical Adjustment and Medium-Term Projections

GDP growth that is above trend tends to buoy the fiscal position by boosting income-elastic tax revenues and, in many countries, diminishing social security expenditures through cyclical declines in unemployment. Estimates of the output gap for Indonesia place actual GDP at about 1 percent higher than potential GDP in 1994 and 1995. Cyclical effects on the budgetary position, therefore, would be expected to be relatively small at this time. Moreover, given that benefit expenditures are zero, public expenditure is largely independent of changes in output (except through discretionary or policy linkages). Neither is oil revenue closely linked to economic activity. With only relatively low non-oil tax revenues linked to GDP growth, the fiscal balance is calculated to be about 0.2 percent of GDP higher in 1995/96 as a result of above-trend output. Thus, the fiscal solvency results appear robust to correction for cyclical factors.

Over the medium term, the following factors are expected to influence public debt sustainability:27

(1) The actual public liabilities position excluding pension liabilities is projected to worsen marginally, increasing by about 7 percent of GDP over the next five years (Table 2). Net public debt falls gradually in proportion to growing GDP, but the projected decline in the value of oil reserves, to 12 percent of GDP by 2000/2001, is greater, so that overall net public liabilities excluding pensions increase.

(2) Projections of the present value of unfunded pension obligations are sensitive to assumptions on wages and investment earnings. Assuming pensions increase by the rate of inflation (close to 10 percent through the next five years, in line with the World Economic Outlook projections, and assumed to be 5 percent in the long run) plus 3 percent real wage growth, the present value of unfunded pension liabilities increases marginally slower than GDP, partially offsetting the rise in other public liabilities. Total public liabilities, including pensions, are projected to increase slightly, from 7 percent of GDP in 1995/96 to 12 percent of GDP in 2000/2001.

(3) The sustainable liabilities position is projected to remain broadly unchanged. Assuming no change in tax policy, non-oil tax revenues are projected to be constant in relation to GDP. Income and sales tax revenues can be expected to grow marginally taster than nominal output, but trade and excise taxes more slowly. Turning to expenditure, Repelita VI (sixth five-year development plan) plans for higher education and medical access, with the likelihood of higher public expenditure in these areas. At the same time, there is some scope for reallocating expenditure and efficiency improvements to allow greater services from existing real expenditure levels (World Bank, 1993). Current expenditure in Indonesia is lower than in other ASEAN countries (Rumbaugh, 1995). However, there has been no discernible trend in current expenditure in relation to GDP in recent years, although it rose somewhat in 1995/96. While expenditure pressures exist, we have chosen to project current expenditure as a constant ratio to GDP over the next five years—implying a real spending increase of over 7 percent a year. With revenues and expenditures stable in relation to GDP, the non-oil current primary balance, obviously, is also projected to remain unchanged, as is the level of sustainable liabilities. (We assume constant real interest and growth rates, at their historical averages.) The overall fiscal balance, on these assumptions, would move into a deficit of about 1 percent of GDP, driven by falls in oil revenue.

(4) Assuming faster growth in current expenditure would lower the level of sustainable liabilities. To illustrate, current expenditure that is 1 percentage point higher as a share of GDP—without a corresponding increase in non-oil revenue—would lower the sustainable level of liabilities by about 25 percent of GDP.

With the level of sustainable liabilities basically unchanged and actual liabilities (including pensions) increasing slightly, net worth (the difference between the two) is projected to decline marginally over the period to 2000/2001. According to our measure, the Indonesian fiscal position would appear to be solvent and is expected to remain so on the basis of these assumptions.

Appendix IV. External Solvency Derivation

A solvent external current account balance is one in which the present value of trade surpluses is sufficient to repay the country’s net external liabilities. We express this condition as the long-run net resource transfer (trade surplus) that a country must make to keep net liabilities constant in relation to GDP. The basic accounting identity is as follows:

where CA is the current account surplus, F is net foreign assets, Y is GDP, C is private consumption, CG is government consumption, and V is investment (private plus government). The domestic GDP deflator, the world rate of interest, and the nominal exchange rate are defined as P, i*, and S, as before, and P* is the foreign GDP deflator. Let the ratio of foreign assets to output ft, be equal to StPt*Ft/PtYt. Dividing both sides of the equation by nominal GDP and rearranging terms, we obtain

where the economy is assumed to grow at rate g, the real rate of depreciation is ε, and the trade balance (or net exports of goods and nonfactor services) is defined as tb = 1 - c - v - cg. Ignoring the term ε·g, and with the economy in a steady state in which consumption, investment, and the stock of foreign assets are constant in proportion to GDP, the sustainable foreign asset position is given by

We simplify this expression and redefine it in terms of sustainable liabilities as f = ψtb, where ψ denotes the discount factor.

As with the fiscal solvency analysis, we adjust for the treatment of the oil resource by including the asset value of oil reserves in the net foreign asset position and assessing sustainability in terms of the trade balance on products other than gas and oil.

Appendix V. Seigniorage Decomposition

The revenue from money creation can be expressed as the sum of two terms; the first is due to inflationary money creation, and the second is that money growth that is attributable to real growth-induced changes in the demand for money (we ignore the term associated with the acceleration of inflation):

where η is the income elasticity of the demand for (base) money.

References

    AsherMukul G.1989“A Comparative Overview of ASEAN Fiscal Systems and Practices,” in Fiscal Systems and Practices in ASEAN: Trends Impact and Evaluationed. byM.Asher (Pasia Panjang, Singapore: ASEAN Economic Research Unit, Institute of Southeast Asian Studies).

    AsherMukul G. and AnneBooth1992“Fiscal Policy,” in The Oil Boom and After: Indonesia Economic Policy and Performance in the Soeharto Eraed. byA.Booth (Singapore; New York: Oxford University Press).

    BarroRobert J.1979“On the Determination of the Public Debt,”Journal of Political EconomyVol. 87 (October) pp. 94071.

    BlanchardOlivier1990“Suggestions for a New Set of Fiscal Indicators,”Working Paper No. 79 (Paris: Organization for Economic Cooperation and Development, Economics and Statistics Department).

    BuiterWillem and UrjitPatel1995“Budgetary Aspects of Stabilization and Structural Adjustment in India: The Painful Road to Sustainable Fiscal-Financial-Monetary Plan,”paper presented at Conference on Economics, Politics, and Policies of Stabilization, Structural Adjustment, and Long-Term DevelopmentYale UniversityNovember.

    DornbuschRudiger and StanleyFischer1993“Moderate Inflation,”World Bank Economic ReviewVol. 7 (January) pp. 144.

    Economic & Business Review Indonesia (1996) “Oil Reserves: Looking at the Bottom of the Barrel,”No. 184 (October21) pp. 611.

    HansonJames1994Indonesia: Stability Growth and Equity in Repelita VI (Washington: World Bank).

    HillHal1996The Indonesian Economy Since 1966: Southeast Asia’s Emerging Giant (Cambridge, United Kingdom; New York, New York: Cambridge University Press).

    International Monetary FundInternational Financial Statistics Yearbook (Washingtonvarious years).

    International Monetary FundGovernment Finance Statistics (Washingtonvarious issues).

    International Monetary FundWorld Economic Outlook: A Survey by the Staff (Washingtonvarious issues).

    LiuksilaClaireAlejandroGarcia and SheilaBassett1994“Fiscal Policy Sustainability in Oil-Producing Countries,”IMF Working Paper 94/137 (Washington: International Monetary Fund).

    Milesi-FerrettiGian Maria and AssafRazin1996“Current Account Sustainability,”Princeton Studies on International Finance (Princeton, New Jersey: Princeton University Press).

    MolhoLazaros1994“Fiscal Adjustment in an Oil-Exporting Country: The Case of Indonesia,”IMF Paper on Policy Analysis and Assessment 94/21 (Washington: International Monetary Fund).

    NasutionAnwar1989“Fiscal Systems and Practices in Indonesia,” in Fiscal Systems and Practices in ASEANed. byM.Asher (Pasia Panjang, Singapore: ASEAN Economic Research Unit, Institute of Southeast Asian Studies).

    Organization of Petroleum Exporting Countries1995Annual Statistical Bulletin (Vienna, Austria).

    PerssonMatsTorstenPersson and LarsSvensson1996“Debt, Cash Flow, and Inflation Incentives: A Swedish Example,”NBER Working Paper No. 5772 (Cambridge, Massachusetts: National Bureau of Economic Research).

    RadeletSteven1995“Indonesian Foreign Debt: Headed for a Crisis or Financing Sustainable Growth?”Bulletin of Indonesian Economic StudiesVol. 31 (December) pp. 3972.

    RumbaughThomas1995“Fiscal Policy: Short-Run Stabilization and Medium-Term Adjustment” (unpublished).

    TanziVito1977“Inflation, Lags in Collection, and the Real Value of Tax Revenue,”Staff PapersInternational Monetary FundVol. 24 (March) pp. 15467.

    World Bank1993“Indonesia: Issues and Policy Options for Repelita VI” (unpublished; Washington).

    World Bank1995Global Economic Prospects and the Developing Countries (Washington).

    World Bank1996“The Indonesia Pension System: Structure, Policy and Implementation” (unpublished; Washington).

    WooWing ThyeBruceGlassburner and AnwarNasution1994Macroeconomic Policies Crises and Long-Term Growth in Indonesia 1965-90 (Washington: World Bank).

Note: Geoffrey Bascand is an Economist in the IMF’s Asia and Pacific Department, and Assaf Razin is Daniel Ross Professor of International Economics at Tel Aviv University anda Research Associate at the National Bureau of Economic Research. The authors thank Christopher Browne, John Hicklin, Sandy Mackenzie, Mahmood Pradhan, David Robinson, Dani Rudrik, Thomas Rumbaugh, and Anoop Singh, along with participants in a seminar held at the IMF in August 1996, for helpful suggestions on earlier versions of this paper.

As far as we are aware, similar treatments do not exist in the literature for Indonesia. See, for example, Asher (1989); Asher and Booth (1992); Hill (1996); Molho (1994); Nasution (1989); Rumbaugh (1995); Woo, Glassburner, and Nasution (1994); and World Bank (1994, 1995) for discussions of focal policy in Indonesia. See Radelet (1995) for a treatment of Indonesia’s public debt sustainability. Liuksila, Garcia, and Bassett (1994) provide a similar treatment of fiscal solvency in the presence of exhaustible resources to that given later, but do not address the other aspects of sustainability considered here.

Assuming the government faces limits in its capacity to borrow, then it is efficient to maintain a positive fiscal position (net worth), which can be run down in the event of adverse shocks, thereby avoiding the need to increase tax rates; see, for example, Barro (1979).

Logically, there is an optimal path for the government’s fiscal position, which will depend on (1) economic structure and risk characteristics (e.g., creditors may accommodate higher government gearing when die economy is diversified and exposure to shocks is low); (2) rates of time preference (the value of expenditures and revenues today compared with their value in future periods); and (3) the efficiency costs of taxation and their relationship to current and future tax rates. Sustainability, which we focus on, is a weaker condition than optimality and generally abstracts from (2) and (3).

Appendix I provides definitions and sources of the data used in this paper.

Because Indonesia is large relative to the other economies, the share of exports in GDP is expected to be lower.

This includes extrabudgetary transactions and treats net foreign borrowing as a financing item.

The Indonesian fiscal year runs from April 1 to March 31.

Asher and Booth (1992); the dates are implementation dates, which are often one to two years later than the legislation dates.

Sec Buiter and Patel (1995) and Blanchard (1990). Gerson and Nelior provide an alternative analysis for assessing fiscal solvency in the Philippines (see Chapter 6 in this volume). While their conceptual approach is similar, the empirical technique they use to assess solvency relies on information contained in a statistically identified trend in public debt, whereas the method used here extrapolates annual fiscal settings into the future.

Given that solvency is a matter of financial viability over a number of years, we do not analyze the effects of transitory revenues or expenditures, which, by definition would be difficult to project into the future.

Our deficit measure therefore includes an accrual estimate for central bank revenue.

We use this formulation with the foreign interest rate and the real rate of depreciation rather than the more traditional domestic real interest rate because Indonesian government debt is denominated almost entirely in foreign currency. Note, however, that in equation (2) we omit the term ε·i*.

Omitted from the standard notion of net worth are real government assets, for which no estimates are available. The estimated value of oil reserves is included below.

To the extent that die returns on public sector capital investment are less than the rate of interest, our measure understates the deficit and overstates the level of liabilities that could be sustained. Appendix II modifies equation (2) to allow for this possibility.

The Organization of Petroleum Exporting Countries (OPEC) (1995) reports oil reserves of 5.13 billion barrels (bbls.) and natural gas reserves equivalent to 19.4 bbls. of oil (of which approximately 7.5 bbls. is the Natuna field), as of December 1994. The current depletion rate is approximately 0.6 bbls. a year for oil and 0.5 bbls. a year for natural gas. We have assumed two-thirds of gas reserves are economically viable, since potentially high costs of extraction render the value of a portion of the gas reserves uncertain. See also Economic & Business Review Indonesia (1995), which reports similar figures for reserves from the Ministry of Mines and Energy.

Oil and gas reserves are estimated now to stand at just under 30 percent of GDP. The decline in value is partly attributable to depletion of the resource, but mostly to lower real petroleum prices over the period, especially when measured in units of Indonesian GDP.

Unfunded liabilities here mean the share of pension obligations not met by ongoing government pension payments, which are included in current expenditure. The calculations of unfunded liabilities draw on work undertaken by the World Bank (1996) and Mani Sundaresan of Watson Wyatt Consulting Actuaries in estimating the government’s pension contribution, hut use our own assumptions of wage growth and the discount rate. Estimates for pension and post-retirement health care liabilities are not available before 1994/95.

In calculations of the discount factor, the nominal interest rate has the value of 6.4 percent (the average rate on external dollar debt for 1982/83–1995/96); the real rate of exchange depreciation averaged 4.6 percent from 1981/82 to 1995/96. Together, these assumptions imply a relatively high real rate of interest. If the real interest rate is calculated on the basis of the real exchange rate experienced in the 1990s—averaging 1.6 percent—the real rate of interest (7.5 percent) is only marginally above the growth rate, averaging just over 7 percent of the long term, implying a higher sustainable debt level.

However, Tanzi (1977) identified another effect of inflationary finance that, in contrast, operates to reduce real tax revenues when inflation rises. Owing to collection lags, defined as the time that elapses between the date when the tax liability accrues and the time when the tax payment is received by the government, inflation causes an erosion of the real tax revenue. The collection lag can be shortened or interest applied to the payment to reduce this loss.

See also Dornbusch and Fischer (1993) for a cross-country study of countries with moderate inflation, and the relative seigniorage and revenue position of Indonesia in the 1970s and 1980s in this group of countries.

For an alternative approach, see Chapter 4 in this volume. See also Radelet (1995) for an examination of Indonesia’s external debt sustainability.

Dayal-Gulati and Thimann (Chapter 7 in this volume) provide statistical evidence supporting this proposition for the group of ASEAN countries. Regressions performed by the authors confirmed this result for Indonesia.

Technically, one standard deviation in the real interest rate (the nominal rate on foreign debt plus real exchange rate depreciation) for the period 1982/83–1995/96 is 10.3 percent, against a mean rate of 11.4 percent.

In 1995, non-oil export prices were 32 percent higher than in 1993, while oil prices were also relatively high in 1995 and 1996.

Seo Chapter 11 in this volume for further analysis of this issue.

Radelet (1995, p. 67) provides a similar analysis and concludes that “Indonesia does not appear to be headed toward a debt crisis, either now or in the immediate future. If the trends in exports and GDP growth recorded since 1989 continue, the debt problem should ease gradually‖. But the debt situation leaves little room for error.”

Projections are derived from the World Economic Outlook for growth and inflation. Our fiscal projections commence with the 1996/97 budget figures, which are then driven forward according to the assumptions noted. The main difference from the projections contained in the World Economic Outlook is a lower overall fiscal balance because the World Economic Outlook assumes that policy adjusts to increase non-oil tax revenue to offset declining oil revenues and maintain fiscal balance.

    Other Resources Citing This Publication