Information about Asia and the Pacific Asia y el Pacífico

CHAPTER 3 Maintaining Fiscal Sustainability Under Uncertainty

Thomas Rumbaugh
Published Date:
January 2012
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

Indonesia’s public debt outlook is stronger than in many advanced and emerging economies. Nevertheless, Indonesia, like other emerging economies with relatively low debt levels, is still exposed to shocks. This chapter presents considerations for setting up a fiscal strategy in Indonesia aimed at maintaining sustainability while managing uncertainties and risks. Stochastic simulations confirm that a medium-term fiscal consolidation strategy, based on subsidy reduction and revenue administration reforms in line with the authorities’ framework, is robust to macroeconomic and oil price shocks. However, delaying subsidy reform could increase fiscal vulnerabilities in the context of rising fuel consumption, volatile oil prices, and oil production shocks.


Indonesia’s public-debt-to-GDP ratio has been on a declining trend since 2000 and fell to 29 percent in 2009, well below the average for emerging and advanced countries (Figure 3.1). Prudent fiscal management resulting in sustained primary fiscal surpluses (1.6 percent of GDP per year on average since 2000), combined with favorable debt dynamics, supported a continuous reduction in the debt ratio, which in 2009 stood at about a third of its 2000 level. Foreign currency debt has also been reduced to less than half of total debt because the improved fiscal position facilitated domestic capital market access. The authorities’ medium-term fiscal strategy targets further gradual fiscal consolidation and reductions in public debt. This strategy is based on improvements in tax administration and other efforts to broaden the tax base, with a reorientation of spending toward development of infrastructure and a phase out of energy subsidies.

Figure 3.1General Government Gross Debt

Sources: IMF, World Economic Outlook database; and IMF, 2010.

Maintaining low levels of public debt is a prudent strategy—the 2007–09 global economic and financial crisis has shown that even emerging economies with relatively low debt levels, like Indonesia, are still exposed to shocks and have low debt tolerance. 1

  • Maintaining a relatively low debt level can help manage risks from increased capital flow volatility. 2 For example, an increase in global risk aversion can trigger a sudden reversal of capital flows (as happened in late 2008), possibly dampening growth, and leading to spikes in borrowing costs and high exchange rate volatility. Recent econometric evidence suggests that higher debt levels in advanced countries are likely to be accompanied by higher long-term real interest rates, which could adversely affect emerging markets’ financing conditions (see IMF, 2010). 3

  • Maintaining low debt levels can also help manage risks from volatile oil and gas revenues. Fluctuating between 35 percent and 20 percent of total revenues, oil and gas revenue remains a significant, but volatile, source of income in Indonesia (Figure 3.2): the standard deviation of the real annual growth rate was more than five times the average annual real growth rate over 2000–09, well above oil price volatility. Recent turmoil in commodity markets has highlighted the relatively high uncertainty and volatility surrounding crude oil prices and oil-related revenues. But this high degree of uncertainty also affects relative prices and therefore can act like a tax on investment, leading to lower investment and growth (see Van Wijnbergen, 1985). This is particularly important in countries with relatively underdeveloped financial sectors, where risk sharing and obtaining bank financing during periods of illiquidity may be difficult to arrange. 4

  • Maintaining low debt levels can help lower the risks caused by the provision of sizeable fuel subsidies, which are highly volatile and procyclically linked to oil prices. Fuel price subsidies—comprising a third of total government spending on average over the last decade—lower fiscal policy effectiveness because they are inefficient, inequitable, environmentally unfriendly, and can crowd out other productive spending. 5 The volatility of spending on fuel subsidies is also far greater than oil price volatility alone. The standard deviation of the annual real growth rate of fuel subsidy spending was more than four times the annual average real growth rate during 2000–09. Most recently, spending on subsidies is projected to increase by 74 percent in 2011, more than twice the increase in the price of crude oil for the same year (which increased 36 percent in current dollar terms). Furthermore, maintaining fuel subsidies also implies that spending will increase automatically during oil price booms; for example, maintaining subsidies despite significant price increases in 2011 resulted in fiscal impulse that was twice as large as the non-oil fiscal impulse, net of oil revenues and subsidies.

Figure 3.2Oil and Gas Revenue Volatility

Sources: Indonesia, Ministry of Finance (various years); CEIC Data Co., Ltd.; and IMF staff estimates.

Assessing risks from these types of shocks is important when designing a medium-term fiscal strategy aiming at maintaining sustainability and managing risks. Increased volatility in macroeconomic variables has the potential to increase uncertainty about projected public debt dynamics. In addition, the experience of many oil-exporting countries shows that high dependence on volatile natural resource revenues can lead to debt problems if markets become inaccessible in downturns. Fiscal policy can help smooth rather than exacerbate oil price and revenue volatility, thereby minimizing its potential negative impact on growth. This chapter shows that anchoring fiscal policy in a medium-term fiscal framework to ensure consistency of annual targets with fiscal sustainability is important, but needs to be supported by fuel subsidy reform to lower fiscal vulnerability.


Risks from potential shocks were analyzed using a framework for assessing fiscal sustainability under uncertainty. The framework also incorporates oil and gas revenue volatility. This framework uses a stochastic simulation approach that derives the probability distribution of future debt stocks based on stochastic simulations of key risk variables (Box 3.1). The framework can also be expanded to include an endogenous fiscal policy reaction rule, whereby there is a partial adjustment of the primary balance to deviations from the baseline. 6

This framework can be used to assess the impact of oil and gas revenue volatility on the fiscal accounts. The model uses simulation methods to forecast the distribution and evolution of (net) public debt to assets explicitly accounting for oil and gas revenue volatility and expenditure policy. Projections of the oil and gas revenue stochastic profile can, in turn, be critical in formulating spending plans using oil income. For example, fiscal policy in countries with limited proven oil reserves (e.g., Mexico) should be very different from the fiscal strategy in countries with vast oil and gas reserves (e.g., the Russian Federation and Kazakhstan) where price volatility is a more important challenge.


This section uses the stochastic simulation approach to assess risks surrounding public debt projections under the authorities’ medium-term fiscal strategy. The analysis is undertaken in various stages, with the aim of assessing the sustainability of Indonesia’s public finance outlook to shocks in oil revenues and fuel subsidies. The first subsection checks the short- to medium-term robustness of the baseline fiscal strategy to oil price risks and other macroeconomic risks. The second subsection expands the analysis by checking the robustness of this strategy to oil and gas production shocks in a longer-term horizon and shows how an endogenous fiscal policy reaction rule can lower the uncertainty surrounding baseline debt projections. The last subsection illustrates the risks associated with delaying fuel subsidy reform.

Box 3.1The Framework

The first step in such an approach is to create a baseline scenario of the likely future time path of the public debt, using the flow budget constraint equation. This equation updates future debt as a share of GDP based on macroeconomic projections of key determinants of public debt dynamics: (a) non-oil primary deficit (net of seigniorage); (b) oil and gas fiscal revenues, which involve projections of the oil and gas extraction profile, prices, and taxation regimes; (c) growth-adjusted real interest payments on public debt; (d) capital gains or losses on net external debt caused by changes in the real exchange rate; and (e) other factors that can lead to debt accumulation. The framework is expressed as

where d is the (net) public-debt-to-GDP ratio, nopd is the non-oil primary deficit as a share of GDP (net of revenue from seigniorage), g is the real GDP growth rate, r is the real interest rate on public debt, ê is the change in the bilateral real exchange rate (local currency unit per US$1) where ê > 0 denotes a real exchange rate depreciation, and Roil denotes oil and gas fiscal revenues. Other factors (OF) could include off-budget liabilities leading to debt increases—for example, implicit contingent liabilities (bank bailouts) and called guarantees.

The framework incorporates two different approaches to analyzing uncertainty. To deal with vulnerability to specific shocks and assess robustness to extreme events, the framework provides a variety of stress tests (IMF, 2003). To get a broader view of the riskiness of the basic projections, the framework incorporates stochastic simulation methods, using empirical information about the distribution of the input variables (Burnside, 2005; Budina and van Wijnbergen, 2008; and IMF, 2008).

The stochastic simulation approach to fiscal sustainability involves simulating the entire distribution of future debt stocks, based on stochastic realizations of key debt determinants (real growth rate, real interest rate, real exchange rate), and accounting for their variances and covariance structure. Using estimated parameters of the joint distribution of debt determinants, the distribution of these variables can be simulated jointly using Monte Carlo methods. This implies that for n variables and a horizon of T years, n X T random numbers are generated repeatedly until the generated and empirical distribution are sufficiently close (by default 5,000 runs are generated). For each run, the model is applied to derive the full path of debt stocks and transform the generated random numbers in such a way that the resulting distribution conforms to the value at risk (VAR) estimates of the true distribution of the input variables. The default uses multivariate normal, but other distributions can be incorporated, too. The probability density of the outcomes of the debt ratio in each year can be plotted from the stochastic simulations, generating a so-called fan chart for the debt-to-GDP ratio.

Sources: Celasun, Debrun, and Ostry (2006); Bandiera, Budina, and van Wijnbergen (2008); and Budina and van Wijnbergen (2008).

Assessing Fiscal Sustainability and Medium-Term Risks

The robustness of the baseline fiscal strategy to shocks from a medium-term perspective was assessed. The baseline fiscal strategy is consistent with the authorities’ medium-term framework and assumes gradual fiscal consolidation, supported by revenue administration reforms and the elimination of fuel subsidies.

The evolution of debt was forecast using the identity equation (Equation 3.1.1, Box 3.1) that relates debt in year t to debt in the previous year, the non-oil primary balance and the projected oil and gas fiscal revenues in year t, and other stock-flow adjustments in year t and existing macroeconomic projections summarized in Table 3.1. The non-oil primary balance declines in line with oil revenue, implying that the overall primary fiscal position will be in balance. The exchange rate is projected to be broadly constant, thus not contributing substantially to the change in the public sector debt ratio. The decline in the public debt ratio is driven mostly by the favorable interest-growth differential.

TABLE 3.1Key Macroeconomic Indicators, 2009–16
Real GDP growth rate (in percent)
Exchange rate, period average (rupiah/US$)10,4079,088n.a.n.a.n.a.n.a.n.a.n.a.
Average inflation rate (percent)
Indonesia crude oil price (US$/barrel)6179107108105104105106
Oil production (thousands of barrels/day)9609549499901,0101,0101,0101,010
Nominal GDP

(trillions of rupiah)
Sources: IMF, World Economic Outlook database; and IMF staff estimates.
Sources: IMF, World Economic Outlook database; and IMF staff estimates.

The medium-term fiscal strategy should strike a balance between further public debt consolidation and the need to reorient fiscal priorities to support growth. Given low and declining public debt (below 27 percent of GDP in 2010) maintaining an overall fiscal deficit of about 1.5 percent of GDP while implementing fuel subsidy reform will create fiscal space for additional capital spending on needed infrastructure areas (e.g., transport, ports, and water systems), and ensure fiscal sustainability. This strategy—based on subsidy reduction, tax administration reforms, and continued strong economic growth—will support a further decline in the public debt to 21 percent of GDP by 2016. Such a strategy will ensure that a nearly balanced primary fiscal position is maintained, while accommodating extra resources for development spending (Figure 3.3). Moreover, to the extent that high-rate-of-return capital and infrastructure projects can be designed and implemented, even higher capital spending could be accommodated, provided that absorptive capacity improves significantly.

Figure 3.3Medium-Term Fiscal Strategy: Revenue, Spending, and Non-Oil Primary Deficit

Sources: Indonesia, Ministry of Finance (various years); and IMF staff estimates.

Such a fiscal strategy is sustainable and robust to macroeconomic and oil price shocks. The public debt ratio is likely to fall further, to about 25 percent of GDP in 2011, because rupiah appreciation and strong economic growth will more than offset the impact of the small primary deficit. A further decline in the public debt ratio to 21 percent of GDP by 2016 is likely, assuming a balanced primary fiscal position is maintained, supported by structural reforms to enhance economic growth and reduce fuel subsidies. Overall, risks from macroeconomic and oil price shocks under such a strategy appear limited. The framework runs Monte Carlo simulations using historical variances of five variables (changes in the real exchange rate, real borrowing costs for external debt and for domestic debt, growth rate, and the price of oil). The simulations indicate that the maximum likely debt ratio will be less than 30 percent of GDP throughout the projection period (Figure 3.4). An important part of this scenario is the assumption that oil and gas production levels remain stable and fuel prices increase gradually to curtail fuel subsidies.

Figure 3.4Fan Chart of Public-Debt-to-GDP Ratio

Source: IMF staff estimates.

However, failure to implement domestic fuel price adjustments could put pressure on fiscal deficits. Maintaining administrative fuel prices at their 2011 levels, given continuous strong energy demand growth, is estimated to keep subsidies at close to 3 percent of GDP per year over the medium term. Assuming that other spending remains as in the baseline scenario, and that oil production stagnates at its 2011 level, the overall deficit is estimated to increase gradually from 1.5 percent to more than 3 percent of GDP by 2016 (Figures 3.5 and 3.6). The fiscal impact could be even larger in the context of further price increases, faster energy demand growth, and lower oil production. Partial estimates show that for every $10 per barrel increase in the crude oil price, the 2011 deficit will increase by 0.2 percent of GDP, factoring in lower oil production estimates than budget estimates, rapidly growing demand for subsidized fuel products, and the decision in 2011 to postpone fuel subsidy reform. 7 Nevertheless, the impact of postponing subsidy reform on public debt over the medium term is still manageable given the relatively strong initial debt position and the favorable automatic debt dynamics suggested by strong economic growth and favorable interest rates. In this case, public debt would remain at 25 percent, while the maximum likely debt ratio increases to more than 30 percent (Figure 3.7). However, higher deficits may increase short-term fiscal vulnerability to the extent that gross financing needs will increase, increasing the vulnerability to changing market sentiments. Postponing subsidy reform would have a much more significant impact on fiscal sustainability in the longer term, when fuel consumption outpaces oil and gas production significantly (see next sections).

Figure 3.5Spending on Energy Subsidies

Source: IMF staff estimates.

Figure 3.6Overall Deficit

Source: IMF staff estimates.

Figure 3.7Fan Chart of Public-Debt-to-GDP Ratio without Subsidy Reform

Source: IMF staff estimates.

Assessing Fiscal Risks from an Oil or Gas Production Shock

This section extends the projection period to assess fiscal risks from an oil or gas production shock (e.g., running out of oil reserves), in addition to other stochastic shocks explored in the previous section. The analysis also shows the way in which an endogenous fiscal policy reaction rule could be used to manage uncertainty surrounding baseline debt projections.

Oil and gas wealth in Indonesia is significant, but much uncertainty surrounds future oil and gas production. Oil and gas production capacity expanded greatly during 1970–2000. However, oil production has declined rapidly since 2000 although gas production remained relatively constant. The decline in oil production could be related to insufficient investment in the sector (Figure 3.8) (Agustina and others, 2008). Proven oil reserves have declined sharply since the 1980s—from 11 billion barrels to 3.7 billion barrels as of end-2008—but have gone up recently to 4.4 billion barrels as a result of new oil discoveries. This level of reserves can sustain slightly more than a decade of production at current levels. Possible reserves could add another decade of current production levels. Proven gas reserves (3.18 billion cubic meters) can sustain 45 years of current gas production, while possible reserves could lengthen the gas production period further (Figure 3.9). 8 However, at current production levels, fiscal revenue from gas is a relatively small fraction of total oil and gas revenue, which means that when oil reserves are exhausted, a large expansion in gas capacity may be needed to maintain similar levels of oil and gas revenues. Although substantial expansion of gas capacity is possible, significant investment requirements and long lead times create considerable uncertainty about the gas extraction profile.

Figure 3.8Oil and Gas Production

Source: BP p.l.c., 2009.

Figure 3.9Proven Oil and Gas Reserves

Source: BP p.l.c., 2009.

The uncertainty surrounding the future oil and gas production profile adds to the challenge of managing oil and gas revenue volatility. This scenario checks the robustness of the baseline fiscal strategy to a large oil production shock resulting from oil reserves depletion that is not compensated for by higher gas production. Oil and gas price projections are the same as in the baseline, assuming constant real oil prices beyond 2015 (Figure 3.10). Based on these assumptions, oil and gas fiscal revenues drop sharply in 2021 as a result of the oil production shock. (Assuming no new oil discoveries are made, oil reserves are projected to be depleted by 2020.) The simulations are built around the baseline strategy for 2010–15 but assume full elimination of energy subsidies and progress toward revenue administration reforms. These reforms will support a gradual adjustment of the ratio of the non-oil primary deficit to GDP (Figure 3.11), in line with declining oil and gas revenues.

Figure 3.10Oil and Gas Price Trends

Sources: Ministry of Finance (various years); IMF, World Economic Outlook database; and IMF staff estimates.

Figure 3.11Projected Oil and Gas Fiscal Revenue and Primary Balance

Sources: Ministry of Finance (various years); and IMF staff estimates.

The baseline fiscal strategy is likely to be sustainable even after such a large shock to oil and gas production. Continuous adjustment in the non-oil primary deficit caused by a gradual reduction of subsidies and revenue administration efforts will prevent sizeable accumulation of public debt with the maximum likely debt ratio at 30 percent of GDP (Figure 3.12).

Figure 3.12Fan Chart for Public-Debt-to-GDP Ratio—The Impact of Fiscal Reaction Rules

Source: IMF staff estimates.

Note: Alpha represents degree of fiscal reaction.

Using an endogenous fiscal reaction rule can help manage uncertainty around the baseline debt projections by lowering the maximum likely debt ratio. The impact of exogenous shocks will be smaller if the government can commit to taking deliberate corrective actions as its debt stock rises. An endogenous fiscal policy reaction rule adjusting the primary balance to deviations from the baseline debt level could be used to lower risks. An application of the stochastic analysis with such an endogenous fiscal reaction rule narrows the confidence interval around baseline debt projections—the maximum likely debt ratio drops to about 25 percent of GDP (compared with about 30 percent without such a rule).

Assessing Fiscal Risks of Delaying Fuel Subsidy Reform

In addition to production uncertainties as discussed above, this section considers risks from delaying fuel subsidy reform in the context of increasing demand for fuel products. 9 This scenario checks the robustness of a fiscal strategy that is similar to the baseline, but in the absence of fuel subsidy reform. Subsidies are modeled as a product of a tax-inclusive fuel price gap and fuel consumption. Furthermore, the demand for subsidized fuel is assumed to grow faster than real income, outpacing oil and gas production in 2019. 10 Under these assumptions, fuel subsidies are likely to surpass the revenue from oil and gas in 2019. The non-oil primary deficit including subsidies is estimated to surpass 4 percent of GDP, while oil and gas revenue is projected to decline from 4 percent to about 1 percent (relative to GDP) over the projection period (Figure 3.13). This implies that beyond 2011, primary deficits will steadily increase, with a big jump in 2020 and beyond if oil reserves run out.

Figure 3.13Fiscal Strategy Without Subsidy Reform

Source: IMF staff estimates.

Not surprisingly, simulations reveal that such an alternative scenario would yield an increasing accumulation of debt. Rising primary deficits lead to increasing debt accumulation in every year after 2014, with public debt reaching 47 percent of GDP in 2023. Risks to public debt are large—the maximum possible debt ratio reaches around 70 percent of GDP within the 95 percent confidence interval (Figure 3.14). Thus, a delay in fuel subsidy reform, in combination with shocks to oil and gas revenue, can increase vulnerabilities and risks.

Figure 3.14Fan Chart for Public-Debt-to-GDP Ratio without Subsidy Reform

Source: IMF staff estimates.


Indonesia’s fiscal and debt outlook has been remarkably resilient to the recent global financial crisis. Public finance improvements, combined with relatively modest fiscal stimulus and strong economic growth, have supported a continuous decline in public debt and lowered the economy’s dependence on short-term foreign financing. The benefits of this situation during the global financial turmoil of 2008–10 are obvious. Nevertheless, risks to the public debt outlook still exist, and managing these risks could enhance policy credibility further.

This analysis used a stochastic simulation approach to assess risks to the public debt outlook. Specifically, this framework derives the probability distribution of future debt stocks based on stochastic simulations of key risk variables. An endogenous fiscal policy reaction function, which adjusts the primary balance to deviations from baseline level debt stocks, can also be used with stochastic simulations. Finally, the impact of oil and gas revenue volatility and the risks from delaying subsidy reform are also assessed.

The authorities’ strategy of gradual fiscal consolidation is sustainable and likely to reduce public debt further in the medium term. Such a strategy implies a gradual adjustment of the non-oil primary deficit in line with the declining ratio of oil and gas revenue to GDP. This adjustment strategy is supported by revenue enhancements and fuel subsidy reform.

Risk assessment revealed that

  • Risks to the debt outlook from macroeconomic and oil price shocks appear manageable in the medium term because the maximum likely debt level under stochastic shocks is relatively moderate. The fiscal strategy is robust to macroeconomic and oil price shocks, if supported by revenue administration reform and fuel subsidy reform.

  • Over the longer term, lack of investment could result in stagnating gas production, declining oil production, and an associated revenue drop. Fiscal risks are still manageable in this scenario but uncertainties are higher. Having an endogenous fiscal policy reaction would lower risks.

  • These robust outcomes are only possible when the fiscal strategy is supported by fuel subsidy reform. Fiscal risks from delaying such reform create fiscal vulnerability in the context of rapid growth in fuel consumption, combined with a negative oil or gas revenue shock from stagnating or even declining oil or gas production.


Reinhart, Rogoff, and Savastano (2003) argued that “safe” external debt levels for emerging-market countries with histories of default and inflation are relatively low (as low as 15 percent of GNP in some countries).

Reinhart and Rogoff (2008) used a new data set to illustrate vulnerabilities of emerging markets, in particular, the fact that sovereign debt defaults are sensitive to global capital flow cycles.

High debt ratios could increase long-term real interest rates by almost 2 percentage points, negatively affecting emerging markets’ financing conditions. In addition, high debt ratios are also likely to negatively affect potential growth in advanced economies, with possible consequences for emerging markets.

Recent literature highlights the potentially negative impact of natural resource revenue volatility on growth in countries highly dependent on natural resources (van der Ploeg and Poelhekke, 2009). Aghion and others (2006) have shown empirically that high volatility slows down productivity growth by a substantial margin in countries with relatively underdeveloped financial sectors.

For a detailed discussion and estimation of such a reaction function for the United States, see Bohn (1998) and Celasun, Debrun, and Ostry (2006) for a panel of emerging markets.

The impact of an oil price increase on the budget requires estimating the sensitivity of all related revenues and expenditures. See IMF (2007) for details on these sensitivity estimates.

See BP p.l.c. (2009). Proven and possible gas reserves may be higher (see Embassy of the United States of America, 2008).

For estimates of fuel subsidies around the world, see Coady and others (2010).

Fuel consumption is benchmarked to grow faster than real income, in line with the relatively high income elasticity estimate found in Agustina and others (2008).

    Other Resources Citing This Publication