Helping Countries Develop
Chapter

6 A Framework for Fiscal Debt Sustainability Analysis in Low-Income Countries

Author(s):
Benedict Clements, Sanjeev Gupta, and Gabriela Inchauste
Published Date:
September 2004
Share
  • ShareShare
Show Summary Details
Author(s)
Emanuele Baldacci and Kevin Fletcher 

Assessing fiscal debt sustainability is an important element of fiscal policy analysis. It aids in evaluating the appropriateness of a country’s macroeconomic policies in both the short and long run by helping to determine whether current and future fiscal policies are sustainable or whether a period of fiscal adjustment is needed.

To perform debt sustainability analysis, it is first necessary to define debt sustainability. While debt sustainability can be defined in several ways, this paper considers a country’s debt to be sustainable if the country can meet its debt-service obligations without recourse to debt relief, rescheduling of debt, or the accumulation of arrears and without unduly compromising growth.1 While the last element of this definition is admittedly somewhat vague, it is meant to capture the notion that fiscal sustainability cannot be predicated on assumed future fiscal adjustments and growth that are so large as to be unrealistic. In other words, fiscal debt sustainability requires that governments be not just able, but also politically willing, to repay their debt.

A variety of methodologies can be used to assess fiscal debt sustainability. In countries with significant access to private capital markets, market signals (such as interest rate spreads) can be useful indicators of the probability of debt default. However, such market signals are typically absent in low-income countries (at least for external debt, which may be almost entirely noncommercial). In these countries, debt sustainability must be assessed more directly by examining whether a government has sufficient resources to meet its debt-service obligations. In theory, this can be determined by calculating whether a government’s debt (D) is less than or equal to the present value of the primary surpluses (P) that will be used to pay the debt:

where t denotes the time period and i is the nominal interest rate.2 However, such calculations are very sensitive to the assumed long-run interest rate and primary balance, which are typically not known with much certainty. Moreover, such a simple budget constraint approach may not adequately account for factors that make debt unsustainable even when the government is technically solvent. For example, debt crises could also be triggered by liquidity constraints, constraints on the degree to which domestic revenue can be converted into foreign currency for debt service, and political constraints on the share of revenue that the public is willing to use for debt service. Thus, a more holistic approach to debt sustainability analysis requires that debt dynamics be assessed under a variety of scenarios to allow for variability in the underlying assumptions and incorporate a variety of indicators to allow for the different types of constraints that could lead to debt crises.

This paper presents one such framework for debt sustainability analysis that is meant to be both pragmatic and accessible to a wide range of policy analysts.3 The framework has the following key features: (1) it allows analysts to break down the evolution of debt into various factors (for example, effects of primary balances, effects of GDP growth); (2) it provides a simple mechanism for clarifying the key underlying assumptions in a baseline scenario and for assessing their realism in light of historical experience; (3) it suggests the use of various indicators to examine the multiple constraints on debt sustainability; and (4) it suggests several “stress tests” to debt sustainability that are aimed to capture the major risks facing most low-income countries.

The framework is intended primarily to be a “toolkit” for analyzing the main factors and risks affecting debt sustainability and is not a complete guide to medium-term macroeconomic programming. The framework provides a check on whether the underlying policies assumed in a baseline scenario would lead to, or maintain, debt sustainability and on how robust they are to shocks. It also provides the means to test the realism of the base case projections as well as whether alternative policies may be more appropriate. The results of this framework should be supplemented by additional work that identifies the links between government policies, debt dynamics, and economic growth. Thus, analysts should take care to try to specify these links as accurately as possible, drawing on both the general economic growth literature and country-specific factors, and to carefully consider both the policy choices that are implicit in their baseline scenarios and the policy implications that should be derived from it.

The Framework

This section describes in detail the framework’s specific features, including the definition of public debt, the time frame of the debt sustainability analysis, the key indicators used in the framework, the assessment of the macroeconomic assumptions, and the stress tests used.

Time Frame

To conduct debt sustainability analysis, it is first necessary to determine the time frame over which debt dynamics will be examined. In this regard, one must consider whether a country is in a “steady state” in regard to its fiscal policies or whether major changes in the fiscal policy environment are expected in the future. In the former case, a relatively short time frame could be used on the assumption that current trends will continue more or less indefinitely. The latter assumption, however, requires that debt sustainability analysis be carried out long enough to capture the effects of major changes, such as the effects of large demographic shifts on pension systems.4

In the case of low-income countries, an important future shift in the fiscal environment that may need to be considered is a country’s future “graduation” from highly concessional assistance. Currently, many low-income countries finance their deficits primarily through borrowing from official lenders on concessional terms, meaning that the interest rates are low (1-2 percent) while the maturities are long (10-40 years), with long grace periods (5-10 years) during which no principal payments are due. Because of this relatively cheap source of financing, many low-income countries may currently be able to maintain their current debt levels despite running significant deficits.5 In the long run, however, one of the goals of concessional financing is for the recipient countries to substantially reduce their reliance on external official flows. Accordingly, it is generally envisaged that low-income countries will eventually graduate from concessional assistance, implying that the rate of interest they pay on their debt will gradually approach market rates as access to concessional financing is phased out.

For most low-income countries, graduation from concessional borrowing is expected to occur only in the long run. In the process, these countries will have to shift their primary balances from deficits in the near and medium terms to significant surpluses in the long run in order to remain solvent while repaying maturing concessional loans and financing any current overall deficits. Moreover, the necessary adjustments in the primary balance will have to reflect any loss in grants that these countries might suffer, placing greater pressure to increase their own revenue and decrease expenditure. To reflect these circumstances, the time frame of the analysis should, if possible, extend to the point of the country’s expected graduation from concessional assistance.6 In most low-income countries, it is envisaged that this will be at least 20 years.

On the other hand, there are costs to using an excessively long time frame for debt sustainability analysis. In particular, projections very far into the future are typically highly sensitive to very small changes in long-run assumptions, which typically are highly uncertain. As a result, a very long timeframe risks diverting attention away from near-term policy needs and instead allows debt sustainability to be predicated on highly uncertain long-run macroeconomic assumptions or on changes in fiscal policy in the distant future that current policymakers cannot credibly commit to. In addition, long time frames increase the computational costs. Thus, even if graduation is not expected after 20 years, analysts may wish to cap the exercise at this point to keep it manageable. In these cases, the analysis will at least indicate how far the country may be from graduation in 20 years.

Defining Public Debt

Another important preliminary issue in fiscal debt sustainability analysis is determining how public debt should be defined. In this regard, there are choices to be made regarding coverage (which could range from a narrow focus on central government to a very broad focus on total public sector), as well as type of debt (gross versus net).

In our framework, we suggest that, in most cases, debt statistics should be compiled for the entire public sector as well as its major subcomponents. The definition of public debt should in principle include both domestic and external debt of the central government, regional or state governments, and local governments. The liabilities of extrabudgetary funds, such as social security funds, and public enterprises, such as publicly owned airlines and utilities, should, if possible, be included in the definition of public debt in most cases (see caveats below). However, it is recognized that in many cases limits on data availability may allow coverage only for a smaller subsector, such as the central government.

In principle, it is useful to focus most public debt sustainability analyses on net debt concepts (liabilities minus the value of assets), since it is important to look at not just the debt a government owes but also the assets that a government has to repay its debt. In practice, however, the value of government assets is difficult to establish in many low-income countries. As a result, rough estimates of net debt may have to be made in many cases. In this regard, it is worth noting that the assets of the general government in most countries are either small or not readily sellable in the market (e.g., a public monument). Thus, in these cases, gross debt may be an acceptable proxy for net debt for the general government sector. Similarly, gross debt may be a good proxy for net debt for public enterprises that are known to have weak financial positions, perhaps due to the quasi-fiscal activities that these enterprises often undertake. However, in cases where public enterprises or extrabudgetary funds (such as social security institutions) are known to have substantial assets, it may often be necessary to take account of these assets in the sustainability analysis. For example, if an enterprise is known to have substantial assets (such as a financial institution) or to operate on a sound commercial basis, analysts might decide to estimate the net debt of these enterprises to be zero, effectively excluding them from the analysis.

Another exception to using broad coverage may be cases in which subnational governments do not contract debt and their revenues cannot be used (either directly or indirectly) to pay central government debt. In these cases, it may be sensible to exclude these subnational governments from both measures of debt and measures of public sector balances, so that, for example, debt-to-revenue ratios are more meaningful.

In any event, care should be taken to ensure that the definition of the public sector is consistent across stocks and flows, so that changes in the flows, such as a larger primary balance, can be reconciled with changes in the stock of debt. To allow reconciliation between stocks and flows, contingent liabilities of the public sector should be excluded from debt stocks until the liabilities are officially recognized. However, if the value of contingent liabilities is known, these can be shown in output tables of the debt sustainability analysis as a memorandum item, as recommended by the Government Finance Statistics Manual 2001 (IMF, 2001). Such contingent liabilities may include government loan guarantees, the expected costs of bank recapitalization, or unfunded pension liabilities. In such cases, the contingent liabilities should be valued at the option price that a financial institution would charge to take responsibility for the government’s potential liability. For example, if the coverage of public sector stocks and flows extends only to the general government, then general government guarantees on debt contracted by public enterprises should be included in the memo items as a contingent liability, valued at the present value of the expected cost of the government’s liability.7 On the other hand, if the coverage of public sector stocks and flows includes public enterprises, then the gross value of such debt will already be included in the public sector debt stock and need not be listed as a contingent liability.

Key Indicators

A third key issue in debt sustainability analysis is to determine which indicators one will use to assess debt sustainability. In this regard, we suggest using multiple indicators to capture the various economic and political constraints on debt sustainability, as discussed in this chapter’s introduction. In particular, the following indicators may be especially useful in the low-income country context.

Nominal level of debt as a share of GDP: This is one of the most commonly quoted debt sustainability statistics, since it compares the level of indebtedness with a measure of a country’s capacity to pay (GDP) in a manner that is easily understood and straightforward to calculate. However, its simplicity also entails at least two important drawbacks: (1) it ignores the terms on which debt is denominated, which can have a major impact on sustainability; and (2) it ignores the pattern of debt-service payments, which can create debt-service problems owing to liquidity or political constraints, even at low debt-to-GDP ratios.

Net present value (NPV) of debt as a share of GDP: The NPV of debt (debt service discounted at the market interest rate) can be another useful summary indicator of indebtedness since, unlike the nominal value of debt, it takes account of the terms on which the debt is denominated—that is, low-interest-rate debt will have a lower NPV while high-interest-rate debt will have a higher NPV. This is especially important in low-income countries, where much of the debt is concessional. As a result, comparisons of these countries’ nominal debt ratios to those in countries with low levels of concessional financing may be misleading.8 In the case of concessional financing, the total amount of debt service is lower and farther in the future, when a country’s capacity to pay will be higher relative to current GDP, if this variable’s growth is positive. Thus, to reflect these effects of concessional lending, a time series for the NPV of debt ratios can be useful.

The NPV concept may be most useful in low-income countries for external debt, since a large portion of external debt is often concessional. Ideally, consistent methods should be used for calculating the NPV of domestic debt by, for example, transforming future domestic debt-service payments into U.S. dollars using exchange rate projections and then using the same U.S. dollar discount rate for domestic debt as for external debt. However, the NPV of domestic debt may not differ significantly from the nominal value since most domestic debt is on market terms and is often very short-term debt. Moreover, in many cases, analysts may have insufficient detail on domestic debt-service payments to calculate the NPV. Thus, it may be necessary to make the simplifying assumption that the NPV of domestic debt equals the nominal value (so that the NPV of total fiscal debt is the NPV of external debt plus the nominal value of domestic debt). In the following discussion, it is assumed that this approach is taken.

Foreign currency-denominated public debt as a share of GDP: Foreign currency debt as a share of GDP can be another useful indicator of fiscal debt sustainability, since countries with a high degree of foreign currency debt may encounter problems generating a sufficient amount of foreign exchange to service this debt, even when total indebtedness is not excessively high.9

Gross financing needs as a share of GDP: Gross financing need is defined as the public sector deficit plus amortization of medium- and long-term debt plus the stock of short-term debt at the end of the previous period.10 This statistic is important because it indicates whether liquidity constraints may become problematic.

Debt-to-revenue ratio: The debt-to-revenue ratio is useful because it gives a further indication of a country’s ability to finance its debt and, unlike the debt-to-GDP ratio, takes account of constraints that may exist in converting national income into public revenue.11 In addition, this ratio is likely to be less sensitive to the definition of the public sector than the debt-to-GDP ratio.12 As a result, the debt-to-revenue ratio may be more useful for cross-country comparisons. One issue in calculating the debt-to-revenue ratio is whether one should include grants in revenue, and the answer to this question may depend largely on the purpose of the analysis. If the goal is to determine the ability of a country to pay its debt obligations given its current and projected level of grants, then it seems reasonable to include grants in revenue; but if the goal is to assess whether a country is able to graduate from external assistance and still service its debts, then it may be reasonable to exclude grants from revenue.

Ratio of debt service to revenue: Debt service is typically defined as interest payments plus amortization of medium- and long-term debt. Debt service as a ratio to revenue is another useful indicator of potential liquidity problems, as well as the degree to which debt may become “politically unsustainable,” in the sense that high levels of debt service relative to other types of spending may create political pressure to renege on debt payments.

Primary deficit that stabilizes the debt-to-GDP ratio: The primary deficit that stabilizes the debt-to-GDP (or NPV of debt) ratio is useful as a measure of the amount of adjustment that is needed to stabilize the debt dynamics. In our framework, we define the debt-stabilizing primary balance as the sum of the actual primary balance (as a share of GDP) and the actual change in the debt ratio in each period of the baseline scenario. Note that, in this case, this estimated primary deficit only stabilizes the debt ratio in the period in question, assuming that all previous periods followed the path of the baseline scenario.

Decomposing Fiscal Debt Dynamics

To better analyze debt dynamics, it is also useful to decompose the evolution of the debt-to-GDP ratio in order to determine the main factors driving the evolution of the debt ratio. In this regard, the change in the debt-to-GDP ratio between any two periods can be described by the following equation:

where dt is the nominal value of public debt at time t expressed as a ratio to GDP, a is the share of debt held in foreign currency, rf is the average real interest rate on foreign currency debt,13rd is the average real interest rate on domestic debt, g is the real GDP growth rate, ε is the per-centage change in the real exchange rate,14p is the primary balance as a share of GDP, and η represents other balance sheet changes that affect debt as a share of GDP This equation is derived in the Appendix.

Equation (2) can be used to calculate the change in the public debt ratio and to separate the different channels that contribute to its evolution: (1) the primary deficit effect (-p); (2) the real foreign-currency interest rate effect (αdrf/(1+g)); (3) the real domestic interest rate effect ((1-α)drd/(1+g)); (4) the real GDP growth effect (-gd/(1+g)); and (5) the real exchange rate effect (αdε(1+rf)/(1+g)). The separation of the different factors allows an assessment of their relative importance for the evolution of the debt ratio. It also serves as the basis for stress tests in which the key assumptions are varied (see the subsection on stress tests below). In addition to these terms, there are other factors (η) that may increase or decrease debt; for example, from recognition of contingent liabilities, debt relief operations, or privatization receipts. Gross debt can also change as a result of other below-the-line operations (such as repayment of debt financed by a reduction in financial assets) as well as cross-currency movements. Finally, factors that cannot be identified are counted as a residual that is equal to the difference between the actual change in debt and the identified factors. Large residuals might indicate problems in the data and/or projections.

The evolution of the NPV of debt can also be decomposed through an analogous equation:

where nt is the NPV of public debt at time t expressed as a ratio to GDP, rδ is the real discount rate on foreign currency debt,15 λ is the grant element of new debt (expressed as a percentage), and bf is new foreign currency borrowing as a share of GDP This equation is also derived in the Appendix.

This equation can be used to separate the different channels contributing to the evolution of the NPV: (1) the primary deficit effect (-p), (2) the real discount rate effect (αnrδ/(1 + g)), (3) the real domestic interest rate effect ((l-α)nrd/(1+g)), (4) the real GDP growth effect (-ng/(1+g)), (5) the real exchange rate effect (αnε(1+rδ)/(1+g)), (6) the impact of grants that are effectively given via the grant element of new borrowing (λbf) and (7) the other balance sheet changes, including a residual for unidentified factors (η).

Assessing Key Macroeconomic Assumptions

Projections of debt sustainability are typically highly sensitive to the underlying assumptions regarding the evolution of real GDP growth, interest rates, exchange rate appreciation, primary balances, and other macroeconomic variables. Thus, it is critical to analyze how realistic these assumptions are in the baseline scenario. A simple check that can be made on the assumptions is simply to compare them to their historic averages. In general, it is probably imprudent to assume that key variables will differ markedly from their historic averages, unless there are clear and specific reasons for doing so.

Stress Tests

Because of the importance of and uncertainty regarding key macroeconomic variables, it is also useful to perform several “stress tests” to assess the sensitivity of the baseline scenario to major departures from the assumed trends in macroeconomic variables.

In choosing the most appropriate stress tests, it is useful to first consider which factors typically have the largest effects on debt sustainability. Important factors that are often cited are real GDP growth,16 exchange rate depreciation, real interest rates, primary balances, and the recognition of previously contingent liabilities. Thus, the following are some suggested stress tests for low-income countries.

  • Real GDP growth and the primary balance in each period are equal to the historical averages over the past 10 years (or latest years available);17 if the results of this stress test differ markedly from the baseline scenario, this might indicate excessively optimistic or pessimistic assumptions in the baseline.

  • The primary balance remains at last year’s level for all future years; this might be a “no-reform” scenario.18

  • Real GDP growth in each of the first two periods is below the historical average by one standard deviation (standard deviations for key variables can be calculated using historical data); this simulates the effects of a severe recession.

  • The primary balance in each of the first two periods is below the historical average by one standard deviation; this simulates the effects of a significant policy slippage.

  • Both real GDP growth and the primary balance in each of the first two periods are below their historical averages by one-half a standard deviation.

  • The real GDP growth rate in each period is below the baseline scenario by one standard deviation of the average real GDP growth rate for the entire projection period.19

  • There is a permanent 30 percent depreciation in the first projection period.

  • There is a 10 percent of GDP increase in other debt-creating flows in the first projection period; this simulates the potential effects of having to assume contingent liabilities such as a bank recapitalization.

  • There is a permanent, one-standard-deviation shock to the most important commodity price in the first projection period.

Of course, this list is only meant to be suggestive; the number and type of stress tests should be tailored to the particular risks facing each country.20

Note that one easy method of performing these stress tests is to use the decomposition in equations (2) and (3) to estimate the effect on debt of changing a particular macroeconomic variable in each period. However, for shocks that have many pass-through effects (e.g., exchange rate depreciation may be associated with higher interest rates and also affect growth), more complicated modeling may be required. In general, the need for more complicated modeling will depend on how important the pass-through effects are judged to be in each case.

Note also that, to perform stress tests, it is necessary to make assumptions about the terms of marginal borrowing. In other words, if a change in macroeconomic assumptions results in more borrowing, then one must make assumptions about the terms on which this borrowing is contracted. Since these assumptions may have a substantial impact on the outcome of the stress tests, it is important that the terms of marginal borrowing be specified realistically. In particular, it may be that, while many low-income countries’ debt is, on average, highly concessional, these countries’ access to concessional financing at the margin may be more limited (especially in the event of policy slippages).

Bringing the Elements Together

The elements of this debt sustainability analysis framework—the key indicators of the baseline scenario, the decomposition of the evolution of debt, the highlighting of key macroeconomic assumptions, and the results of the stress tests—are all brought together in Tables 1 and 2, which show how one might present these elements for a particular country (the details of this particular case study are discussed in the next section). Table 1 shows how the decomposition of nominal debt evolution could be presented, while Table 2 shows the decomposition of the NPV of debt.

Table 1.Public Sector Debt Sustainability Framework, 2000-2023, Nominal Debt(Percent of GDP, unless otherwise indicated)
ActualsEstimateProjectionsProjections10-Year Standard Deviation10-Year Historical AverageProjected Average 2004-08Projected Average 2009-23
20002001200220032004200520062007201020152023
I. Baseline Projections
Public sector debt1222.2171.1171.9125.3113.6105.697.992.279.666.342.6
of which, foreign currency-denominated217.1150.2152.9105.997.891.885.580.871.160.940.0
Change in public sector debt39.3−51.10.9−46.6−11.7−8.1−7.7−5.8−3.5−3.7−3.2
Change in public sector debt39.3−51.10.9−46.6−11.7−8.1−7.7−5.8−3.5−3.7−3.2
Identified debt-creating flows28.5−45.32.4−45.5−10.3−7.1−5.7−5.8−4.3−3.7−2.8
Primary deficit2.84.71.60.5−1.5−1.0−0.4−0.90.00.30.03.9−1.1−0.90.1
Revenue and grants25.124.826.325.525.524.524.023.022.421.320.02.026.223.921.2
of which, grants5.75.78.37.17.46.35.94.84.33.21.93.73.45.83.1
Primary (noninterest) expenditure27.929.527.825.924.023.523.622.022.521.620.13.825.123.121.3
Automatic debt dynamics33.5−62.50.8−12.8−8.8−6.1−5.2−4.9−4.4−4.0−2.8
of which, contribution from
average real interest rate−1.1−3.0−0.8−0.20.60.1−0.2−0.4−0.5−0.7−0.6
of which, contribution from
real GDP growth−6.3−10.4−5.0−7.4−5.4−5.4−5.0−4.2−3.6−3.0−2.0
of which, contribution from
real exchange rate depreciation40.9−49.16.6−5.3−4.0−0.80.0−0.3−0.3−0.3−0.3
Other identified debt-creating flows−7.912.50.0−33.10.00.00.00.00.00.00.0
Privatization receipts (negative)0.00.00.00.00.00.00.00.00.00.00.0
Recognition of implicit or
contingent liabilities0.00.00.00.00.00.00.00.00.00.00.0
Debt relief (HIPC and other)−7.90.00.0−33.10.00.00.00.00.00.00.0
Bank recapitalization0.012.50.00.00.00.00.00.00.00.00.0
Residual, including asset changes10.9−5.9−1.5−1.1−1.4−0.9−2.00.10.80.0−0.4
Other sustainability indicators
NPV of public sector debt140.2115.0123.161.155.452.447.343.938.434.021.5
Ratio of NPV of public sector debt to revenue and grants (percent)558.8463.2468.2240.0217.1214.4197.3191.4171.1163.6114.8
Gross financing need7.08.06.48.610.98.98.37.05.02.71.3
Ratio of public sector debt to revenue and grants (percent)885.5689.0654.1492.1445.6431.7408.0401.4355.1310.8212.7
Ratio of debt service to revenue and grants (percent)12.110.115.532.031.425.222.421.114.89.86.5
Primary deficit that stabilizes the debt-to-GDP ratio−36.555.80.747.110.27.17.24.83.64.03.2
Key macroeconomic and fiscal assumptions
Real GDP growth (percent)3.64.93.04.54.55.05.04.54.54.54.55.60.64.74.5
Average real interest rate on domestic debt (percent)16.312.80.11.63.34.65.55.45.04.84.38.59.84.84.6
Average real interest rate on foreign exchange debt (percent)−1.1−1.8−0.6−0.3−0.1−0.7−1.1−1.2−1.3−1.5−1.60.6−1.1−0.9−1.5
Real exchange rate depreciation (percentage change, + = depreciation)24.0−24.14.6−3.6−3.9−0.90.1−0.4−0.5−0.5−0.7−1.1−0.5
Inflation rate (GDP deflator; percent)30.024.319.719.913.77.95.25.05.05.05.044.243.77.45.0
Growth of real primary spending (deflated by GDP deflator; percent)9.210.9−2.9−2.7−3.22.65.4−2.24.43.33.617.03.91.63.8
Major commodity price: copper $/lb0.80.70.70.80.80.90.90.90.90.90.90.20.90.90.9
II. Stress Tests for Public Debt Ratio
Alternative scenarios
A1. Real GDP growth and primary balance are at historical averages125.3114.0106.398.492.477.157.425.4
A2. Primary balance is unchanged from 2002 (no reform)125.3116.7111.1105.4102.194.586.671.1
A3. Long-run real GDP growth is at baseline minus one standard deviation125.3115.2108.8102.999.093.194.8102.4
Bounds tests
B1. Real GDP growth is at baseline minus one standard deviation in 2004-05125.3121.2121.2114.7110.4102.597.083.5
B2. Primary balance is at baseline minus one standard deviation in 2004-05125.3117.5113.3105.699.786.972.848.1
B3. Combination of Bland B2 using one-half-standard-deviation shocks125.3118.8115.5107.1100.886.771.244.9
B4. One-time 30 percent real depreciation in 2004125.3177.0165.1154.1145.1124.2100.265.2
B5. 10 percent of GDP increase in other debt-creating flows in 2004125.3123.6115.5107.7101.988.974.649.6
B6. A permanent, one-standard-deviation negative shock to the major commodity price in 2004125.3116.0109.2102.998.589.982.667.1
Sources: Country authorities; and IMF staff estimates and projections.

Central government; gross debt.

Sources: Country authorities; and IMF staff estimates and projections.

Central government; gross debt.

Table 2.Public Sector Debt Sustainability Framework, 2000-2023, NPV of Debt(Percent of GDP, unless otherwise indicated)
ActualsEstimateProjectionsProjections10-Year Standard Deviation10-Year Historical AverageProjected Average 2004-08Projected Average 2009-23
20002001200220032004200520062007201020152023
I. Baseline Projections
NPV of public sector debt1140.2115.0123.161.155.452.447.343.938.434.021.5
of which, foreign currency-denominated135.194.1104.041.739.538.734.932.629.828.619.0
Change in NPV−0.5−25.28.1−62.0−5.8−3.0−5.1−3.4−1.3−1.2−1.6
Identified debt-creating flows6.4−19.44.0−58.9−6.1−3.7−2.1−2.5−1.4−0.9−0.7
Primary deficit2.84.71.60.5−1.5−1.0−0.4−0.90.00.30.03.9−1.1−0.90.1
Revenue and grants25.124.826.325.525.524.524.023.022.421.320.02.026.223.921.2
of which, grants5.75.78.37.17.46.35.94.84.33.21.93.73.45.83.1
Primary (noninterest) expenditure27.929.527.825.924.023.523.622.022.521.620.13.825.123.121.3
Grant element of new concessional loans−4.7−3.1−3.0−2.4−2.8−2.1−1.4−1.3−1.2−0.9−0.5
Automatic debt dynamics33.7−33.65.4−4.3−1.8−0.7−0.3−0.3−0.3−0.3−0.3
of which, contribution from
real discount rate
on foreign-currency debt5.04.64.44.41.91.61.41.31.11.10.8
of which, contribution from
real interest rate on domestic debt0.70.60.00.30.60.70.70.60.50.30.1
of which, contribution from
real GDP growth−4.9−6.6−3.4−5.3−2.6−2.6−2.5−2.0−1.7−1.5−1.0
of which, contribution from
real exchange rate depreciation32.8−32.24.4−3.7−1.6−0.30.0−0.1−0.1−0.1−0.1
Other identified debt-creating flows−25.412.50.0−52.70.00.00.00.00.00.00.0
Privatization receipts (negative)0.00.00.00.00.00.00.00.00.00.00.0
Recognition of implicit or contingent
liabilities0.00.00.00.00.00.00.00.00.00.00.0
Debt relief (HIPC and other)−25.40.00.0−52.70.00.00.00.00.00.00.0
Bank recapitalization0.012.50.00.00.00.00.00.00.00.00.0
Residual, including asset changes−6.9−5.84.1−3.00.30.8−3.0−0.90.1−0.3−0.9
Other sustainability indicators
Nominal value of public sector debt222.2171.1171.9125.3113.6105.697.992.279.666.342.6
NPV of contingent liabilities
(not yet officially recognized
in public sector debt)
Ratio of NPV of public sector debt to revenue and grants (percent)558.8463.2468.2240.0217.1214.4197.3191.4171.1159.2107.5
Gross financing need7.08.06.48.610.98.98.37.05.02.71.3
Ratio of public sector debt to revenue and grants (percent)885.5689.0654.1492.1445.6431.7408.0401.4355.1310.8212.7
Ratio of debt service to revenue and grants (percent)12.110.115.532.031.425.222.421.114.89.86.5
Primary deficit that stabilizes the NPV of debt-to-GDP ratio3.430.0−6.562.44.32.04.62.51.41.51.7
1. Baseline Projections
Key macroeconomic and fiscal assumptions
Real GDP growth (in percent)3.64.93.04.54.55.05.04.54.54.54.55.60.64.74.5
Average nominal interest rate on foreign exchange debt (in percent)1.00.60.61.11.11.00.90.90.70.40.30.30.70.90.5
Average real interest rate on domestic debt (in percent)16.312.80.11.63.34.65.55.45.04.84.38.59.84.84.6
Real exchange rate depreciation (percentage change, + = depreciation)24.0−24.14.6−3.6−3.9−0.90.1−0.4−0.5−0.5−0.7−1.1−0.5
Inflation rate (GDP deflator, percent)30.024.319.719.913.77.95.25.05.05.05.044.243.77.45.0
Growth of real primary spending (deflated by GDP deflator, percent)9.210.9−2.9−2.7−3.22.65.4−2.24.43.33.617.03.91.63.8
Major commodity price: copper $/lb0.80.70.70.80.80.90.90.90.90.90.90.20.90.90.9
Grant element of new external borrowing (share of total external borrowing, percent)62.362.362.362.362.362.362.362.362.362.362.318.288.762.362.3
II. Stress Tests for NPV
Alternative scenarios
AI. Real GDP growth and primary balance
are at historical averages61.155.752.947.443.835.625.65.7
A2. Primary balance is unchanged
from 2002 (no reform)61.158.357.754.453.251.852.347.4
A3. Long-run real GDP growth is at baseline
minus one standard deviation61.156.354.350.448.247.755.971.5
Bounds tests
B1. Real GDP growth is at baseline minus one
standard deviation in 2004-200561.159.561.657.956.055.258.756.6
B2. Primary balance is at baseline minus one
standard deviation in 2004-200561.159.159.754.450.844.839.926.6
B3. Combination of B1 and B2 using
one-half-standard-deviation shocks61.158.859.253.449.542.736.922.7
B4. One-time 30 percent real depreciation in 200461.181.177.670.765.656.046.630.1
B5. 10 percent of GDP increase in other
debt-creating flows in 200461.165.061.756.352.746.641.527.9
B6. A permanent, one-standard-deviation
negative shock to the major
commodity price in 200461.157.255.551.749.547.548.543.6
Sources: Country authorities; and IMF staff estimates and projections.

Central government; gross debt.

Sources: Country authorities; and IMF staff estimates and projections.

Central government; gross debt.

Once one has projected the indicators and run the stress tests, the key remaining question is what do they imply for the prospects for debt sustainability? In this regard, a few econometric studies have tried to map indicators such as the debt-to-GDP ratio into probabilities of experiencing a debt crisis or a period of low growth.21 However, most of these studies either focus on emerging markets or on external debt; the analysis of total public sector debt indicators (external plus domestic debt) in low-income countries has been more problematic. To some degree this should not be surprising, given the wide differences across low-income countries in the quality and coverage of public sector data.

Thus, while acknowledging that the literature on external debt and emerging markets can provide some useful guideposts, our framework eschews a purely formulaic approach to mapping various public debt indicators into a specific probability of default, given the still early stage of the literature and the heterogeneity in conditions across countries. Rather, some judgment is required on the part of the analyst, taking into account his or her knowledge of country-specific details,22 to assess whether the indicators suggest that debt sustainability is a sufficiently low risk or whether corrective action is needed, given the path of the key variables under the baseline scenario and stress tests. In this regard, scenarios that show rapidly rising debt or debt-service ratios should raise warning flags. Similarly, projections that show debt-service or gross financing need spiking in a particular period may signal a need to restructure debt payments to reduce the risk of liquidity crises.

Case Study

In this section we present simulation results illustrating the use of the debt sustainability analysis framework for a typical low-income country with a large stock of debt. We use the framework described in the previous sections to decompose the factors underlying the trends in the nominal value and NPV of public debt under the baseline scenario. We then analyze the range of risks to the baseline projection by using the results derived from the stress tests.

Data used for the simulation are based on a typical low-income country. Parameters of the simulation are calibrated to replicate the case of a highly indebted country that reaches the completion point under the HIPC Initiative during 2003. In this year, the country benefits from substantial debt relief. The policy issue in this country is whether the post-relief profile of debt will be sustainable or not under a set of economic policies aimed at achieving macroeconomic stability and poverty reduction. From 2004 onward this country is expected to start a gradual process toward graduation from exceptional financing. This process will not be finished by the time of the end of the simulation. The simulation period spans between 2004 and 2023, with model calibration based on historical data referring to the 1993-2002 period. The first projection period in the simulation is 2004. Results are presented on an annual basis for the 2004-07 period, for 2010, 2015, and 2023 (Tables 1 and 2).

The coverage of public debt is critical to the interpretation of the results. In this country, public debt is defined as the sum of the gross liabilities of the central government, as local government debt is generally not significant and data on state-owned enterprises are usually incomplete in a typical low-income country. With incomplete data on the broader definition of public sector, the results of the simulation have to be assessed with caution. Other fiscal risks reflect the presence of institutions outside the coverage of the public sector chosen in this analysis that can engage in important quasi-fiscal operations. The simulation results have to be analyzed in combination with a broader analysis of the possible risks to fiscal sustainability stemming from quasi-fiscal operations of off-budgetary units and contingent liabilities.

At the start of the simulation, the country has a high level of public debt (Table 1). In 2003, the NPV of public debt was above 60 percent. The nominal value of public debt reached 125 percent of GDP, a level well above that of emerging and industrial countries (IMF, 2003). The large stock of debt also implied large flows of payments for debt service. Gross financing needs were high, at nearly 9 percent of GDP, and debt service represented 32 percent of overall revenue collection. Although approximately 85 percent of the debt was contracted from external lenders, the share of domestic debt was substantial for a low-income country. Domestic debt represented almost 20 percent of GDP at the beginning of the simulation period, reflecting a large reliance on domestic sources of deficit financing in the past in the absence of sufficiently large inflows of foreign aid.

In this country, the NPV of public debt has been on a downward trend since 2000 when it reached 140 percent of GDP. The subsequent fall in the ratio was a result of improved macroeconomic conditions and the beneficial impact of debt relief (nearly 80 percent of GDP). The reduction in the stock of debt did not necessarily, however, lead to a more manageable debt-service burden, for the ratio of debt service to revenue increased from 12 percent to 32 percent in the same period, reflecting higher reliance on more costly domestic debt and a decline in domestic revenue collection that largely offset higher grant flows.23 The latter suggests that the volatility of revenue collection may be an important source of vulnerability for debt sustainability in this country.

For the projections, the baseline scenario is one in which tight fiscal policy is assumed to be conducive to a noninflationary environment that would maintain moderate real interest rates and foster growth.24 Thus, real GDP growth is projected to be buoyant, averaging 4.7 percent a year in the first five years of the simulation and stabilizing at 4.5 percent a year in the following period. These buoyant growth rates compare with the poor performance in the previous decade when GDP growth was limited to 0.6 percent a year with large annual fluctuations.25 The average real interest rate on domestic debt is expected to fall from an average 10 percent in 1993-2002 to 4-5 percent in the projection period. Inflation, as measured by the GDP deflator, is projected to subside, reaching 5 percent by the end of the simulation period from high initial values (20 percent a year in 2003).

The improved macroeconomic framework is expected to be sustained by a prudent fiscal policy that would maintain a small primary surplus during the initial part of the simulation period. Maintaining this fiscal stance is achieved by reductions in primary expenditure, since grants are projected to fall while domestic revenue collection is projected to remain broadly constant over the projection period.

The reduction in primary expenditure consistent with this scenario is large. The ratio of public spending to GDP is projected to fall by almost 6 percentage points of GDP from 26 percent in 2003 to slightly more than 20 percent at the end of the simulation period. This implies that the growth rate of real primary spending will be limited to 3.8 percent a year (the historical average growth rate) in the period 2009-23, while real spending is projected to grow only slightly in the initial years of the simulation. Given the projected growth of the population in the same period, the assumed growth rate of primary spending will reflect only a moderate increase in real per capita primary spending. To achieve this target and make progress to reduce poverty in the medium term would entail enhancing the effectiveness of public expenditure and improving its composition in favor of poverty reduction programs that benefit the vulnerable groups in the population. However, this level of expenditure compression may not be politically sustainable in this country, giving rise to risks of political instability that could affect the persistence of the fiscal consolidation effort and highlighting a possible need for more grant aid.

Debt dynamics are projected to improve markedly under the baseline scenario, with the NPV of public debt projected to fall steadily as a share of GDP from 61 percent in 2003 to 22 percent in 2023. Gross financing requirements would be dramatically reduced from 9 percent of GDP in 2003 to less than 2 percent in 2023. Debt service as a share of total revenue is also projected to fall to more sustainable levels—about 6.5 percent at the end of the simulation. However, debt-service indicators are expected to remain relatively high in the next few years. This result highlights the vulnerability of the country adjustment if it does not succeed in stabilizing the domestic revenue base.26

The decline in the NPV of public debt to GDP projected under the baseline scenario primarily reflects the maintenance of prudent fiscal policies, buoyant economic growth, and continued access to concessional financing. In the absence of large primary deficits (as in recent years) and high real interest rates, robust real GDP growth drives a continuous reduction in the NPV of public debt to GDP, as can be seen in the decomposition in Table 2. Exchange rate dynamics are conducive to a lower stock of public debt in the first period of the simulation, while the contribution of exchange rate dynamics to debt stabilization becomes marginal thereafter.

The results of the “no-reform” alternative scenario highlight the importance of achieving the targeted fiscal consolidation to maintain the NPV of debt-to-GDP ratio on a declining path. This scenario, which is based on the assumption that the primary fiscal deficit remains unchanged at the level of 1.6 percent of GDP observed in 2002, does not lead to a large reduction in the debt stock. Rather, the ratio of the NPV of public debt to GDP remains significant at 47 percent of GDP at the end of the projection period.

The stress tests show that growth and exchange rate assumptions have a large impact on public debt projections. Under many alternative simulations, results for both the ratio of public debt to GDP and the ratio of NPV of debt to GDP depart significantly from the baseline (Figure 1). Results are very sensitive to a fall in the real GDP growth rate in both the short term and long run. Debt levels are also sensitive to changes in the exchange rate, especially in the short run. The stress tests also show that excessive debt service continues to be a source of risk for this country despite the large debt relief received in 2003. In the event of negative shocks, debt service could remain at high levels indefinitely (Figure 1).

Figure 1.Indicators of Public Debt Under the Baseline Scenario and Selected Stress Tests, 2003-23

(Percent)

Source: IMF staff projections and simulations.

The fiscal stance projected under the baseline scenario at the end of the simulation (no primary deficit) is consistent with a sustainable path of public debt, which is defined in this case as a nonincreasing ratio of the NPV of public debt to GDP. If, however, long-run growth rates fall below the real interest rate, a tighter fiscal stance would be required.

In sum, achieving fiscal sustainability will be a challenge for this country. Even though the fiscal stance projected under the baseline scenario is consistent with a declining debt-to-GDP ratio, there is great vulnerability to depreciation and income growth shocks. This is especially pertinent in light of the relatively optimistic assumptions for growth relative to the historical record. Furthermore, the success of the adjustment is predicated on cuts in spending (relative to GDP), which will pose a challenge as the government simultaneously attempts to reallocate the budget toward poverty-reducing activities. In this context, an adjustment strategy that relies on raising the revenue effort might be politically more viable. In addition, higher external grants could help alleviate the pressure to cut spending.

Policy Implications of Debt Sustainability Analysis

Debt sustainability analysis can be an important aid in determining the appropriate policy stance to achieve a country’s policy objectives. The policies underlying the baseline scenario are critical, since alternatives can lead to substantially different debt dynamics by affecting growth and other key macroeconomic variables. Debt sustainability results should be assessed in light of the policy options that are available to low-income countries to correct possibly unsustainable paths of public and external debt. In this context, the stress tests can help shed light on the robustness of these baseline scenarios. For example, while the baseline scenario could indicate that debt would remain sustainable, stress tests may indicate that the country would be unduly vulnerable to lower growth or adverse exogenous developments. Debt sustainability analysis could also facilitate an assessment of the extent to which the policies are consistent with the country’s broader development agenda. For example, countries may need higher public spending to achieve the United Nations’ Millennium Development Goals. In these cases, debt sustainability analysis can be useful in highlighting the additional resources that may be needed for countries to achieve the Millennium Development Goals while maintaining a sustainable debt position.

An unsustainable debt position can be addressed in a number of ways—for example, through increased foreign grants, higher domestic revenue mobilization, or lower public spending. As such, an assessment that debt is rising over time, or that a higher primary surplus must be obtained to achieve debt sustainability in the future, does not necessarily imply that public expenditures must be reduced. The feasibility and desirability of the alternative options would need to be judged on a case-by-case basis.

In this regard, it is important to keep in mind that growth is an important determinant of debt sustainability in low-income countries; thus, options to reduce the fiscal debt burden must be assessed in light of their effects on growth, which makes a thorough analysis of the determinants of growth in each country essential.

Nonetheless, many countries may find that debt sustainability requires the implementation of more prudent fiscal policies. Fiscal prudence can contribute to debt sustainability directly, through a sufficiently large primary surplus, and indirectly, through the beneficial effect of low fiscal deficits on growth and macroeconomic stability.

In this regard, the composition of fiscal policy also has an important role. Fiscal consolidations based on expenditure savings stemming from efficiency-enhancing reforms, such as the reduction of subsidies, tend to be more effective at fostering growth than fiscal adjustments based on curtailing productive outlays. In fact, protecting the capital budget is associated with higher growth in low-income countries.27 Over the long run, expanding basic social services and investing in human development and infrastructure are also likely to promote growth. In cases where such investments cannot be funded by cuts in less productive spending, analysts may need to carefully weigh the possible trade-offs between the growth-enhancing effects of such investments and the costs of the borrowing needed to finance them.

Conclusion

In this paper, we have presented a simple, pragmatic, and flexible framework that analysts might use to assess debt sustainability in low-income countries. The framework allows one to analyze the main determinants of debt dynamics, takes account of multiple constraints on debt sustainability via multiple indicators, examines the main risks to debt sustainability via stress tests, and emphasizes the importance of clarifying the underlying assumptions behind debt projections. This framework is simply meant to be suggestive, and it can and should be adjusted and tailored to the specific issues in the country of interest. Similarly, the framework is not intended to yield a simple “yes” or “no” answer to the question of whether a country’s debt is sustainable; rather, it provides the basic building blocks for formulating an assessment of the balance or risks, taking into account both the empirical literature on key determinants of debt crises28 and the analyst’s knowledge of country-specific factors. In doing so, it is hoped that such a framework can assist in formulating appropriate macroeconomic policies.

Technical Appendix

Derivation of Fiscal Debt Equations

We model public debt dynamics as follows:

The debt stock is composed of debt instruments denominated in both domestic and foreign currencies. Domestic currency debt (Dtd) evolves according to the real interest rate in the domestic market (rd) and the change in the domestic GDP deflator (πd), while the evolution of the foreign currency debt (Dtf), expressed in domestic currency, is affected not just by the foreign real interest rate (rf) and inflation rate,29 but also by changes in the real exchange rate (ε).30 Finally, H is a residual that captures all other debt-creating flows, such as cross-currency movements.

The analysis looks at debt stocks relative to GDP. Therefore, defining lower-case variables as upper-case variables expressed as a proportion of nominal GDP (e.g., dt+1 = Dt+1/Yt+1) and dividing both sides of equation (A1) by nominal GDP, Yt+1 yields the following:

where g = the real GDP growth rate. Letting dtf=αtdt (where α is the share of debt held in foreign currency) and dtd=(1αt)dt, subtracting dt from both sides, and rearranging yields equation (2) in the main text:

The fiscal template uses this equation to calculate the change in debt and to separate the different channels that contribute to the evolution of the debt-to-GDP ratio.

To see how the equation for the evolution of the NPV of debt-to-GDP ratio is derived (equation (3) in the main text), note first that the NPV of foreign currency debt at the end of time period t is equal to the following:

where Ntf = the NPV of foreign currency debt (expressed in units of local currency), DSf* is the foreign currency debt service (* denotes a value denominated in foreign currency) in each period on the stock of existing debt as of time period t, δ is the foreign-currency discount rate, and e is the nominal exchange rate (expressed in local currency per unit of foreign currency).

In time period t+1, DSt+1f will be paid off and new debt will be contracted, so that the NPV of foreign currency debt at the end of t+1 will be the following:

where the NPV of new debt contracted in time period t+1 is equal to nominal gross new borrowing (BO times one minus the grant element (X) of this borrowing. Inserting equation (A4) in (A5), one can see that there is a clear relationship between Ntf andNt+1f

(Note that debt service is now expressed in local currency units). Gross new foreign currency borrowing will equal foreign currency debt service minus the portion of the primary balance that is funded via foreign currency borrowing:

Combining equation (A6) with equation (A7) and rearranging yields

It is assumed that all domestic debt is nonconcessional, so that the present value of domestic debt equals the nominal value,31 which evolves as follows:

where Nd is the NPV of domestic debt, Dd is the nominal value of domestic debt, id is the nominal interest rate on domestic debt, and Pd is the portion of the primary surplus financed by domestic debt.

Combining equation (A8) with (A9) and letting α equal the share of the NPV of debt held in foreign-currency yields the following expression for the evolution of the NPV of total debt:

Dividing both sides of the equation by GDP in t+1 and letting smaller case symbols indicate amounts as a share of GDP yields:

Now define the “real discount rate,” rδ, the real domestic interest rate, rd, and the rate of real exchange rate depreciation, ε, as follows:

Inserting these expressions into equation (A11), subtracting nt from both sides, and adding a residual term, η, to capture all other effects on the NPV yields equation (3) in the main text:

Intuitively, equation (A13) states that the NPV grows at the difference between the real interest rate and the real growth rate (with the real interest rate defined as a weighted average of the real discount rate on foreign-currency debt and the real interest rate on domestic debt), plus real exchange rate depreciation, minus the primary balance, and minus the grant that is implicit in new concessional borrowing.

References

    BaldacciEmanueleBenedictClementsSanjeevGupta and CarlosMulas-Granados2004“Persistence of Fiscal Adjustments and Expenditure Composition in Low-Income Countries,”Chapter 3 in Helping Countries Develop: The Role of Fiscal Policyed. bySanjeevGuptaBenedictClements and GabrielaInchauste (Washington: International Monetary Fund).

    BaldacciEmanueleAryeHillman and NaokoKojo2003“Growth, Governance, and Fiscal Policy in Low-Income Countries,”Chapter 4 in Helping Countries Develop: The Role of Fiscal Policyed. bySanjeevGuptaBenedictClements and GabrielaInchauste (Washington: International Monetary Fund).

    ChalkNigel and RichardHemming2000“Assessing Fiscal Sustainability in Theory and Practice,”IMF Working Paper 00/81 (Washington: International Monetary Fund).

    CohenDaniel1997“Growth and External Debt: A New Perspective on the African and Latin American Tragedies,”CEPR Discussion Paper No. 1753 (London: Centre for Economic Policy Research).

    CuddingtonJohn1997“Analyzing the Sustainability of Fiscal Deficits in Developing Countries,”World Bank Policy Research Working Paper No. 1784 (Washington: World Bank).

    EasterlyWilliam2001“Growth Implosions, Debt Explosions, and My Aunt Marilyn: Do Growth Slowdowns Cause Public Debt Crises?”World Bank Policy Research Working Paper No. 2531 (Washington: World Bank).

    GuptaSanjeevBenedictClementsEmanueleBaldacci and CarlosMulas-Granados2004“Fiscal Policy, Expenditure Composition, and Growth in Low-Income Countries,”Chapter 2 in Helping Countries Develop: The Role of Fiscal Policyed. bySanjeevGuptaBenedictClements and GabrielaInchauste (Washington: International Monetary Fund).

    GuptaSanjeevBenedictClementsAlexanderPivovarksy and ErwinTiongson2004“Foreign Aid and Revenue Response: Does the Composition of Aid Matter?”Chapter 14 in Helping Countries Develop: The Role of Fiscal Policyed. bySanjeevGuptaBenedictClements and GabrielaInchauste (Washington: International Monetary Fund).

    HellerPeter2003Who Will Pay? Coping with Aging Societies Climate Change and Other Long-Term Fiscal Challenges (Washington: International Monetary Fund).

    HemmingRichardMichaelKell and AxelSchimmelpfennig2003Fiscal Vulnerability and Financial Crises in Emerging Market Economies IMF Occasional Paper No. 218 (Washington: International Monetary Fund).

    International Monetary Fund2001Government Finance Statistics Manual (Washington).

    International Monetary Fund2003World Economic Outlook September 2003: Public Debt in Emerging Markets Chapter 3 World Economic and Financial Surveys (Washington: International Monetary Fund).

    International Monetary Fund2002“Assessing Sustainability,”prepared by the Policy Development and Review Department (Washington: IMF). Available via the Internet: http://www.imf.org/external/np/pdr/sus/2002/eng/052802.htm.

    KraayAart and VikramNehru2004“When Is Debt Sustainable?”World Bank Policy Research Working Paper No. 3200.

    PatilloCatherineHelenPoirson and LucaRicci2002“External Debt and Growth,”IMF Working Paper 02/69 (Washington: International Monetary Fund).

Alternative definitions of debt sustainability have been presented in the literature. See, for example, IMF (2003) and Chalk and Hemming (2000).

For expositional simplicity, it is assumed that the interest rate is constant over time. This condition also assumes that the present value of the government’s debt in the indefinite future converges to zero, so that no “Ponzi game” is possible. See Cuddington (1997) for further discussion.

The framework is based largely on an approach that has been developed primarily by the IMF’s Policy Development and Review Department (PDR) for emerging market countries (IMF, 2002). However, the framework in this chapter is modified somewhat to address some specific issues that are common in low-income countries, and the authors take responsibility for any errors.

See Heller (2003) for a discussion of the importance of framing fiscal policy in a long-run context.

Indeed, it is often a condition of concessional financing that a country increase its social spending, reducing or eliminating (increasing) any fiscal surplus (deficit) that it may have been running.

This is the point at which all new debt is on commercial terms, although some existing debt may still be on concessional terms.

This assumes that the general government has not yet officially assumed the liability, in which case it would become part of general government debt and not be shown as a contingent liability.

Similarly, NPV calculations may facilitate comparisons of indebtedness across time within the same country when the terms on which the debt is denominated are changing.

In this regard, it is useful to complement fiscal debt sustainability analyses with external debt sustainability analyses, in which the focus is on the evolution of the balance of payments, and total external debt (both public and private) as a share of exports is typically a key indicator of vulnerabilities in generating sufficient foreign exchange. However, this chapter restricts its focus to fiscal debt sustainability issues.

Short-term debt is debt with a maturity of one year or less.

On the other hand, some may argue that GDP is better than current revenue at approximating the potential revenue that a country could raise in the event of a debt-servicing crisis.

The coverage of public sector debt in debt sustainability analysis should be consistent with the coverage of public sector flows; for example, if public sector debt is defined for the general government, then revenue should also be defined for the general government. As a result, when the definition of the public sector differs across countries, the debt-to-revenue ratio is more likely to be similar across countries than is the debt-to-GDP ratio; for example, the ratio of general government debt to general government revenue in Country X may be somewhat comparable to the ratio of central government debt to central government revenue in Country Y, whereas the ratio of general government debt to GDP in Country X is likely to differ significantly from the ratio of central government debt to GDP in Country Y simply because of the different coverage.

rf - (1+if)/(1+πf)-1, where if = the average nominal interest rate on foreign currency debt and πf = the GDP deflator in the country in whose currency the foreign currency debt is denominated. If the debt is denominated in several currencies, this could reflect an average of partner country inflation or, as a simplifying approximation, inflation in the country of the most important partner (e.g., the United States or the euro area).

A real depreciation of the local currency (ε > 0) leads to an increase in foreign currency debt, expressed in local currency terms. If the debt is denominated in several currencies, this could reflect the average real depreciation against these currencies or, as a simplifying approximation, real depreciation against the most important currency could be used as a proxy (e.g., the U.S. dollar or the euro).

If δ is the nominal, time-invariant U.S. dollar discount rate, then (1+δ)=(1+πtf)(1+rtδ), where πtf = the U.S. GDP deflator inflation rate.

For example, Easterly (2001) finds that a slowdown in growth rates was a primary determinant of the debt crises in Heavily Indebted Poor Countries (HIPCs).

For countries where exceptional noneconomic episodes (e.g., wars) have severely affected economic Performance in the recent past, the period covered by such episodes may have to be excluded when computing historical averages. Alternatively, historical averages for similar countries could be used.

Unless, for example, there has already been a substantial adjustment that the baseline scenario predicts will be difficult to sustain.

If the standard deviation of the one-period growth rate is SD, then the standard deviation of the average growth rate over n periods is SD/√n. This is to test the sensitivity of the results to a lower-than-expected long-run growth rate.

It may also be useful, in some cases, to assess the favorable impact on debt sustain ability of substituting grants for loans.

In this regard, Kray and Nehru (2004) find that the quality of a country’s policies is a key factor affecting the relationship between debt indicators and the likelihood of a debt crisis.

For more on the relationship between domestic revenue and grants, see Clements and others (2004).

On the relationship between fiscal policy and growth and its determinants in low-income countries, see Baldacci, Hillman, and Kojo (2003).

The standard deviation of GDP growth rates in the period 1992-2002 is estimated at 5.6 percent, or more than nine times the average growth rate in the same period.

For more on the persistence of fiscal adjustment, see Baldacci and others (2004).

See, for example, Gupta and others (2004).

See the studies cited in footnote 21.

See footnote 13.

(1+ɛt+1)=et+1et(1+πt+1f1+πt+1d), with e defined as the nominal exchange rate in units of local currency per unit of foreign currency.

See the subsection in the main text on key indicators.

    Other Resources Citing This Publication