Abstract

This chapter reviews different concepts of debt sustainability and gives illustrative results on debt limits for economies based on macroeconomic characteristics prevailing in the Caribbean region. In particular, we deal with three important policy-related issues: First, we delineate key aspects of the different approaches to measure fiscal sustainability and public debt limits; second, we measure the sustainability of fiscal policy and the extent of over- or under-borrowing by the public sector over the last two decades; and third, we derive debt benchmarks through illustrative scenarios, using reasonable assumptions about growth and interest rate shocks from the region’s economies.

This chapter reviews different concepts of debt sustainability and gives illustrative results on debt limits for economies based on macroeconomic characteristics prevailing in the Caribbean region. In particular, we deal with three important policy-related issues: First, we delineate key aspects of the different approaches to measure fiscal sustainability and public debt limits; second, we measure the sustainability of fiscal policy and the extent of over- or under-borrowing by the public sector over the last two decades; and third, we derive debt benchmarks through illustrative scenarios, using reasonable assumptions about growth and interest rate shocks from the region’s economies.

After 50 years of leading Caribbean economic transformation, the public sector is overburdened with debt and is restraining economic recovery. This has instigated a policy debate on the sustainability of fiscal policy and public debt amid new strains created by the global financial and economic crisis, which exposed longstanding structural weaknesses in the region. In particular, in 2011 the median public debt ratio was about 71 percent of GDP, having risen from about 65 percent before the global crisis in 2008.1 Importantly, many in the Caribbean now recognize that their fiscal policy is headed down an unsustainable path, their public debt is too high, and fiscal adjustment strategies are required to lower public debt ratios to a level that would restore sustainability and promote growth. Without adjustment, the region would remain highly indebted for a long time owing partly to the slow recovery of key trading partners and continuing fiscal deficits. Yet some important questions need to be answered, including questions about the pace of fiscal consolidation, the short-term costs of consolidation, and the levels to which the region’s economies should reduce their debt ratios over the medium term.

The literature on fiscal and debt sustainability is vast and expanding, but it yields no single agreed-upon definition. Fiscal sustainability concepts range from simple to complex approaches. Simple approaches include, as an example, the debt stabilizing primary balance, whereas complex approaches aim to measure optimal debt levels within welfare-maximizing frameworks. In addition to the lack of a single definition of fiscal sustainability, a number of other gaps remain in the existing literature as well. These include the general absence or inadequacy of the treatment of economic uncertainty in existing debt sustainability models and the over-concentration of empirical work on developed and large emerging market economies.

Overall, the results suggest that the region is pursuing policies that are not in accord with a sustainable path for fiscal policy and that it has over-borrowed relative to the calibrated debt limits for the region.2 Owing to this, Caribbean governments will need to adopt ambitious fiscal adjustment or other debt reduction strategies over the medium term to reduce the drag of high public debt on economic activity.

This chapter begins by describing the Caribbean context that motivates the analysis of fiscal sustainability. It then sketches definitions of fiscal sustainability and debt limits, with a discussion of their operational measures and implications for public debt ratios. After that, it discusses some illustrative results based on macroeconomic characteristics of the countries analyzed. It ends with concluding comments.

Context and Motivation

The public sector in Caribbean economies has been a key driver of growth and transformation for a long time. For historical reasons, the original leading role of this sector was focused on two goals: the creation of public goods and the transformation of the economy and society from a colonial orientation to a modern one. In this context several indicators of human well-being have indeed improved, revealing strong social, economic, and political progress, including high per capita incomes and health standards, a high ranking on the United Nations’ human development index, and thriving and competitive parliamentary democracies.

However, in the last 20 years the public sector has also been used to somewhat compensate for the collapse of traditional export sectors, including sugar and banana exports, as competitive pressures have intensified. This enhanced role for the public sector, facilitated by deficit financing, caused a rapid growth in public debt, sometimes financing projects without a direct cash return (including poorly managed public enterprises). Added to this has been the very high cost of coping and rebuilding after destructive natural events, such as annual hurricanes. These factors have contributed to high public debt levels, which have increased further as a consequence of the global crisis and brought the sustainability of fiscal policy and public debt to the forefront of the policy debate.

Today, large public debt imposes a high cost on the region’s economies and puts at risk the social and developmental gains made over the last 50 years. There are several channels through which these costs and risks are transmitted to the economy, described next.

Higher Debt May Lower Growth

There is some empirical evidence that high public debt reduces long-run economic growth (see, e.g., Kumar and Woo, 2010; and Greenidge and others, 2012). For the region’s countries, Greenidge and others (2012) show that gross debt beyond a threshold of 55–56 percent of GDP is associated with lower economic growth. Indeed, there is a nonlinear relationship between debt and growth. Excessive public debt crowds out private sector investment and lowers economic growth; higher debt requires higher taxes to service the debt, which reduces investment and growth. Further increases in government expenditure financed by higher debt from already elevated levels are likely to be self-defeating, leading to lower long-term growth, as public sector debt crowds out private investment.

Higher Debt Heightens Roll-Over Risks

Higher public debt is likely to increase the public sector borrowing requirement for two reasons. First, a higher debt ratio requires a larger share of GDP to service this debt. Second, as the debt ratio becomes larger, investors are likely to demand higher interest rates and shorter maturities, as has happened for some of the region’s countries (see Chapter 3). As the size and frequency with which the government needs to tap the debt market increases, this directly increases the roll-over risk.

Higher Debt Increases Vulnerability Through Sovereign–Bank Interlinkages

High public debt held largely by the domestic banking system increases a country’s vulnerability to shocks. For example, a shock that negatively affected the sovereign’s ability to repay its debt would impact the domestic banking system, which could lead to a negative feedback loop. Further, where the debt is largely held by domestic banks, in the event of a debt restructuring the scope for debt relief is reduced considerably (see Chapter 3).

Higher Debt Reduces Policy Flexibility and Increases Vulnerability

A high debt ratio reduces the space for policy flexibility and the ability to respond to shocks. Countries that had a high debt ratio at the time of the global crisis were unable to respond with countercyclical fiscal policies due to the lack of fiscal space. Instead, many countries had to tighten their fiscal stance to stave off a financing crisis, thus pursuing procyclical policies. Furthermore, increases in interest rates raise the debt service burden and may hasten debt distress.

Two policy questions are worth answering: Taking into account their stage of development and vulnerability to shocks, what public-debt-to-GDP ratio should the region’s economies adjust to? And what should be the speed of adjustment to that ratio? While theory offers little or no guidance on these matters, some Caribbean countries adopted a 60 percent debt ratio before the global crisis, and most considered a 60 percent debt ratio as a safe medium-term public debt target. Yet many countries have encountered fiscal and debt distress and are unlikely to meet these medium-term targets without ambitious adjustments. Against this background, it would appear that a total rethinking of fiscal sustainability and safe debt limits for all of the region’s economies is required.

Debt Sustainability and Debt Limits: Theory

Definitions

Debt sustainability is an elusive concept. Determining whether a government’s debt is sustainable is not a straightforward matter (Chalk and Hemming, 2000). In the context of IMF programs, a government’s fiscal policy is regarded as sustainable if its path does not imply an abrupt change in primary balances, and the government has sufficient financial resources to meet all of its maturing obligations. In addition, the fiscal adjustment path pursued by the government must be economically and politically feasible.

The concept of a debt limit is closely associated with fiscal sustainability. It represents the explicit stock implications to the flow variables that are used to calibrate fiscal sustainability. There are at least three ways debt limits can be defined: First, a debt limit can be the debt ratio to which the economy converges in the steady state (see, e.g., Blanchard, 1990; and Blanchard and others, 1990). Second, it could be the level of debt that the economy can service or carry without generating debt distress and requiring debt restructuring. Third, a debt limit could be the optimal level of debt given an economy’s policy objectives, stage of development, and policy environment.

A number of countries have adopted debt limits to anchor fiscal policy. They have done so in the context of fiscal responsibility legislation, which provides the legal basis for a medium-term macroeconomic framework. Some of these debt limits or targets have been adopted after a major crisis, others to protect prudent countries from the spillover effects of other countries’ fiscal distress, typically in a monetary and financial integration arrangement. Examples of the latter include the limits adopted by the euro area and the Eastern Caribbean Currency Union (ECCU). Both have a 60 percent debt-to-GDP-ratio limit on public debt. This limit is included in the accession principles for the euro area, whereas for the ECCU countries the limit is a target to which they are aiming to reduce their debt by 2020.3 However, there is no theory underpinning their adoption of such limits. Instead, their adoption largely rests on the degree of comfort policymakers feel with particular targets relative to where their own country’s debt ratios currently stand.

Measures of Fiscal Sustainability

Several indicators have been proposed in the literature to measure fiscal sustainability. These include, among others, the debt stabilizing primary balance, the reaction of fiscal policy to higher public debt, and debt benchmarks. The first two are discussed here, whereas debt benchmarks and debt limits are presented in the next section.

Debt Stabilizing Balances and Fiscal Efforts

This approach involves calculating the primary budget balance that would stabilize public debt at its current level, thus “debt stabilizing primary balance” (see Box 6.1).4 Fiscal sustainability is then determined by comparing the actual primary balance with the debt stabilizing primary balance. If the actual balance is less than the debt stabilizing one, this implies that public debt is rising and unsustainable. This approach is consistent with the IMF framework for assessing sustainability.

Although this measure of fiscal sustainability has many advantages, it also has some important drawbacks. One advantage is its simplicity: it is simple and easy to apply and the results are easy to interpret. This model, however, does not incorporate uncertainty and its key parameters—growth, interest rate, and the primary balance—are assumed to follow deterministic paths. As a result, there is no presumed feedback from the debt stock to fiscal policy or the environment within which policy is made. Finally, the model presumes that the authorities would be able, somehow, to achieve the required primary surplus and therefore avoid defaulting on the country’s public debt obligations.

Fiscal Policy Reaction to Debt Levels

Another way of assessing debt sustainability is to look at the reaction of fiscal policy to rising debt levels. Under this approach, the primary balance adjusted for the effects of temporary factors is presumed to respond to public debt, where a positive response to debt implies a policy with long-term solvency (Bohn, 1998). By taking account of the constraints and objectives of policy, this framework is more flexible than the simple debt stabilizing primary balance approach.

A number of studies on fiscal sustainability apply this approach. These include Melitz (1997), Égert (2010), and Ghosh and others (2013) for OECD countries; Debrun and Wyplosz (1999) and Gali and Perotti (2003) for the euro area; and Burger and others (2011) for South Africa. A key drawback, however, is that this approach is silent on when and how fiscal policy should adjust to rising debt. It only requires the policymaker to commit to adjust at some time in the future. Therefore, the reliance on the credibility of the policy framework and the policymaker has to be considered. As a result, it may take a very long time to achieve fiscal sustainability, in particular if the market does not believe the policymaker’s commitment.

The Long-Run Fiscal Sustainability Condition

This approach uses the government’s flow budget constraint

bt=(1+r)bt1Xtσt,(B1)

where bt is real government debt, rt the real interest rate, xt the real primary balance, and st real seigniorage.

Forward iteration on (B1) combined with the condition

lim(1+r)(j+1)bt+j=0(B2)

implies

bt1=i=0(1+r)(j+1)(Xt+1+σt+1).(B3)

This equation is the government’s lifetime budget constraint, thus the government finances its debt at the end of the period t−1 by raising seigniorage revenue and running primary surpluses with an equal present value.

The most basic tool for fiscal sustainability analysis uses a steady-state version of the lifetime budget constraint. Equation (B3) can be rewritten in terms of stocks and flows expressed as fractions of GDP. Letting y, represent real GDP and defining b¯t=btyt,X¯t=Xtyt.σ¯t=σtyt equation (B3) can be rewritten as follows:

b¯t-1=i=0(1+r)(i+1)(X¯t+i+σ¯t+i)yt+iyt-1.(B4)

In steady state, real GDP grows at a constant rate g, and the primary surplus (X¯) as a fraction of GDP and seigniorage (σ¯) as a fraction of GDP are constant. As a result, (B4) reduces to

b¯t1=i=0(1+g1+r)i+1(X¯+σ¯).(B5)

Assuming that r > g, equation (B5) becomes

b¯t1=b¯=(X¯+σ¯)/r¯,(B6)

where r¯ = (r − g)/(1 + g).

Two measures of fiscal sustainability can be derived from this equation. First, using medium-term values of X¯,σ¯, r, and g we can derive an estimate of b¯. If the government’s actual stock of debt exceeds this estimate, then a government’s finances are unsustainable. Second, by rearranging equation (B6) to

X¯=r¯b¯σ¯,(B7)

we can use equation (B7) to determine the necessary size of the primary balance to ensure fiscal sustainability given estimates of σ¯,r¯,, g, and b¯.

This box is adapted from Burnside (2004).

Debt Limit Measures

Debt limits are meant to reflect the maximum level of debt that can be contracted without imposing undue welfare costs or instigating debt distress and default. The literature distinguishes between the optimal debt level, a normative concept, and the crisis-free debt level—the so-called “safe debt” level.5 We follow this distinction and discuss these two broad concepts separately. First, we discuss the safe debt concept, starting with debt benchmarks, before turning to debt thresholds based on uncertainty and probabilistic methods (natural debt limits, value-at-risk and fair spread approaches). Second, we briefly touch on the literature about optimal debt. The concepts under the first category are used in the illustrative scenarios later in this chapter, and the comparison with actual debt ratios will give a sense of over- and under-borrowing.

Debt Benchmarks

Debt benchmarks can be used to judge whether a country has over-borrowed and may face debt distress at some time in the future. A country’s debt benchmark is derived using historical values for its ratio of revenues to GDP (T/GDP) and its ratio of primary spending to GDP (G/GDP), both of which are discounted by the differential between its real interest (r) and GDP growth (g) rate. Thus, each country’s debt benchmark has the following form:

bbm=T/GDPG/GDPrg=PB/GDPrg.(6.1)

Of the different approaches to calculate debt benchmarks, we will discuss four methods. The first approach is the long-term debt benchmark, also referred to as the Blanchard ratio (see, e.g., Blanchard, 1990; and Blanchard and others, 1990). This standard approach is based on steady-state or long-term values for the fiscal indicators. Basically, a country’s historical track record forms the basis for the interest rate, growth rate, and primary balance paths.

Second, there is an approach based on the exceptional fiscal performance of a country. This approach uses the maximum primary surplus during a predefined period and discounts it with the average interest and growth performance.

A third approach is the signal approach. It relies on the signaling effect of debt distress and uses the debt ratio at which a country experiences debt distress as the debt benchmark. Thus, this ratio is based on a country’s history with debt distress. For example, if a country experienced debt distress above a debt ratio of 50 percent of GDP, then a debt ratio of 50 percent or higher is considered unsafe. In such a scenario, there is a signal that a crisis is likely. If the country’s debt level does not exceed this threshold then there is no signal of debt distress.

The fourth approach unifies the fiscal reaction function methodology and the long-term approach to measuring fiscal sustainability. Based on the parameters of the fiscal reaction function and estimates of the interest rate-growth differential, a country’s debt limit and its associated fiscal space is calibrated (see Ghosh and others, 2011). This approach makes a distinction between the long-term public debt to which an economy converges in the steady state, on the one hand, and the maximum sustainable public debt, the level of debt immediately before a country loses market access, on the other.

Later in this chapter we will provide illustrative scenarios based on the long-term and exceptional fiscal performance debt benchmarks.

Debt Thresholds Based on Uncertainty and Probabilistic Methods

Debt limits based on probabilistic methods explicitly take into account that governments face high amounts of uncertainty regarding their revenues and expenditures and how these affect sustainable debt ratios. Since fiscal revenues and expenditures are subject to shocks, a steady-state level of public debt that ignores downside risks might not give a full picture of sustainable debt. Instead, the methods discussed here ask whether a current debt ratio is sustainable given the current macroeconomic environment and future prospects. In addition, interest and exchange rate movements add to the uncertainties. However, one caveat should be kept in mind: While these methods relax some of the problematic assumptions used in the steady-state methods, new assumptions have to be added, making them susceptible to criticism as well (see Burnside, 2004).

Strategies for Including Macroeconomic Uncertainty

There are different strategies for including macroeconomic uncertainty in the assessment of fiscal sustainability. Each of the approaches is specialized and looks at a particular aspect of fiscal sustainability. We will consider three approaches: the natural debt limit based on Mendoza and Oviedo (2009), the value-at-risk (VaR) approach, and the fair spreads approach.6

Natural Debt Limit Approach

The natural debt limit, based on Mendoza and Oviedo (2009), takes into account the uncertainty and risks faced by policymakers. In particular, it considers that volatile revenues and expenditures can have a devastating effect on debt sustainability if negative shocks are realized. A government that wants to credibly commit to service its debts in the future under all circumstances has to bear in mind that it could be faced, for example, with a prolonged period of below-average revenues. To account for this uncertainty, the natural debt limit is calculated by adjusting average revenues and expenditures downward by two standard deviations:

bnl=TMin/GDPGMin/GDPrg.(6.2)

To further take into account growth and interest rate shocks, we propose an additional measure for the natural debt limit. This measure—our preferred one—adjusts real interest rates upward and growth rates downward by one standard deviation:

bnl=TMin/GDPGMin/GDPrMaxgMin.(6.3)

The key idea is that the minimum primary balance generated indicates the fiscal stance that the authorities can credibly commit to in the presence of economic shocks going forward. The advantage of the debt limit approach is that it explicitly incorporates uncertainty in the assessment of fiscal/debt sustainability. Therefore, it takes into account the possibility of default when the primary balance cannot be increased beyond a certain level.

Value-at-Risk Approach

The approach used by Barnhill and Kopits (2004) adapts the value-at-risk (VaR) methodology, which is commonly used to assess the risk of financial assets, to the government’s net worth. In particular, this measure of the government’s net worth compares the value of outstanding debt to the present value of net flows used to service the debt. However, it excludes measures that are only used to close the budget constraint, as this concept is an ex-ante assessment of a government’s finances (Burnside, 2004). By modeling the government’s net worth as a stochastic process, the probability of the net worth becoming negative can be assessed, and that can be interpreted as the probability of a government default. Thus, in this approach fiscal risks and their impacts on the government’s net worth position are modeled and estimated explicitly, showing the vulnerability of the net worth. While in theory this approach can factor in contingent liabilities, it requires a vast amount of information from public sector balance sheets, thus precluding its application to a large set of countries.7

Fair Spreads Approach

The method by Xu and Ghezzi (2003) proposes to compute “fair spreads” on public debt to assess default probabilities. The flows of the government budget are modeled as stochastic processes, which are used to estimate default probabilities. These probabilities are then mapped into term structure models to compute fair spreads. However, in contrast to the other approaches discussed, this measure is more closely linked to liquidity than to pure solvency (Burnside, 2004).

The Optimal Debt Level

The literature on optimal debt levels contains two strands (IMF, 2013b). The first strand focuses on calibrating an absolute level of the debt-to-GDP ratio. Most of the models attempting this have faced difficulties modeling the complex interplay of objectives, costs, and distributive effects in a tractable way and have instead opted for a simple approach, modeling just one aspect of the issue from which welfare costs and gains are then derived. An important work belonging to this strand is that of Aiyagari and McGrattan (1998), who calibrate the optimal debt level for the U.S. economy. Their results are based on certain special assumptions about the behavior of the government, households, and borrowing constraints. At the time of their work (1998), their result suggested that the debt level of the United States then was optimal and that further increases over a wide policy space did not seem to yield measurable welfare costs.

Another recent strand of the literature on optimal debt ratios looks into the optimal debt profile. These models essentially focus on the tax smoothing properties of rolling over debt, in the spirit of Barro (1979).

Illustrative Results of Fiscal Sustainability and Debt Limits for Caribbean Economies

In this section, we apply some of the fiscal sustainability and debt limit concepts discussed above to illustrate scenarios based on the 13 Caribbean countries. This will give us illustrative results about how high the debt limits are and how they depend on underlying macroeconomic characteristics. First, we discuss debt stabilizing primary balances, before turning to debt limits.

Debt Stabilizing Balances and Fiscal Efforts

An analysis of debt stabilizing primary balances for the region shows that policies in many countries are not in line with a sustainable fiscal policy. The stabilization of public-debt-to-GDP ratios at 2011 levels would require adjustment in seven countries (Table 6.1). The fiscal adjustment efforts required range from 1.3 percent of GDP for Barbados to 5.6 percent in St. Lucia. Since the 2011 debt ratios are high, stabilizing at this level would not help to lower the vulnerability to shocks, particularly in the highly indebted countries. In this context, it would be important for most countries to reduce their debt ratios to levels that make them less vulnerable to shocks and bring fiscal policy onto a sustainable path. We choose 60 percent of GDP by the end of the decade as the debt ratio limit, since the ECCU countries have committed themselves to reduce their debt ratios to this level by 2020.

Table 6.1

Debt Stabilizing Balances: Illustrative Fiscal Adjustments

(In percent of GDP)

article image
Source: Authors’ calculations.

Reducing public debt ratios to 60 percent of GDP by 2020 would require large fiscal adjustments (above 2 percent of GDP) in eight countries. Of these, five countries—Antigua and Barbuda, Barbados, Grenada, Jamaica, and St.

Lucia—would require a fiscal adjustment above 5 percent of GDP relative to their primary balances in 2011. In particular, Barbados would require an adjustment of 6.1 percent of GDP. The fiscal adjustment would be somewhat smaller in the other three countries (Dominica, St. Kitts and Nevis, and St. Vincent and the Grenadines), ranging between 2 and 4 percent of GDP. One country would need an adjustment below 2 percent (Belize).

Debt Limits

In this section, we show illustrative results based on the region focusing on two debt limit concepts: long-term debt benchmarks and natural debt limits. In addition to deriving results under various specifications, we compare the calculated debt ratios with the actual ones as of end-2011 to convey a sense of whether countries have over- or under-borrowed. Finally, we also show debt benchmarks for exceptional fiscal performance.

For the calculation of these concepts, we construct three groups based on the macroeconomic characteristics of the region:

  • Group 1: countries with debt ratios above 90 percent of GDP

  • Group 2: countries with debt ratios between 60 and 90 percent of GDP, and

  • Group 3: countries with debt ratios below 60 percent of GDP.

In addition, we calculate the average for the Caribbean. We use data on revenues, primary expenditures, and real growth rates from the October 2012 World Economic Outlook database. Ideally, we would like to use averages over the last 20 years to derive steady-state values for the key parameters. However, since in some cases the available data is limited to a shorter horizon, we use the longest available data for each country case (Table 6.2, column 2). Table 6.2 gives an overview of the averages for revenues, primary expenditures, and real GDP growth rates.

Table 6.2

Measures of Debt Sustainability: Illustrative Underlying Variables

(In percent of GDP)

article image
Source: Authors’ calculations.

Another important parameter for the calculation is the real interest rate. However, these rates are hard to measure, as debt instruments differ in their characteristics.8 Another complication for measuring real interest rates in the region’s countries arises from the lack of fully developed debt management practices and underdeveloped capital markets, in particular for domestic debt. This might lead to a downward bias in the interest rates paid on domestic debt instruments, as risk factors are not properly taken into account. Using average interest rates on public debt, calculated as expenditure on interest divided by the previous period’s debt stock—as is done in debt sustainability analyses—might not give an adequate picture of the risks associated with government debt. To estimate real interest rates, we use the average of JPMorgan’s Central American and Caribbean Index for the period 2002–11, discounted by the U.S. GDP deflator and, where available, we use yields to maturities on a country’s bonds (see Table 6.2). We also explore the effects of increasing or decreasing the interest rate in a sensitivity analysis.

However, a number of countries in the Caribbean can borrow on concessional terms, which might render a real interest rate oriented at the market rate too high. Therefore, we consider that countries that are eligible for concessional assistance from the IMF’s Poverty Reduction and Growth Trust might be subject to lower real interest rates.

Long-Term Debt Benchmarks

First, we focus on the long-term debt benchmark (compare equation 6.1). The derivation of the long-term debt benchmark raises two issues: The average primary balance needs to be in surplus and the real interest rate needs to be higher than the real GDP growth rate in order to get meaningful debt benchmarks. If the average primary balance is negative, the growth rate would need to exceed the real interest rate in order to get a positive debt benchmark. The case of growth rates exceeding real interest rates, also known as “dynamic inefficiency,” implies that there would not be a need to generate primary surpluses in order to achieve fiscal sustainability (see Blanchard and others, 1990). However, in the long run it is expected that real interest rates would be higher than real growth rates.

Thus, the primary balance needs to be, on average, in surplus. However, for quite a few Caribbean countries, average primary balances were negative over the observation period. Nevertheless, to calculate the debt benchmark we use the smallest positive primary balance over the observation period (see Table 6.2). This means the calculated long-term debt benchmark should be interpreted as an upper limit, since on average the primary balance is actually lower.9

In the baseline case, long-term debt benchmarks are on average 40.3 percent of GDP (Table 6.3, column 4). However, there are big differences across countries with different characteristics. For example, a country with exhaustible natural resources would be able to generate high primary surpluses more easily than other countries. However, such a country would also need to run primary surpluses, since the revenue stream would be based on a finite and possibly price-sensitive resource and it would be necessary to build buffers for the future. In addition, the natural resource boom in the years leading up to a financial crisis also would have helped in generating strong GDP growth.

Table 6.3

Measures of Debt Sustainability: Illustrative Long-Term Debt Benchmarks

(In percent of GDP)

article image
Source: Authors’ calculations.

This combination of factors leads to a high long-term debt benchmark as exemplified by country B in group 3. However, we need to keep in mind that the long-term debt benchmark might not capture all aspects in a country with exhaustible natural resources. Similarly, country D in group 1 shows that high primary surpluses would lead to a long-term debt benchmark above the Caribbean average. Now, we can compare the derived long-term debt benchmark with actual debt-to-GDP ratios at the end of 2011. This gives a sense of whether countries are over- or under-borrowing. A ratio higher than 1 shows that a country has over-borrowed compared to its debt benchmark, while a ratio lower than 1 shows that a country has under-borrowed.10 On average, Caribbean countries have borrowed twice as much as the long-term debt benchmark would suggest. However, the over-borrowing ratios vary considerably by country characteristics and depend on the specific combination of fiscal performance and growth and interest rates. The illustrative results also show that, on average, over-borrowing ratios are not necessarily higher in countries with high debt ratios (group 1) than they are in countries with somewhat smaller debt ratios (group 2).

Next, we explore the effects of the discount factor, that is, the effect of real interest and growth rates on the debt benchmarks. From a simple, comparative static point of view, a higher interest rate would lead to a lower debt benchmark, all else being equal, while a higher growth rate would lead to a higher debt benchmark.11 Of course, by taking this perspective, we ignore the effects that the interest rate or GDP growth might have on the primary balance. For example, one would expect higher GDP growth to have a positive effect on tax revenue, while certain expenditure categories might be lower, thus increasing the primary balance. Because these effects depend on each country’s economic structure, taking them into account would be beyond the scope of this chapter. Instead, we will keep the primary balance constant and just vary interest and growth rates.

In a first step, we analyze by how much the debt benchmark in each country changes if in turn the interest rate is raised by 1 percentage point or the average real GDP growth rate increases by half a percentage point. Secondly, we explore how a negative shock would affect the debt benchmark, by shocking both the real interest rate and the real GDP growth rate. To take into account the volatility of each country, the interest rate is increased by one standard deviation, while the GDP growth rate is decreased by one standard deviation.

The sensitivity analysis of the long-term debt benchmarks is displayed in Table 6.3, columns 6–8. On average, debt benchmarks decrease by 13 percentage points of GDP for a 1 percentage point increase in real interest rates, while a higher average GDP growth of half a percentage point increases the debt benchmark by 18 percentage points of GDP. For the interest rate increase, the average debt benchmark falls to 27.5 percent of GDP, down from 40.3 percent of GDP, which shows the high sensitivity of the debt benchmark to interest rate changes. This also drives up over-borrowing ratios. On the other hand, higher growth has a beneficial impact on debt benchmarks, bringing the average up to about 57.9 percent of GDP.

A negative shock, raising interest rates and lowering average growth rates at the same time, leads to considerable lower debt benchmarks compared to the baseline case (see last column in Table 6.3). This negative shock paints a particularly dire picture as the Caribbean average drops to 12.3 percent of GDP; depending on country characteristics the debt benchmark may even fall below 10 percent of GDP. This shows that an increase in interest rates coupled with subdued GDP growth would pose a high risk for sustainability in Caribbean countries.

Natural Debt Limits

The second debt limit we focus on is the concept of a natural debt limit as described earlier (see equation 6.3).12 This takes into account the volatility of revenues and expenditures, as well as the volatility of growth and interest rates. The average natural debt limit for the Caribbean is 30 percent of GDP (Table 6.4). Thus, taking into account the volatility of fiscal and macroeconomic variables reduces the debt levels that could be sustained. Natural debt limits vary across different country characteristics, but overall they are below 100 percent of GDP. On average, higher revenue variability leads to a lower natural debt limit. For example, country E in group 1 has a very high coefficient of variation of revenues (see Table 6.2), while its natural debt limit is only about 14.3 percent of GDP.

Table 6.4

Measures of Debt Sustainability: Illustrative Natural Debt Limits

(In percent of GDP)

article image
Source: Authors’ calculations.

As for long-term debt benchmarks, we can calculate over- and under-borrowing ratios. For the Caribbean average, the over-borrowing ratio increases to almost 3, that is, the average actual debt ratio is three times higher than the suggested natural debt limit. There is a high variability of these ratios across countries. Again, some country characteristics suggest an under-borrowing (country B of group 3 and country A of group 2). There are also examples of countries with relatively low over-borrowing ratios of less than 2 (countries C and D of group 1 and country A of group 3). The highest ratio is found for country C in group 3, which has a relatively high revenue and relatively low expenditure variability.

The Caribbean average of natural debt limits is higher compared to the one for long-term debt benchmarks (by 10 percentage points of GDP), while correspondingly the over-/under-borrowing ratios are higher as well (see Table 6.5). Turning to individual countries, a comparison shows that for the majority of cases the natural debt limit is smaller than the long-term debt benchmark. This would be expected if a high volatility of revenues would restrain governments in their repayment capacity. However, there are also some cases for which the natural debt limit is higher than the long-term debt benchmark. These include countries with relatively low revenue volatility, for example country C of group 1 and country B of group 2. Some country characteristics can also lead to the case where the two debt ratios are quite close, with a difference of less than one percentage point (country E in group 1).

In a next step, we conduct a sensitivity analysis by varying the real interest rate. In particular, we show how the natural debt limits change for a low interest rate scenario (2 percentage points lower than the baseline) and a high interest rate scenario (2 percentage points higher than the baseline). On average, natural debt limits are 5 percentage points of GDP higher for the low interest rate case, while they are 3 percentage points of GDP lower for the high interest rate case. Figure 6.1 depicts the corresponding over- and under-borrowing ratios for the different interest rate scenarios. Only one case, country B in group 3, shows under-borrowing even for the high interest rate case. All other countries are more or less very susceptible to interest rate increases.

Figure 6.1
Figure 6.1

Over- and Under-Borrowing, Natural Debt Limit—Interest Rate Scenariosa

Source: Authors’ calculations.a The abbreviations correspond to the grouping from Table 6.2, which means G1A corresponds to Group 1, Country A, and so on.

An important question surfaces. The analysis of long-term debt benchmarks and natural debt limits for the Caribbean average and different country cases raises one important issue: Why is the actual average debt ratio in the Caribbean two or three times as high as the respective debt benchmark? These discrepancies might be due to factors that the debt sustainability measures are unable to capture but that are specific characteristics of the Caribbean region.

One of these factors might be related to the interest rates on government debt. We have mentioned that it is complicated to measure real interest rates in the Caribbean due to the lack of fully developed debt management and underdeveloped capital markets. This, coupled with further distortions in domestic financial markets, might shelter these governments from real capital costs and allow them to borrow more than the debt benchmarks would suggest. Indeed, the sensitivity analysis hints at the possibility that this issue exists by showing that reducing the real interest rate by 200 basis points would increase the average natural debt limit by 10 percent of GDP.

Underdeveloped capital markets also imply that domestic investors (like banks, nonbank financial institutions, and national insurance schemes) lack viable and profitable investment alternatives and therefore invest heavily in government debt instruments. In addition, foreign investment options are often limited due to capital controls or foreign investment caps. All of this would limit the power of interest rates to work as true indicators for market rates.

While the high domestic investor base might, on the one hand, lead to adverse sovereign–bank interlinkages in cases where the sovereign faces debt distress, on the other hand it might also help to sustain a higher debt ratio. Arslanalp and Tsuda (2012) show that for a set of advanced economies, countries with a high share of domestic investors (domestic banks or central banks) have a low risk of pressures from financial markets. In several Caribbean countries, domestic investors hold a large share of government debt (see Chapter 3).

The issue of low real interest rates also touches upon the previously mentioned “dynamic inefficiency” issue. If a government were sheltered from facing the true interest rate costs, it might not have needed to run fiscal surpluses. But instead, the combination of real interest rates lower than the (albeit low) average growth rates might have lead to a negative discount factor combined with a primary deficit. Normally, market mechanisms would have led over time to rising interest rates; however, distortions might have suspended this mechanism from working correctly. Thus, governments may have been able to borrow above the debt benchmarks based on the measures for real interest rates that we have used in this analysis. Indeed, Escolano, Shabunina, and Woo (2011) find that persistently negative interest rate-growth differentials in non-advanced economies are related to captive financial markets, financial repression, and lack of financial development.

However, even though a negative interest rate-growth differential should play a debt-stabilizing role, keeping debt stable even in the presence of persistent primary deficits, debt has increased strongly over the last years in the Caribbean region (see Chapter 2). This casts doubt on whether this mechanism was at play.

Exceptional Fiscal Performance

As a third measure, we look at a debt benchmark based on exceptional fiscal performance and compare it to the long-term debt benchmark as well as the natural debt limit. This measure uses the highest attained primary surplus of each country together with the baseline interest and growth rate differential. The exceptional fiscal performance debt ratio shows the highest possible debt limit a country could commit to if fiscal outcomes were particularly favorable. However, we have to stress that this is based on the fiscal outcome of one particular year, and maintaining such primary surpluses over a longer period of time or raising average primary balances to such dimensions would be a different story. In particular, Chapter 7 of this book shows that maintaining fiscal consolidation efforts in the Caribbean is not an easy endeavor.

In this case, the average debt ratio for the Caribbean is almost 300 percent. However, the adjustments from the average to the highest attained primary balance are substantial for some cases. A comparison between the Caribbean average of average primary surpluses and the highest attained primary surpluses shows a difference of 5 percentage points of GDP (see Table 6.2). For individual countries, the differences vary between 1.9 percentage points (country C of group 3) and 8 percentage points (countries A and E of group 1). Therefore, the results on exceptional fiscal performance should be seen as illustrative, showing that debt problems could be mitigated by embarking on strong fiscal consolidation.

Most Caribbean countries would not run into debt problems if they would maintain their highest ever attained fiscal surplus (see Table 6.5).13 However, in some cases, even using the exceptional fiscal performance measures indicates over-borrowing compared to actual debt ratios. Examples include country A in group 1 and country C in group 3.

Table 6.5

Measures of Debt Sustainability: Illustrative Results—Comparison of Different Measures

(In percent of GDP)

article image
Source: Authors’ calculations.

Fiscal Contingent Liabilities

The analysis of over- and under-borrowing ratios so far has only taken into account actual debt ratios as of end-2011 but has not included contingent fiscal liabilities. However, these liabilities can be high and can add a substantial amount of debt to already high debt ratios, jeopardizing fiscal sustainability. Recent examples include Antigua and Barbuda and Belize. For the latter, IMF staff estimate that gross contingent liabilities are about 17 percent of GDP (see IMF, 2011). However, since a big part of these contingent liabilities stems from nationalized companies and their valuation is surrounded by a great deal of uncertainty, the value of Belize’s contingent fiscal liabilities is also uncertain. Similarly, in Antigua and Barbuda contingent liabilities from state-owned enterprises pose fiscal risks, with the government-guaranteed debt of these enterprises amounting to 14.6 percent of GDP (see IMF, 2013a).

We check how our illustrative results would change if contingent liabilities were included. Results so far have already shown substantial amounts of over-borrowing for the Caribbean average. Including contingent liabilities would only exacerbate this situation. Table 6.6 shows an illustrative scenario for the way over- and under-borrowing ratios would increase for the hypothetical case of adding contingent liabilities of 15 percent of GDP to each country’s debt ratio. In this case, the average of the over-borrowing ratio would be about 2.4 in the case of the long-term debt benchmark (up from 2) and about 3.2 in the case of natural debt limits (up from 2.7). Once more, this scenario illustrates the high vulnerability of many Caribbean economies to any type of debt shock.

Table 6.6

Measures of Debt Sustainability: Illustrative Results—Over/Under-Borrowing and Contingent Liabilities

(In percent of GDP)

article image
Source: Authors’ calculations.Note: LTDB = long-term debt benchmark; NDL = natural debt limit; O.B. ratio = over-/under-borrowing ratio.

Policy Implications and Conclusions

In this chapter, we have discussed measures of fiscal sustainability and related the concepts to fiscal performance and debt ratios in the Caribbean. The illustrative results have revealed an over-borrowing ratio for the Caribbean average of 2, casting doubt on the sustainability of public debt in the Caribbean region. Several sensitivity analyses have shown the high vulnerability to negative shocks from interest rates and contingent liabilities, but they have also shown that higher growth and fiscal consolidation could lead the way back to sustainable paths.

There are several important implications for policy that follow from these insights. In particular, policy needs to refocus on reducing public debt to sustainable levels. Moreover, these sustainable levels are likely to be different for each country, given their varying growth prospects and the volatility of revenues and spending. Further, public debt policy needs to target a debt ratio in normal times that would not create financing difficulties in the presence of negative economic shocks. Finally, the authorities need to develop a comprehensive framework to fully account for all public debt obligations, including contingent fiscal liabilities. This is important to fully determine debt sustainability.

Appendix 6.1

Appendix Table 6.1

Measures of Debt Sustainability: Illustrative Natural Debt Limits

(In percent of GDP)

article image
Source: Authors’ calculations, based on the Mendoza-Oviedo (2009) formulation.
1

It must be emphasized that individual country circumstances vary greatly using this measure.

2

We should stress that while we take every effort to calibrate the necessary parameters carefully, the results should be taken as illustrative scenarios subject to the caveats related to the different methods. In particular, the fiscal sustainability measures are highly dependent on the underlying assumptions about growth and interest rates. Therefore, these calculations should not be seen as the ultimate debt target, but should be seen as a way to think about the high debt levels and fiscal sustainability and how it is affected by prevailing and prospective macroeconomic conditions.

3

Actually, the euro convergence criteria include a “debt criterion” that provides for a limit on gross government debt of 60 percent of GDP or, if the debt ratio is higher, it shall at least be found to have “sufficiently diminished and must be approaching the reference value at a satisfactory pace” (see European Monetary Institute, 1995).

4

See, for example, IMF (2003) for an application to emerging market and industrial economies.

5

See the discussion on debt limits in IMF (2013b).

6

For a detailed overview of these approaches, see Burnside (2004) and Tanner (2013). Additional probabilistic methods include approaches that formally model stochastic processes for the macro determinants of debt ratios. These estimates are then used to simulate corresponding debt ratios and to derive their probability distributions. For applications to selected emerging market economies see, for example, Celasun, Debrun, and Ostry (2006), Tanner and Samake (2006), and di Giovanni and Gardner (2008).

8

See Mendoza and Oviedo (2009) for a discussion.

9

An alternative way to derive the primary balance would be to adjust tax revenue and government spending as was done for the natural debt limit by Mendoza and Oviedo (2009). The results for the natural debt limit in the Mendoza and Oviedo formulation can be found in Appendix Table 6.1.

10

See IMF (2003) for an overview of over-borrowing ratios in emerging market and industrial countries.

11

This can be seen by computing the marginal derivatives of equation (6.1) with respect to the interest and growth rates, respectively.

12

For completeness, we also report the country group medians for the natural debt limit in the Mendoza-Oviedo formulation in Appendix Table 6.1. However, as it does not take into account growth and interest rate volatility, we focus our discussion on the adjusted natural debt limit.

13

In addition, we need to stress that this measure does not take into account possible positive or negative effects such as those that strong fiscal consolidation would have on growth in the short and medium term.

References

  • Aiyagari, S. Rao, and Ellen McGrattan, 1998, “The Optimum Quantity of Debt,” Journal of Monetary Economics, Vol. 42, pp. 44769.

  • Arslanalp, Serkan, and Takahiro Tsuda, 2012, “Tracking Global Demand for Advanced Economy Sovereign Debt,” IMF Working Paper 12/284 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Barnhill Jr., Theodore M., and George Kopits, 2004, “Assessing Fiscal Sustainability Under Uncertainty,” The Journal of Risk, Vol. 6, No. 4, pp. 3153.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Barro, Robert, 1979, “Optimal Debt Management,” NBER Working Paper No. 5327 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Blanchard, Olivier J., 1990, “Suggestions for a New Set of Fiscal Indicators,” OECD Working Paper No. 79 (Paris: Organization for Economic Co-operation and Development).

    • Search Google Scholar
    • Export Citation
  • Blanchard, Olivier J., Jean-Claude Chouraqui, Robert P. Hagemann, and Nicola Sartor, 1990, “The Sustainability of Fiscal Policy: New Answers to Old Questions,” OECD Economic Studies, No. 15 (Autumn), pp. 736.

    • Search Google Scholar
    • Export Citation
  • Bohn, Henning, 1998, “The Behavior of U.S. Public Debt and Deficits,” Quarterly Journal of Economics, Vol. 113, pp. 94963.

  • Burger, Philippe, Alfredo Cuevas, Ian Stuart, and Charl Jooste, 2011, “Fiscal Sustainability and the Fiscal Reaction Function for South Africa,” IMF Working Paper 11/69 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Burnside, Craig, 2004, “Assessing New Approaches to Fiscal Sustainability Analysis,” unpublished, Duke University, September 2004.

  • Celasun, Oya, Xavier Debrun, and Jonathan D. Ostry, 2006, “Primary Surplus Behavior and Risks to Fiscal Sustainability in Emerging Market Countries: A ‘Fan-Chart’ Approach,” IMF Working Paper 06/67 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chalk, Nigel, and Richard Hemming, 2000, “Assessing Fiscal Sustainability in Theory and Practice,” IMF Working Paper 00/81 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • da Costa, Carlos Eugenio, 2010, “Public Debt Sustainability,” in Public Debt: the Brazilian Experience, ed. by Anderson Caputo Silva, Lena Oliveira de Carvalho, and Otavio Ladeira de Medeiros (Brasília: National Treasury Secretariat and World Bank).

    • Search Google Scholar
    • Export Citation
  • da Costa, Carlos Eugênio, Anderson Caputo Silva, and William Baghdassarian, 2004, “Assessing Three Models for the Analysis of Debt Sustainability,” unpublished, Getúli Vargas Foundation, April 2004.

    • Search Google Scholar
    • Export Citation
  • Debrun, Xavier, and Charles Wyplosz, 1999, “Onze Gouvernements et une Banque Centrale,” Revue d’Economie Politique, Vol. 109 (May–June), pp. 387424.

    • Search Google Scholar
    • Export Citation
  • di Giovanni, Julian, and Edward Gardner, 2008, “A Simple Stochastic Approach to Debt Sustainability Applied to Lebanon,” IMF Working Paper 08/97 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Égert, Balázs, 2010, “Fiscal Policy Reaction to the Cycle in the OECD: Pro- or Countercyclical?OECD Economics Department Working Paper No. 763 (Paris: Organization for Economic Co-operation and Development).

    • Search Google Scholar
    • Export Citation
  • Escolano, Julio, Anna Shabunina, and Jaejoon Woo, 2011, “The Puzzle of Persistently Negative Interest Rate-Growth Differentials: Financial Repression or Income Catch-Up?” IMF Working Paper 11/260 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • European Monetary Institute, 1995, “Progress Towards Convergence” (November 1995). www.ecb.int/pub/pdf/othemi/emiprogresstowardsconvergence199511en.pdf.

    • Search Google Scholar
    • Export Citation
  • Gali, Jordi, and Roberto Perotti, 2003, “Fiscal Policy and Monetary Integration in Europe,” Economic Policy, Vol. 18, No. 37, pp. 53372.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ghosh, Atish R., Jun I. Kim, Enrique G. Mendoza, Jonathan D. Ostry, and Mahvash S. Qureshi, 2013, “Fiscal Fatigue, Fiscal Space and Debt Sustainability in Advanced Economies,” Economic Journal, Vol. 123, No. 566, pp. F4F30.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Greenidge, Kevin, Roland Craigwell, Chrystol Thomas, and Lisa Drakes, 2012, “Threshold Effects of Sovereign Debt: Evidence from the Caribbean,” IMF Working Paper 12/157 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), 2003, “Public Debt in Emerging Markets: Is It Too High?” Chapter 3 in World Economic Outlook, September (Washington).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF) , 2011, “Belize 2011, Article IV Consultation,” IMF Country Report No. 11/340 (Washington).

  • International Monetary Fund (IMF), 2013a, “Antigua and Barbuda, Staff Report for the 2012 Article IV Consultation, Seventh Review under the Stand-By Arrangement and Financing Assurance Review, Request for Waiver of Nonobservance of Performance Criterion and Request for Waiver of Applicability,” IMF Country Report No. 13/76 (Washington).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (IMF), 2013b, Fiscal Monitor, April (Washington).

  • Kumar, Manmohan, and Jaejoon Woo, 2010, “Public Debt and Growth,” IMF Working Paper 10/174 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Melitz, Jacques, 1997, “Some Cross-Country Evidence about Debt, Deficits and the Behaviour of Monetary and Fiscal Authorities,” CEPR Discussion Paper No. 1653 (Washington: Center for Economic Policy Research).

    • Search Google Scholar
    • Export Citation
  • Mendoza, Enrique, and Marcelo Oviedo, 2009, “Public Debt, Fiscal Solvency and Macroeconomic Uncertainty in Latin America,” Economia Mexicana, Vol. 18, No. 2, pp. 13373.

    • Search Google Scholar
    • Export Citation
  • Tanner, Evan, 2013, “Fiscal Sustainability: A 21st Century Guide for the Perplexed,” IMF Working Paper 13/89 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tanner, Evan, and Issouf Samake, 2006, “Probabilistic Sustainability of Public Debt: A Vector Autoregression Approach for Brazil, Mexico, and Turkey,” IMF Working Paper 06/295 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xu, David, and Piero Ghezzi, 2003, “From Fundamentals to Spreads: A Fair Model for High Yield Emerging Markets Sovereigns,” Deutsche Bank, July.

    • Search Google Scholar
    • Export Citation
Tackling Fiscal and Debt Challenges
  • View in gallery

    Over- and Under-Borrowing, Natural Debt Limit—Interest Rate Scenariosa