2. Medium-Term Fiscal Adjustment in an Uncertain World

International Monetary Fund. Fiscal Affairs Dept.
Published Date:
April 2013
  • ShareShare
Show Summary Details

Almost all advanced economies have implemented significant fiscal adjustment since 2010. Nevertheless, their current fiscal positions differ significantly, primarily reflecting uneven starting conditions and differences in the impact of the crisis on their fiscal accounts, rather than variations in the extent of postcrisis adjustment.

Broadly speaking, advanced economies can be classified into three groups (Figure 7, left panel). The first group comprises countries that have relatively low debt-to GDP ratios; most of them have already stabilized or even lowered them compared to 2007. The second group includes those countries where the debt ratio has recently peaked or is still rising, but remains at a fairly contained level. These countries will require further adjustment, but should be able to generate positive debt dynamics with a fairly contained fiscal effort. The third group consists of 10 countries where the debt ratio is high (above 90 percent of GDP) and rising, reflecting still-large deficits (on average about 5½ percent of GDP). It is within this group that most fiscal vulnerabilities are concentrated, and these are therefore the countries where the focus of fiscal adjustment will be in the coming years. Although these countries are few in number, they account for more than 40 percent of global output, meaning the success or failure of their efforts will have profound implications for the world economy.

Figure 7.Country Groups According to Debt Level and Trend

(Percent of GDP)

Source: IMF staff projections.

Note: Figure shows gross general government debt, except in the cases of Australia, Canada, Japan, and New Zealand, for which net debt ratios are used. For advanced economies, Group 1: Debt less than 60 percent of GDP; Group 2: Debt between 60 and 90 percent of GDP; Group 3: Debt greater than 90 percent of GDP. For emerging market economies, Group 1: Debt less than 40 percent of GDP; Group 2: Debt between 40 and 70 percent of GDP; Group 3: Debt greater than 70 percent of GDP.

Emerging market economies have, as a group, come out of the crisis in better fiscal shape than many advanced economies. As a result, their future adjustment needs are typically smaller, even if one accepts that their historically more volatile financial environment suggests more prudent benchmark debt levels than those used for advanced economies (Figure 7, right panel). Although only three emerging market economies fall in the high-debt group (debt ratios in excess of 70 percent of GDP), those emerging market economies in the middle category (i.e., those with still-rising but fairly contained debt ratios) might still have relatively large adjustment needs because of their high deficits.

This section takes a fresh look at the medium-term fiscal adjustment needs in advanced and emerging market economies, underscoring the high current uncertainty, particularly in those advanced economies in which public debt has reached its highest level since the immediate post–World War II period, when the outlook was in many respects more supportive of fiscal adjustment than now:

cuts in military spending provided an easy way to consolidate, labor force and output growth prospects were strong (in contrast, the labor force is now projected to decline in many advanced economies), and interest rate caps and restrictions on bank assets kept sovereign borrowing costs relatively low. The current environment is much less friendly and carries potentially high, although uncertain, risks. There are three major sources of uncertainty:

  • The debt consolidation target: Should debt ratios just be stabilized at their currently historically high levels or should they be brought down (and by how much)? Should strategies target a specific debt level, or would it be preferable to target a specific fiscal balance (for example, a balanced budget)?

  • The projected interest rate–growth differential: For a given debt stock, higher interest rates mean that a larger share of public resources needs to go toward paying interest, leaving fewer resources to pay down the debt. In contrast, faster growth brings down debt ratios more quickly, by increasing the denominator in the debt-to-GDP ratio and by making it easier to run larger primary balances. But with the crisis, the dispersion of interest rate–growth differentials across countries has increased. Not all countries are recovering at the same speed, and while the interest rate has risen sharply in countries under market pressure, it has fallen in countries benefiting from safe-haven flows. Predicting the future path of the interest rate–growth differential is thus not easy. The incidence of cyclical versus structural factors in accounting for the decline in output after 2007 remains unknown, resulting in large revisions of potential growth projections in advanced economies while bond yields have fluctuated widely.

  • The feasibility of implementing large, sustained fiscal adjustment: An increase in the primary balance can bring the debt ratio down and avert a painful debt restructuring or monetization of an otherwise unsustainable debt. But what constitutes a politically and socially acceptable pace of fiscal adjustment, and for how long can large primary surpluses be maintained before pressures to raise spending or reduce taxes become overwhelming? Overall, the empirical evidence suggests that in some countries, either the size of the improvement in the primary fiscal balance needed to bring the debt ratio to a more sustainable level, or the period over which such an improvement would need to be maintained, would be unprecedented.6 This does not mean that the task is impossible, but it does underscore the need to use many levers to facilitate the adjustment. Keeping interest rates low over an extended period and boosting potential growth will be key to successful debt reduction efforts. There is evidence that markets are forward looking, attaching importance not just to the level of the debt but also to the direction in which it is moving (see the October 2012 World Economic Outlook), suggesting that once investors are confident that the debt ratio is safely on a downward path, a virtuous cycle of lower interest rates and higher growth can be triggered. But for this to occur, the credibility of the fiscal adjustment path is critical.

The costs and risks of high debt

What should be the ultimate objective of fiscal adjustment? Stabilizing the public debt ratio has intuitive appeal, as it means the government will be able to finance its operations and remain solvent over time.7 However, there are many reasons why merely stabilizing public debt at high levels would not be optimal. A large body of research, summarized in previous issues of the Fiscal Monitor, concludes that high public debt leads to higher interest rates and slower growth (Table 8).8 Most studies find that high debt levels (above 80–90 percent of GDP) have a negative effect on growth (some 0.15–0.20 percent per 10 percentage points of GDP). High debt also makes public finances more vulnerable to future shocks, because it constrains governments’ ability to engage in countercyclical policies and because the larger the initial debt ratio, the bigger the increase in the primary surplus required to stabilize that ratio after an adverse shock to growth or interest rates. Indeed, when debt is high, there is a risk of falling into a bad equilibrium caused by self-fulfilling expectations (high debt is unsustainable because markets believe it is so and set interest rates accordingly).

Table 8.Recent Empirical Evidence on the Impact of Debt on Interest Rates and Growth
Debt and Long-Term Interest RatesDebt and Growth
StudySampleEffect of a 1 percentage point increase in public debt–to–GDP ratio on long-term interest ratesStudySampleEffect of a 10 percentage point increase in public debt–to–GDP ratio (beyond threshold) on average annual growth
Engen and Hubbard (2004)United States, 1976–20032–3 basis points (projected debt-to-GDP ratio on real rates)Kumar and Woo (2010)38 advanced and emerging market economies, 1970–2007–0.17 percentage point (beyond 90 percent)
Kinoshita (2006)19 OECD countries, 1971–20032–5 basis points (current debt-to-GDP ratio on real rates)Caner, Grennes, and Koehler-Geib (2010)79 advanced and developing economies, 1980–2008–0.17 percentage point (beyond 77 percent)
Ardagna, Caselli, and Lane (2007)16 OECD countries, 1960–2002Up to 3.8 basis points (for nonlinear estimation, at maximum value of debt-to-GDP ratio in the sample)Cecchetti, Mohanty, and Zampolli (2011)18 OECD economies, 1980–2006–0.13 percentage point (beyond 84 percent)
Laubach (2009)United States, 1976–2006 (semiannual)2–4 basis points (projected debt-to-GDP ratio on real rates)Baum, Checherita-Westphal, and Rother (2012)12 euro area economies, 1990–2010–0.59 percentage point (beyond 96 percent)
Baldacci and Kumar (2010)31 advanced and emerging market economies, 1980–20083–5 basis points (current debt-to-GDP ratio on nominal and on real rates)
Alper and Forni (2011)53 advanced and emerging market economies, 2002–101–7 basis points (expected debt-to-GDP ratio on real rates, nonlinear for emerging market economies)
Poghosyan (2012)22 advanced economies, 1980–20102 basis points (in the long run, current debt-to-GDP ratio on real rates)
Jaramillo and Weber (2012)15 emerging market economies, 2007–11 (monthly)4 basis points (projected debt-to-GDP ratio on nominal rates)6 basis points (if high global risk aversion)
Note: OECD: Organisation for Economic Co-operation and Development.
Note: OECD: Organisation for Economic Co-operation and Development.

The ease with which the surge in public debt ratios has been financed in most countries may suggest that the risks arising from high debt levels are overstated. As noted, the lower is the interest rate–growth differential, the higher is the amount of debt that can be sustained over time. After spiking in 2009, that differential has declined in most advanced economies and remains below the precrisis average in spite of higher debt. The failure of market interest rates to respond to rising sovereign indebtedness (except in some euro area countries) could be taken to suggest that many advanced economies have little to fear from high public debt. There are reasons to believe that this trend will not persist, however, and that high debt will expose countries to larger risks in the future.

The exceptionally low borrowing costs enjoyed by some high-debt countries reflect, in addition to still-weak economic activity, the influence of institutional investors—pension, mutual, and insurance funds—as well as national and foreign central banks (Box 3).9 The importance of some of the factors that have helped insulate many countries from debt-related vulnerabilities may gradually diminish, and borrowing costs could increase or become more volatile as a result:

  • First, the capacity of domestic investors to absorb public debt is likely to decline over time for some countries. For example, the aging of the population is expected to reduce savings in Japan, curbing the growth of nonbank financial institutions. Chinese central bank holdings of U.S. Treasury bonds may also decline owing to diversification away from U.S. dollars or as a consequence of smaller current account surpluses.

  • Central bank purchases of government debt have continued in the largest advanced economies in 2012. But as market conditions in advanced economies normalize and demand for base money declines, domestic central banks, to avoid inflationary pressures, may choose to unwind their asset purchases undertaken for purposes of monetary policy. As a result, the share of public debt they hold would decline.

  • In many emerging market economies, debt dynamics are benefiting from spillovers from accommodative monetary policies abroad, as well as from a combination of regulatory constraints and the relative shallowness of domestic financial markets. These factors are likely to decline in importance as monetary policies normalize in advanced economies and as domestic financial intermediation deepens, easing financial repression.

  • The gap between market and concessional rates in emerging market economies is about 4½ percentage points. The share of official lending provided to these countries is already declining and is likely to continue doing so, pushing up their interest rate–growth differentials as official financing is replaced with more expensive market borrowing. In addition, fiscal risks are affected not only by what is already in the government’s balance sheet but also by what could potentially be there.10 In other words, looking only at current debt ratios may result in understatement of the fiscal risks a country faces because it does not take account of contingent liabilities. Explicit government guarantees for a representative sample of advanced economies are estimated at 2½ percent of GDP–with some variations across countries–mostly related to public enterprises. Implicit guarantees could be far larger, as preliminary IMF staff estimates put the outstanding debt of these enterprises at about 16½ percent of GDP on average (Figure 8).11 Of course, contingent liabilities are not exclusive to advanced economies, as implicit and explicit guarantees—for example, related to subnational governments and the financial sector—can also be found in emerging market economies (Box 4).12

Figure 8.Net Consolidated Government and Central Bank Debt, Outstanding Government-Guaranteed Bonds, and Debt of Government-Related Enterprises

(Percent of GDP)

Sources: Dealogic; and IMF staff estimates.

Note: Amounts for some countries are likely to be underestimated given data constraints. See Statistical Table 15 for details on net consolidated government and central bank debt.

1 Outstanding government-guaranteed bonds correspond to bonds that are issued by private and public banks and financial institutions and carry state guarantees. Short-term debt is not included.

2 Bonds issued by government-owned or government-related institutions; includes both financial and nonfinancial institutions, subject to data availability. For the United States, includes mortgage-backed securities and other guarantees of government-sponsored enterprises.

In practice, about one-fifth of the largest unexpected increases in general government gross debt during 2007–10 were due to government support to the financial sector and hidden or implicit obligations to public corporations and public-private partnerships outside the general government perimeter.13 Experience suggests that countries with large or quickly expanding financial sectors or with sizable state-owned enterprises may find that current debt and deficit levels are an imperfect indicator of risks to their fiscal positions. For example, Iceland and Ireland saw their government debt ratios increase by 60–70 percent of GDP, despite seemingly safe precrisis budget positions, as a result of outsized financial sectors that eventually needed massive public support.

In sum, high debt significantly increases a country’s fiscal vulnerabilities. These vulnerabilities may rise in the future as the result of changes in the investor base and other structural factors. And the risks in some countries may already be higher than they appear because of contingent liabilities that are not recorded in debt statistics. When the certainty of medium-term spending pressures associated with population aging (see Clements and others, 2013; IMF, 2010b; and Statistical Tables 12a and 12b) is added to this, it makes a compelling case for high-debt countries to continue the process of gradual but sustained deficit reduction that began in 2010, or to get onto that path without delay, aiming not just at stabilizing the debt ratio but at reducing it.

The magnitude of the required adjustment

By how much should public debt be lowered, and over what time frame, and what would it take in terms of spending cuts or tax increases—that is, improvement in the primary balance—to lower it? As noted above, there is no straightforward answer to these questions.

First, although the economics literature can provide guidance about the costs associated with high debt, it is less helpful in identifying what an ideal debt ratio would be.14 Empirical studies yield a wide range of debt objectives depending on the approach used to identify them. In practice, many countries have adopted debt ceilings in their fiscal responsibility laws or in the context of supernational agreements (for example, the EU’s Economic and Monetary Union and the Eastern Caribbean Currency Union) that are close to or below 60 percent of GDP (IMF, 2011a), although these levels are usually motivated as being prudent rather than optimal. The standard Fiscal Monitor long-term adjustment needs scenario has used benchmark debt ratios of 60 percent of GDP for advanced economies and 40 percent of GDP for emerging market economies, in both cases close to the precrisis median for these country groups.15 But the appropriate debt target need not be the same for all countries. If the investor base is such as to allow countries to finance themselves at low rates, it will be easier for them to sustain a higher debt level. Volatility in the interest rate–growth differential is also important: because high public debt and high volatility in growth and interest rates may be a particularly toxic combination, countries subject to relatively large shocks to growth and interest rates may want to be conservative in choosing debt targets. In addition, contingent liabilities have proven very important for certain countries, although they are not typically embedded explicitly in debt benchmarks. The implication is that to the extent that policies can diminish the degree of uncertainty, they can also allow countries to target a higher level of debt.

Second, once a long-term debt target has been identified, the required pace of primary adjustment will still depend on the length of the adjustment period and on interest rate–growth differential assumptions. Countries with high debt and a high interest rate–growth differential may prefer more front-loaded adjustment over a shorter period. But here again, the credibility of the adjustment process is critical: as noted, markets are likely to be more tolerant of high debt levels if they are convinced that those levels are being put on a downward path, resulting in a lower interest rate–growth differential and allowing a longer adjustment duration, and in both cases lowering primary adjustment targets. Box 5 illustrates how different assumptions can result in a wide range of estimated adjustment needs in the case of the United States.

Given these uncertainties, this Fiscal Monitor presents not only a baseline medium-term fiscal adjustment scenario, but also alternative scenarios based on different debt targets and interest rate–growth differential assumptions. As in the past, the baseline scenario targets a reduction in the debt ratio to 60 percent by 2030 (40 percent for emerging market economies), with the primary balance rising to the required level by 2020 (the “primary surplus benchmark”) and then remaining at that level for another decade. The projected interest rate–growth differential remains at relatively low levels until 2018, in line with the World Economic Outlook forecast of a slow recovery, and then follows a model-based simulation reflecting the assumed normalization of monetary policy, in which differences in interest rate–growth differentials across countries reflect variations in debt levels and their projected paths (see Statistical Tables 12a and 12b).16 Alternative scenarios gauge the magnitude of the adjustment challenges in advanced economies under different hypotheses:17 using a debt target of 80 percent rather than 60 percent, adopting a balanced-budget target rather than a debt target, and employing sensitivity analyses in which the interest rate–growth differential is 100 basis points higher or lower than under the baseline. For emerging market economies, the interest rate–growth differential is assumed to converge gradually from its current (in most cases negative) level to 1, the average estimate of the future interest rate–growth differential.18

Figure 9.CAPB in 2020–30 and Required Adjustment Needs, 2013–30, across Different Scenarios

(Percent of GDP)

Source: IMF staff estimates and projections.

Note: For selected advanced economies (seven scenarios for each country), the figure plots the average 2020–30 cyclically adjusted primary balance (CAPB) against the residual adjustment need under various scenarios. Triangles illustrate scenarios with a benchmark level of debt of 60 percent by 2030, circles illustrate scenarios with a benchmark level of debt of 80 percent by 2030, and squares illustrate scenarios with no indicative debt benchmark by 2030, but in which countries reach at most overall budget balance. Baseline interest rate–growth differential assumptions are shown in blue; risk scenarios with interest rate–growth differentials 100 basis points above the baseline are shown in red; and favorable interest rate–growth differentials that are 100 basis points below the baseline are shown in green. Thus, the blue triangles correspond to the numbers reported in Statistical Table 13a.

Statistical Tables 13a and 13b update the baseline adjustment needs. Despite steady consolidation in advanced economies over 2011–13, the average additional adjustment required to meet primary surplus benchmarks by 2020 is still substantial, at 5 percent of GDP.19 With many emerging market economies having paused adjustment efforts, and with their borrowing costs relatively low, their residual adjustment needs (i.e., the required improvement in the cyclically adjusted primary balance between 2013 and 2030) are broadly unchanged from October 2012 estimates, and quite small.

Figure 9 presents the results of the various scenarios for the group of advanced economies with the largest adjustment needs. In most of these countries, residual adjustment needs differ significantly under alternative scenarios (for most, the range of residual adjustment needs equals 2–3 percentage points of GDP across scenarios), but some clear differences across the countries also emerge.

  • Three countries (Belgium, France, and Italy) have already undertaken a large share of the adjustment needed to bring their debt ratios down to safer levels over time and, assuming the 2013 projections materialize, would need to increase their cyclically adjusted primary balances only by relatively small amounts (between 1 and 3 percentage points of GDP). For Italy, the scenarios suggest that little or no further adjustment is required. However, owing to its higher debt level, Italy would need to maintain much larger primary surpluses than France or Belgium over the next 10 years.

  • A second group—comprising Ireland, Portugal, Spain, the United Kingdom, and the United States—still has some way to go in terms of the residual adjustment (close to 5½ percent of GDP, unweighted average). Once this consolidation is achieved, all these countries would need to maintain large primary surpluses over the medium term. In the absence of entitlement reforms, projected increases in age-related spending mean that additional measures will still be needed over time, however, to keep the primary surplus constant, particularly in the United States.

  • The largest consolidation requirement is in Japan (close to 16 percentage points of GDP, in order to reach a primary surplus of about 7 percent of GDP). Japan’s large residual adjustment need reflects both its very high public debt ratio and the fact that its cyclically adjusted primary deficit is still very large (in part because of the impact of natural disasters) (Figure 10). Clearly, this implies that a longer time horizon will be required to bring public debt down to the scenario levels.

Figure 10.Advanced Economies with Largest Adjustment Needs: Required Changes in the Cyclically Adjusted Primary Balance

(Percent of GDP)

Source: IMF staff estimates and projections.

Note: Figure shows the advanced economies with the 10 largest illustrative adjustment needs between 2011 and 2030, based on the Fiscal Monitor baseline scenario. The red bars show the adjustment expected to take place between 2011 and 2013. For details, see “Data and Conventions” in the Methodological and Statistical Appendix.

Daunting as these adjustment needs are, in many cases they are little different from those that would be required to balance countries’ budgets in cyclically adjusted terms, something many observers endorse as an appropriate medium-term policy objective. Indeed, balancing the budget would put half of the high-debt cases (Belgium, France, Spain, the United Kingdom, and the United States) within close distance of the benchmark debt ratio by 2030. Although merely balancing the budget would leave debt ratios above 60 percent of GDP by 2030 in other high-debt countries, it is reasonable to expect that markets would easily tolerate the higher—but still steadily declining—debt ratio that would follow from such a policy. A commitment to maintain balanced budgets might also be more palatable politically than one that involves larger and larger headline budget surpluses over time, as would emerge if primary balances were constant and interest payments declined in line with debt.

In most emerging market economies, the benchmark primary surpluses under the baseline illustrative scenario are significantly lower than among advanced economies. However, in Egypt, Hungary, and Jordan, the target exceeds 3 percent of GDP, reflecting high debt (80 percent of GDP or more), compounded by the projected gradual increase in interest rates in the context of a normalization of the economic environment. In Egypt and Jordan, more than 5 percent of GDP in adjustment will be needed to achieve the benchmark primary surpluses. India has large adjustment needs too (6¾ percent of GDP), but it does not have to maintain as high a target cyclically adjusted primary balance, partly thanks to a very favorable interest rate–growth differential.

The adjustment needed to achieve debt-stabilizing primary balances is relatively small in low-income countries, given a negative interest rate–growth differential and low levels of debt (see Box 6). In many sub-Saharan African countries, the primary balance gap, or the difference between the projected primary balance and the primary balance that would stabilize debt at its current level, is relatively small, the major exception being in some fragile states (Figure 11). In countries with small primary gaps but high debt ratios, an additional consolidation effort aimed at reducing rather than just stabilizing debt should be considered. Countries with primary surpluses that are currently higher than those required to stabilize debt may want to lock in surpluses by rebuilding their liquid-asset buffers.

Figure 11.Sub-Saharan Africa: Average Primary Balance Gap, 2012–17

(Percent of GDP)

Source: IMF staff estimates and projections.

Note: The interest rate–growth differential used to calculate the debt-stabilizing primary balance is based on medium-term projections (2013–17 average) for the nominal interest rate on public debt (from each country’s debt sustainability analysis) and nominal GDP growth rate. A positive primary balance gap indicates a tendency of the debt-to-GDP ratio to increase over time unless fiscal policies are tightened. The actual need for tightening will be country specific and will depend on, among other factors, the initial debt level.

Historical evidence on fiscal adjustments

The fiscal effort required to lower debt ratios to more prudent levels remains substantial by any metric for the 10 advanced economies in which debt is high and still rising. It is thus natural to wonder about possible historical precedents. History is not destiny: several countries among the 10 may not have run large primary surpluses in the past because they did not need to, as their debt was much lower. This does not mean that they will not be able to run large primary surpluses in the future. Nevertheless, a look at historical precedents can illustrate the scale of the present challenge. The following analysis is based on a new fiscal balances historical database covering 55 advanced and emerging market economies and developing countries dating back to 1800 in some cases.20

Historical evidence suggests that high primary surpluses may be easier to achieve than to maintain. Large sustained postwar debt reductions have typically involved a combination of high primary surpluses and other policies.21 Since the 1950s, the distribution of the maximum annual primary surplus shows a median of about 6½ percent of GDP for advanced economies and 6¼ percent for emerging market economies, albeit with a greater dispersion for the latter group.22 Using 5-year moving averages, the median falls to 3½–4 percent of GDP; it declines steadily as the length of the moving average window increases, to only 2½–3¼ percent of GDP over a 10-year period.23 One possible explanation is that, as credibility of adjustment efforts is established, interest rates fall and there is no longer a need to maintain as high a primary surplus in the medium to long term.

The largest improvement in the primary balance achieved by an advanced economy, equaling more than 10 percent of GDP, took place shortly after World War II (Table 9).24 In contrast, among emerging market economies, the three largest episodes (2–6 percentage points of GDP) occurred in the 1990s. For both groups, these consolidation episodes took place against the backdrop of large initial deficit positions. In all but one case, the resulting end-level primary surpluses were still below the maximum sustained primary surplus of 3½–4 percent of GDP identified above.

Table 9.Summary Statistics for Three Largest Consolidation Episodes(Percent of GDP)
YearPrimary BalanceΔDebtΔYearPrimary BalanceΔDebtΔYearPrimary BalanceΔDebtΔ
Advanced economies19522.810.458.85.319843.15.742.50.119892.94.458.40.2
Emerging market economies19913.16.744.67.819962.23.739.41.019880.72.431.80.0
Source: Mauro and others (2013).Notes: Table provides information on the three largest changes in the two-year moving-average primary balance over a six-year period, with at least five years elapsing between consolidations. Median of the first-, second-, or third-largest consolidation episode, respectively, for each country grouping is shown in each panel.
Source: Mauro and others (2013).Notes: Table provides information on the three largest changes in the two-year moving-average primary balance over a six-year period, with at least five years elapsing between consolidations. Median of the first-, second-, or third-largest consolidation episode, respectively, for each country grouping is shown in each panel.

Growth appears to have been an important element for achieving high headline primary surpluses.25 In particular, maximum primary surpluses are lower for advanced economies once the sample is restricted to those episodes in which growth was below trend, and lower still for those in which growth was negative (Figure 12). In the latter circumstances, the maximum sustained primary balance was just ½ percent of GDP for advanced economies and 1¼ percent of GDP for emerging market economies. The susceptibility of fiscal outturns to growth shocks may therefore also help explain why it is difficult for countries to maintain large primary surpluses for long periods. Although the data do not allow measurement of the impact of business cycles, the results are in line with evidence from Mauro and others (2013) that points to a lower response of the headline primary balance to debt in the face of negative growth surprises, even after the output gap is controlled for.

Figure 12.Maximum Primary Balance and Growth, 1950–2011

(Percent of GDP)

Sources: Mauro and others (2013); and IMF staff estimates.

Note: For each country grouping in each of the three conditions characterizing growth, the figure shows the median of the maximum five-year moving average of the primary balances. “Unconstrained” takes into account all the years in the sample. “Below-trend growth” refers to years in which the five-year moving average of the real GDP growth rate fell below the trend growth rate. “Negative growth” refers to years in which real GDP growth was negative.

An event study undertaken in connection with the database research looks in more detail into the circumstances characterizing large and sustained improvements in primary balances. Some 22 episodes (12 among advanced economies and 10 among emerging market economies) were identified as falling above the median of the distribution of maximum five-year rolling-average primary balances. The sample can be further divided into those cases that fall between the 50th and 75th percentiles (large improvements) and those at the 75th percentile or above (extraordinary improvements).

The adjustment strategy and magnitude of debt reduction varied across advanced and emerging market economies, but there were some common elements too:

  • In advanced economies, more than 75 percent of the improvement in primary balances was driven by a reduction in primary expenditure as a share of GDP; the situation was more diverse among emerging market economies.

  • Debt reduction was much larger for some emerging market economies, above 60 percent of GDP, compared to about 15 percent for advanced economies, but this was in some cases associated with debt restructuring (for example, in Argentina and Bulgaria).26

  • Across advanced and emerging market economies, extraordinary improvers typically did better not because of one exceptional year, but rather because of an extended period of larger annual increases in their primary balances (Figure 13). Whereas large improvers gave back gains in their balances relatively quickly, possibly because cyclical effects played an important role in their good performance, extraordinary improvers managed to preserve some of their gains, sustaining a net increase in their primary balances of about 2–3 percentage points 10 years after beginning their adjustment efforts.

  • Both country groups benefited from high growth rates (up to 6 percent during the event study periods).

Figure 13.Event Study of the Maximum Sustained Primary Surplus, 1950–2011

(Percent of GDP, except where otherwise indicated)

Sources: Mauro and others (2013); and IMF staff estimates.

Note: An event is defined as the five-year window in which a country achieves its maximum sustained primary surplus, based on five-year moving averages. Year t is the center of the event window, which includes years t – 2 to t + 2. The red line corresponds to countries for which the maximum sustained surplus is between the 50th and 75th percentiles of the distribution and the yellow line to countries with maximum sustained surpluses that equal at least the 75th percentile of the distribution.

The event study also suggests that large adjustments were based on improvements in the primary balances that took place over six to eight years. A natural question to ask is to what extent current adjustment needs in high-debt advanced economies would be consistent with what is observed in the historical evidence. To address this question, illustrative scenarios are presented for a sample of six representative countries. The historical database is used to generate medians of the maximum historical primary balances over windows varying in length from 10 to 30 years. These medians are then juxtaposed against the average cyclically adjusted primary balance that each country would have to maintain in order to bring its debt ratio down to 60 percent of GDP over the corresponding time frame. Other things being equal, the longer the period over which the debt ratio is to decline to the target, the smaller the primary surplus that would need to be maintained over this period. Three alternative scenarios are considered, each corresponding to different assumptions about the interest rate–growth differential.27Figure 14 shows that with an interest rate–growth differential of 1, most of the high-debt advanced economies could achieve the benchmark target while maintaining primary surpluses consistent with previous historical maximums, but only if they choose a very long horizon (of the order of about 30 years) over which to achieve the debt target. Shorter horizons would demand primary surpluses that are larger than those that have been maintained by any advanced economy in the past over the relevant period. To the extent that policies can contribute to lower interest rate–growth differentials—for example, by boosting credibility and growth—the required adjustment would become more consistent with past experience, thereby allowing debt ratios to converge to prudent levels earlier on. If, on the other hand, plans are not credible, and as a result the interest rate–growth differential increases above 1, most countries would not be able to achieve the 60 percent debt ratio even in 30 years without fiscal efforts that would be without historical precedent.

Figure 14.Feasible Adjustment Paths over 20 Years

(Percent of GDP)

Source: IMF staff estimates.

Note: The blue line in each panel denotes the maximum moving average of the primary balances over different time horizons (n years), calculated using the Mauro and others (2013) data set. The yellow (green, red) line in each panel denotes the average cyclically adjusted primary balance needed for the country to achieve its debt target—80 percent of GDP for Japan, 60 percent of GDP for all other countries—within each time horizon, given an interest rate–growth differential of 1 (0, 2). CAPB: cyclically adjusted primary balance.

Can other policies ease the fiscal adjustment process?

Given the size of the challenge facing fiscal policymakers, can other policies (alternative to improving the primary balance) facilitate their task?

Inflating the debt away

Higher inflation could help reduce public debt through three main channels. First, governments can capture real resources by base money creation (seigniorage). Second, inflation can erode the real value of the debt. The impact of this channel will depend on the maturity structure and currency denomination of the debt, as well as on the interest rate response to higher inflation, with inflation having the largest impact on long-term, fixed-rate, and local-currency-denominated debt: short-term debt and maturing long-term debt will need to be refinanced at higher interest rates, floating-rate debt will adjust automatically to higher rates, and the local-currency value of foreign-currency-denominated debt will rise as a result of the currency depreciation that will accompany higher inflation. Third, inflation can affect the primary balance, including if brackets are not indexed under a progressive income tax. Akitoby, Komatsuzaki, and Binder (2013) simulate the effect of the first two channels for Group of Seven (G-7) countries. Given the relatively low levels of base money in most advanced economies, seigniorage from higher inflation would play only a limited role in bringing down debt ratios: raising inflation from World Economic Outlook baseline projections to 6 percent for five years would generate cumulative seigniorage revenue of about 2½ percentage points of GDP. The debt erosion channel could have a stronger impact.

The same increase in inflation, under assumptions of a constant debt maturity structure, no impact of inflation on economic growth, and a one-for-one adjustment to inflation of nominal interest rates on newly issued debt (full Fisher effect), would reduce the average net debt–to–GDP ratio by less than 10 percentage points by the end of the period for most countries (other than Japan and Italy, where the effect would be larger) (Figure 15).28 The erosion effect would drop rapidly after five years, because an increasingly large share of securities would have been issued at higher interest rates, including to replace maturing debt that had been issued at lower rates.

Figure 15.Impact of Inflation on Net Debt Reduction, 2017

(Percent of GDP)

Sources: Organisation for Economic Co-operation and Development; and IMF staff estimates.

Note: The scenario depicted in the figure implies an increase in inflation by 4.4 percentage points over the projected average inflation of 1.6 percent.

Thus, although higher inflation could have some effect on debt stocks, it could hardly solve the debt problem on its own and would raise significant challenges and risks. As a practical matter, it might be difficult to lift inflation to a meaningful level in the current economic environment, as evidenced by Japan’s experience in the last decades, and in any case, countries in a monetary union would not be able to use this tool on their own. More importantly, reliance on inflation to erode debt could lead to fiscal dominance,29 with inflation rates drifting even higher as confidence in the future value of money is lost. As a result, inflation expectations could be unanchored, and the costs of bringing inflation down later could become even more onerous. This would undermine the credibility of the framework built over the past three decades to control inflation. Real interest rates could also rise as the result of unanchoring of inflation expectations, which would exacerbate the debt trajectory. Introducing some form of financial repression could keep interest rates low, but such policies may be difficult to enforce in a complex financial environment and could cause additional collateral damage to the economy. Altogether, the output costs of restoring inflation to more moderate levels would be substantial, based on the experience in the advanced economies in the 1980s (see the October 2012 World Economic Outlook). And inflation would have a highly regressive impact on incomes: although higher inflation would be a tax on bondholders, it would also disproportionately affect lower-income households, which tend to have more limited access to indexed assets.

Restructuring debt

Another option to reduce debt is to restructure it or, in the extreme, to default on it.30 Debt restructuring has been almost unknown in advanced economies in the postwar period. There have been episodes among emerging market economies, but these experiences may not be entirely relevant for the typical advanced economy case. Domestic residents hold a large share of government debt in most advanced economies (see Statistical Tables 12a and 12b), whereas in many instances of debt restructuring among emerging markets, foreign holdings were prevalent. This matters because restructuring is essentially a one-time tax on bondholders. If the tax is imposed on foreign residents, it leads to an increase in the country’s net wealth, whereas if it is imposed on domestic agents, it is equivalent to a large and sudden fiscal tightening. Even when the tax falls on foreign residents, its feedback effect can be large when foreign residents are highly integrated with the defaulting country, as suggested by the recent experience in the euro area following the Greek debt restructuring.

Restructuring is a costly and risky option. In addition to the adverse effects related to fiscal tightening and spillover, it mars a country’s reputation as a reliable borrower, which would be particularly damaging for first-time defaulters. Past debt restructuring may weigh on a country’s borrowing costs and access to international financial markets.31 Sovereign default can result in substantial short-term output losses as well as negative spillovers to other countries (Das and others, 2012). Finally, sovereign debt restructuring can affect the financial sector via two channels: on the asset side, banks will take a hit from the loss of restructured assets, and on the liability side, they may experience deposit withdrawals and the disruption of credit lines (see, for example, Borensztein and Panizza, 2009, and Das, Papaioannou, and Trebesch, 2012).

This said, there may be circumstances in which debt restructuring is unavoidable. If the primary surplus needed to make the fiscal path sustainable is too large (in terms either of adjustment or of the level that would have to be maintained over time) to be sustained by an economy without unbearable economic costs, debt restructuring would become inevitable. A debt restructuring could deal with the issue of high public debt quickly and may thus be appealing to those who do not believe that gradual fiscal adjustment is possible from a political economy perspective (and may be forced by markets if they regard orthodox adjustment impossible or unpalatable). Nevertheless, further work is needed to assess whether, and under what circumstances, debt restructuring can lead to more sustainable debt ratios.32 Moreover, the way public debt is lowered—through orthodox fiscal adjustment or by not paying creditors—could lead to reputational effects that might have a negative impact on investment and growth. Altogether, although debt restructuring is sometimes inevitable, it is never an easy way out for countries looking for a solution to their fiscal problems.

Figure 16.Key Components of Nonfinancial Assets, 2011

(Percent of GDP)

Sources: Organisation for Economic Co-operation and Development; IMF, Government Finance Statistics; and national authorities.

Note: Based on latest available data. Not all countries report all components. For New Zealand, breakdown for total nonfinancial assets is not available. For Italy and Korea, data are from national sources. For Italy, data refer to 2004.

Managing government assets

Selling government assets is a more benign option to lower gross debt ratios, although the revenue loss arising from the sale of those assets would also have to be considered to assess the longer-term impact on public finances. Public financial assets are quite large in advanced economies (more than 40 percentage points of GDP on average, half of that in the form of shares and equities) and some could potentially be disposed of. Privatization has yielded substantial proceeds in the past (see the September 2011 Fiscal Monitor) and may also help boost overall productivity if assets are managed more efficiently in private than in public hands. In many advanced economies, however, core public corporations have been privatized over the past two decades and remaining equity holdings are often in the hands of subnational governments. In Germany, for example, the general government owns about 11 percent of GDP in shares and equity and about one-tenth of this is in the hands of the federal government. Similarly, shares and equities of the Italian general government amount to about 8 percent of GDP, and it is estimated that the sale of the central government’s shares quoted on the stock market could yield about 1 percent of GDP (IntesaSanPaolo, 2012). The dispersed ownership could complicate further the use of assets for debt management purposes, including because subnational governments with limited debt may have little motivation to privatize their assets to reduce debt owed by the central government. That said, with regional and local public debt being sizable in some countries, privatization of their assets may be helpful.

Public holdings of nonfinancial assets are larger than those of financial assets, but drawing resources from them may be more difficult. In 32 advanced and emerging market economies, average nonfinancial assets amount to 67 percent of GDP (with a median of 50 percent) although with very wide cross-country dispersion (Figure 16): from more than 120 percent of GDP for the Czech Republic, Japan, Korea, and Latvia to less than 20 percent of GDP for Bolivia, Hong Kong Special Administrative Region, and Switzerland. These sharp differences reflect a range of factors, such as economic structure, but also coverage of the data and evaluation methods. A large portion of nonfinancial assets (on average more than half) is owned by lower-level governments. The share of regional and local governments is particularly high in federal states where subnational government assets exceed central government holdings by a ratio of at least four to one.

Receipts and savings from sales and management of nonfinancial assetts have been rather small so far. 33 For example, disposal of federal nonfinancial assets in the United States (mainly through sale of electromagnetic spectrum rights and leasing of offshore drilling rights) is reported to have yielded only about ½ percent of GDP over the past two decades. In France, a 2006 initiative to dispose of public buildings (parc immobilier) yielded only 0.2 percent of GDP. In the current environment, asset liquidation may be difficult and market values may be low, reducing the immediate cash potential. In addition, only a very small share of nonfinancial assets are considered by the authorities to be “salable” (for example, 4 percent of GDP in Italy and up to 7 percent of GDP in Japan).34 The collection of user charges where they are not yet imposed (such as road tolls), including through public-private partnerships, could be a more promising source of revenues, since a large share of nonfinancial assets consist of public infrastructure. For example, in Germany, the toll for trucks on federal highways has created annual revenues of about 0.2 percent. In most cases, however, more effective asset management must start with better and more comprehensive reporting.

The spoonful of sugar: Faster growth

The historical evidence reviewed earlier suggests that growth is key for sustained fiscal consolidation. Empirical studies have found that a 1 percentage point increase in growth has led on average to an improvement in the headline primary balance of about ½ percent of GDP (see, for example, Woo, 2003, and Zeng, 2013). In an economy with revenues equivalent to 40 percent of GDP, a 1 percentage point increase in potential economic growth would result in an improvement of 0.4 percent of GDP in the structural fiscal balance if the resulting higher revenue were entirely saved. In addition, the denominator of the debt-to-GDP ratio would also rise. Through these two channels, a country with a debt ratio of 100 percent of GDP could reduce its debt by 30 percent of GDP in 10 years with one additional percentage point of potential growth. This could eventually give rise to a virtuous circle in which lower debt levels would raise potential growth, further facilitating debt reduction.

One critical assumption in this scenario is that the higher revenues associated with faster growth are saved. In practice rising revenue may lead to strong spending pressures that would have to be resisted. Moreover, raising potential growth may not be an easy task, as some advanced economies are already at the production possibility frontier. Boosting growth in these countries will require the introduction of extensive structural reforms that, in any event, may take time to have an effect.

How it can be done

The foregoing analysis shows that the scale of the challenge involved in bringing debt ratios to prudent levels varies significantly across countries. Many have already put their debt ratios on a downward path or need only modest additional adjustment to do so. But for the most highly indebted advanced economies, the adjustment required remains substantial and largely unprecedented. This does not mean, however, that fiscal sustainability is out of reach. One critical factor for success is that these countries maintain low interest rates over an extended period of time and in a noninflationary way. Credibility in fiscal adjustment is essential to achieving this aim. Faster growth will also help. This will require a combination of structural policies to improve productivity and continued monetary accommodation in advanced economies—provided that fiscal adjustment continues to avoid the risk of fiscal dominance and that appropriate safeguards for financial stability are in place. And in some countries, proceeds from the sale of government assets can be used to bring debt ratios down. But most important, policies should be geared toward replacing uncertainty about the future with confidence. From that vantage point, there are two priorities on the fiscal front. First, those advanced economies with medium-term fiscal adjustment plans should ensure that uncertainties about implementation measures are minimized and that adjustment tools are conducive to growth.35 Those without such plans, in particular Japan and the United States, should rapidly introduce them (the Japanese authorities plan to announce a medium-term fiscal consolidation plan this summer). Second, fiscal institutions should be strengthened to enhance the prospects for long-lasting fiscal discipline. In this respect, recent reforms to establish stronger fiscal institutions in advanced economies are welcome steps that should be promptly complemented by effective enforcement and accountability mechanisms.

Box 1.How Can Fiscal Councils Strengthen Fiscal Performance?

There is a growing interest in the role that fiscal councils can play in promoting sound fiscal policies. Although a handful of such councils have been in place for some time in advanced economies, their number has risen in recent years and is expected to continue to do so in the short to medium term (Figure 1.1). In the European Union, the Fiscal Compact explicitly refers to the monitoring of compliance with national fiscal policy rules by an independent body in each euro area member state, and a draft regulation (part of the so-called two-pack) mandates that each euro area country establish an “independent fiscal body” tasked with producing “independent macroeconomic forecasts” to be used for budget preparation. Beyond Europe, the creation of parliamentary budget offices in Australia, Canada, and Kenya suggests a broader interest in these institutions.

Figure 1.1.Number of Active Fiscal Councils

Source: National data.

Fiscal councils are typically established to promote fiscal responsibility, notably by raising the reputational costs of unsound policies and broken commitments. Their tasks usually follow the budget process, from the provision of unbiased budgetary forecasts to the scoring of specific policy initiatives, the preparation of long-term sustainability analyses, the monitoring of compliance with fiscal policy rules, and nonpartisan assessments of fiscal outcomes. By providing unbiased technical inputs to the budget process, fiscal councils increase fiscal transparency and improve the quality of the public debate on fiscal policy.

Two main features distinguish fiscal councils from other institutions that perform similar tasks, such as central banks, research institutes, or international organizations: the specific mandate they receive from the government to perform such tasks independently and the corresponding accountability, and the need to benchmark their assessments against the government’s stated objectives (to avoid being drawn into partisan considerations and guarantee their legitimacy).

IMF (2013b) explores empirically the effectiveness of fiscal councils. Bearing in mind the methodological challenges inherent in the small and highly heterogeneous population of councils, the study finds that the establishment of a fiscal council is on average associated with stronger budget balances and less procyclical policies. Fiscal councils also appear to limit certain sources of policy bias that can occur early in the budget process, such as the tendency to base budgets on optimistic macroeconomic and revenue forecasts and to abuse the uncertainty inherent in the implementation of fiscal rules contingent on the business cycle, such as structural balance rules. Several aspects of fiscal council design and modes of operation seem to enhance their effectiveness, including strict operational independence, strong and effective media presence, and a proper sequencing of the council’s activity with the budget process.

Establishing a fiscal council is not, however, a magic bullet against fiscal biases. Two specific considerations are worth emphasizing:

  • A fiscal council can be effective only if policymakers have internalized the merits of fiscal discipline to start with. If that has taken place, establishing a council—and the enhanced transparency that comes with it—is a signal to the public and to markets that fiscal policy is consistent with long-term sustainability and aligned with policymakers’ preferences. In addition, as even the best intentions can often dissipate in the face of short-term pressures, the watchdog role of a fiscal council helps raise the reputational costs of unjustified policy slippages, enhancing the credibility of commitments to discipline.

Regardless of whether the fiscal council is a signaling tool, a commitment device, or both, likely benefits include lower risk premiums and reduced uncertainty.

  • A critical mass of expertise and resources is needed for the council to deliver credible inputs to the budget process (forecasts, costing) and public analyses that are perceived as nonpartisan.

Box 2.The Appropriate Pace of Short-Term Fiscal Adjustment

Whereas there was broad agreement among economists at the height of the financial crisis about the need for expansionary fiscal policies to help support demand, there has been no consensus about how rapidly—if at all—fiscal stimulus should now be unwound. Some point to still-high unemployment rates and other indicators of weak demand to argue against deficit reduction, which would hurt output in the short run, or even in support of additional fiscal stimulus. Others see the rapid buildup of debt and still-high deficits in some of the largest advanced economies as harbingers of a future fiscal crisis, especially in the context of projected trend increases in pension and health care spending. Do policymakers really face a choice between inflicting high social costs on their populations through fiscal austerity to bring down deficits and debt and playing with fire by delaying adjustment to support growth and employment?

There is now considerable evidence that fiscal multipliers—which measure the impact on output of discretionary changes in the fiscal balance—could be large in the current environment. There are two main reasons for this. First, during a recession, when there is already significant unused productive capacity in the economy, cuts in demand caused by fiscal tightening will have a large impact on real activity. This is in contrast to periods of very strong demand and high capacity utilization, when some of the impact of fiscal tightening will show up in reduced pressure on prices, rather than lower output. Second, with nominal interest rates already close to zero and credit channels impaired in many advanced economies, there is limited scope for monetary policy to offset the contractionary impact of fiscal tightening. But this situation is unlikely to persist indefinitely. As private sector balance sheets are mended and banks recover their lending capacity, private demand and capacity utilization should pick up, helping fiscal multipliers decline to more moderate levels.

This suggests that for many countries, deficit reduction would ideally be deferred to the future, when its output costs would likely be lower. Of course, countries under market pressure may find that limited access to financing leaves them with no option but to front-load their adjustment. Even countries with relatively unimpeded market access operate under some constraints, however. A mere promise to reduce deficits at some point in the future is unlikely to be seen by markets as credible, especially if debt ratios continue to spiral in the meantime. Concerns about credibility may be especially acute in countries where political gridlock may stand in the way of any consensus on fiscal policy. The fact that many advanced economies are still able to borrow at historically low interest rates means that market confidence in these sovereigns remains intact for now. However, this should not be interpreted as evidence of permanent immunity from costly rises in the risk premium. After all, many of the advanced economies now facing market pressures were recently seen as risk-free bets, too.

For countries with adequate financing space, then, the safest course of action is to undertake a path of gradual but sustained adjustment that aims at steady progress over the medium term toward a clearly defined objective. In this context, the adjustment undertaken in the advanced economies over the last three years, averaging about 1 percent of GDP annually in cyclically adjusted terms, seems about right. Where needed, measures to address rising entitlement costs could also contribute to building confidence.

Of course, even modest up-front adjustment will involve output and employment costs, and it will therefore be essential to ensure that other policies remain as supportive as possible. In particular, monetary policy should remain accommodative for the foreseeable future, and structural policies to promote growth should also be pursued. The composition of fiscal adjustment could also be tilted to mitigate its adverse impact on the most vulnerable.

Box 3.Bond Yields and Stability of the Investor Base

Since the start of the global crisis, the sharp contrast in the level and volatility of bond yields among countries with similar deteriorations of their fiscal accounts suggests that one must look beyond just fiscal fundamentals for an explanation. The composition of countries’ investor bases sheds light on the structural factors that drive these developments. Indeed, “real-money investors”—comprising domestic nonbank financial institutions as well as national and foreign central banks—are unleveraged and typically are “buy and hold” investors and are thus able to provide a more stable source of demand for government debt.1

Figure 3.1 illustrates that countries with the highest share of real-money investors at the end of 2007 were also among those that saw the lowest volatility during the crisis. Empirical analysis confirms this relationship over a longer period: while the level and volatility of bond yields rise with the debt ratio, they fall with the share of the debt that is in the hands of real-money investors.2 Interestingly, while large holdings by domestic nonbank institutions depress yields and volatility in both advanced and emerging market economies, the impact of higher central bank holdings differs, putting downward pressure on yields and volatility in the former but increasing them in the latter. This may reflect concerns that central bank holdings of government debt in emerging market economies could be the result of inflationary budget financing.

Figure 3.1.Real Money Investors before the Crisis, 2007

(Percent of total debt, except as otherwise indicated)

Sources: Jaramillo, Zhang, and Gomez (2013); and IMF staff estimates.

Note: Bond yield volatility is measured as the annual standard deviation in basis points of sovereign bond yields, based on daily data.

In recent years, the mitigating role of central bank holdings in advanced economies most likely reflects quantitative-easing strategies undertaken for monetary policy purposes by the Bank of England, the Bank of Japan, and the U.S. Federal Reserve (and to a lesser extent the European Central Bank).3 As documented in the April 2012 Fiscal Monitor, with large central bank purchases of government debt and other assets, consolidated general government and central bank debt is, on average, 30 percent of GDP lower than gross general government debt (Statistical Table 15). Central bank claims on the general government continued to increase in 2012, particularly in the United Kingdom.

In emerging market economies, a substantial share of gross debt comes from foreign official (non-market-determined) creditors, lowering borrowing costs and their volatility. In 2012, official loans accounted for 17 percent of the public debt of emerging market economies, with one-third of countries relying on non-market-determined creditors for more than 30 percent of their debt (Figure 3.2).4 Effective interest rates for countries where official lending accounts for at least 17 percent of their government debt are, on average, 60 basis points lower than those for countries below this threshold, contributing to a generally lower interest rate–growth differential.

Figure 3.2.Emerging Market Economies: Share of Official Lending in Total Debt Stock, 2012


Source: IMF staff estimates.

1This definition of institutional investor base is chosen to ensure comparability across a wide set of advanced and emerging market economies. Lack of data for this broad set of economies does not allow for the distinction of other buy-and-hold investors such as households and unleveraged nonresident funds, which could be important sources of financing in some countries.2Analysis of the determinants of bond yield level and volatility is based on an annual panel data set of 30 advanced and 13 emerging market economies between 2000 and 2012. A Hausman-Taylor model is used to control for endogeneity occurring through the unobserved individual effects and allows for the inclusion of time-invariant variables. Explanatory variables, all lagged one period, include real-money investor holdings, expectations of macroeconomic and fiscal variables, global interest rates, volatility of bank returns, and dummy variables for euro area countries, reserve currency countries, and large emerging market economies. See Jaramillo, Zhang, and Gomez (2013).3The Eurosystem of central banks’ holding of sovereign debt due to monetary policy operations under the Securities Market Programme is much smaller (2 percent of total euro area gross government debt) but could potentially be expanded significantly under the Outright Monetary Transactions.4In 2010 (latest data available), the share of official lending in low-income countries’ external loans was 95 percent; the share of concessional loans was about 75 percent.

Box 4.Potential Sources of Contingent Liabilities in Emerging Market Economies

In emerging market economies as well as in advanced economies, contingent liabilities can arise from multiple sources. This box discusses specific examples in China and India.


The financing of local infrastructure projects is a potential source of fiscal risk in China. Subnational governments in China are generally prohibited from borrowing directly, but on-budget revenue has not been adequate to finance their current spending (subnational governments account for the majority of social spending) and infrastructure. Therefore, subnational governments have financed infrastructure spending off budget, through the creation of local-government financing vehicles (LGFVs) that borrow directly from banks and capital markets, possibly collateralized by state-owned assets or land. Many existing LGFVs also rely on proceeds from land sales to repay their debt (Figure 4.1).

Figure 4.1.China: Financing Sources of Infrastructure Investment

(Percent of GDP)

Sources: CEIC; National Audit Office; Soufun.

Note: 2012 data are estimates.

LGFV borrowing accelerated as subnational governments were assigned the responsibility to spend three-fourths of the Y4 trillion stimulus package (about 12 percent of GDP) in response to the global economic crisis, but lacked the financial resources to do so. The resulting infrastructure spending provided significant countercyclical support to the economy, but has raised concerns about the size of the public debt, the sustainability of subnational government finances, and the risk of deterioration in bank asset quality. A nationwide survey undertaken by the National Audit Office estimated subnational government debt at 27 percent of GDP1at the end of 2010, of which nearly half (47 percent) was channeled through LGFVs. Subnational government debt has almost doubled since 2008, driven largely by the booming LGFV-originated debt. Bank loans account for the majority of subnational governments debt (close to 80 percent), whereas direct government bond issuance amounts to only 7 percent.2 Infrastructure spending accelerated again in 2012, to more than 15 percent of GDP, with subnational governments expected to finance about one-third. LGFVs are reportedly making more recourse to corporate bonds and trust products than before, but no updated data on the level of subnational government debt are yet available.

Available data suggest that the broadly balanced cyclically adjusted fiscal position and the favorable interest rate–growth differential currently mitigate LGFV-related near-term fiscal risks. Some local governments may be in a more vulnerable position than others, however, particularly those in the less-developed western provinces (Figure 4.2). Governments below the provincial level, which according to 2010 data together account for 70 percent of total subnational government debt, are also exposed to higher debt-servicing risks (Figure 4.3). More broadly, the debt-servicing capacity of local governments is vulnerable to large shifts in the housing market, given their reliance on land sales—about 80 percent of surveyed cities and 40 percent of surveyed counties promised to repay their debt using proceeds from land sales (Figure 4.3).

Figure 4.2.China: Debt Repayment Capacities by Regions, 2010

Sources: CEIC; National Audit Office.

Figure 4.3.China: Repayment Capacities at Lower-Government Level

(Percentage of cities and counties surveyed)

Source: National Audit Office.

Because of the large number of LGFVs and the unclear boundary between an LGFV and a commercially oriented, locally owned state enterprise, it is difficult to estimate the extent of LGFV borrowing that is fiscal in nature and the potential implications of this borrowing for government liabilities. The data reported in Tables 1 and 2 may not fully reflect the true extent of government deficits and debt. International experience suggests that subnational government could be an important source of fiscal risks over the medium term, especially in the context of rapid urbanization and related rising expenditure needs. Authorities have taken initial steps to address the fiscal risks, including conducting local government debt surveys, cleaning up LGFVs that run large deficits, forbidding government guarantees and public-property-based collateral for bank loans, encouraging market-based financing channels such as corporate bond markets and municipal bonds in pilot provinces, and promoting private engagement in public infrastructure investment. Establishment of the appropriate legal framework, reporting requirements, and accounting standards can support effective implementation of these initiatives and help further contain fiscal risks.


Credit growth was rapid in India in the years before the crisis, with lending to the private sector expanding by 20 percentage points of GDP during 2001–08. It remained strong in the aftermath of the crisis, with a growing concentration on infrastructure projects, in response to the government’s ambitious investment targets. India’s banks remain well capitalized, and the likelihood of financial sector stress is low (as noted in the October 2012 World Economic Outlook). But credit quality has tended to deteriorate recently, particularly among the state-owned banks, which account for 73 percent of banking assets. Gross nonperforming assets in public banks reached 3.3 percent of advances in 2012. However, the long-run risk may be underestimated, as historically about 15 percent of assets reported as “restructured” (a category that likely accounted for 7.3 percent of the public banks’ assets as of September 2012) are eventually classified as nonperforming.

The Reserve Bank of India has recently taken important steps to tighten bank reporting requirements to get a more accurate picture of asset quality. But state-owned bank portfolios remain vulnerable to losses from delayed infrastructure projects and, most importantly, to the recent growth slowdown that has dented the profits of the large companies that account for the bulk of Indian banks’ loan portfolios. The economy now appears to have bottomed out, but this may not yet be fully reflected in banks’ credit quality. The new and higher capital standards under Basel III will also demand an increase in bank capital if credit growth is to continue. Although precise estimates of state-owned banks’ future capital needs are difficult to compute, these needs are expected to be about 1 percent of GDP cumulatively between 2013 and 2019, depending on the growth trajectory. Further asset quality deterioration could raise needs substantially, but reducing government ownership (which would require legislative action) could bring them down. Beyond bank capitalization, potential losses among India’s state-owned enterprises and the large program of public-private partnerships also represent contingent liabilities.

1 Of which 17 percent pertains to direct repayment obligations, 6 percent is explicit guarantees, and another 4 percent refers to implicit liability for certain debt relief.2 This includes bonds issued by the central government on behalf of local governments since 2009.

Box 5.Fiscal Adjustment in the United States: Making Sense of the Numbers

There is hardly any disagreement that fiscal consolidation is essential for many advanced economies over the medium term. However, views often diverge about exactly how much adjustment is required, reflecting important differences in assumptions. This box illustrates this issue by looking at the case of the United States.

In the United States, several agencies make projections of the federal budget balance and debt level, including the Office of Management and Budget (OMB), which is responsible for the formulation and execution of the budget, and the Congressional Budget Office (CBO), an independent agency that provides analysis to support the congressional budget process. The CBO provides 10-year projections on the basis of legislated policies, whereas the OMB also incorporates the estimated budgetary effects of new measures proposed by the administration. While the OMB’s latest estimates1 show that the federal debt held by the public2 will stabilize by 2022 as a share of GDP, the CBO’s estimates show debt still increasing at that date. The implied improvements in the headline primary balance are 7½ percentage points for the OMB during 2012–22 and 3½ percentage points for the CBO during 2013–23. In contrast, the illustrative baseline adjustment scenario in the current issue of the Fiscal Monitor shows a required improvement in the headline primary balance of about 10 percent of GDP by 2022.3 What explains these sharp differences?

  • Comparing apples with oranges. Whereas the CBO and OMB focus on the federal government debt held by the public, the Fiscal Monitor looks at gross debt of the general government, which also includes state and local governments. The difference between the two concepts was about 33 percent of GDP in 2012.

  • Different policies, different outcomes. The policy assumptions underlying the CBO and OMB projections, on the one hand, and the Fiscal Monitor projections, on the other, differ completely. CBO and OMB project the fiscal path under legislated and/or proposed policies, with no explicit assumptions about debt targets. Both offices find that the federal debt would be in the neighborhood of 75 percent of GDP by 2022. By contrast, the Fiscal Monitor calculates the adjustment required to achieve a general government debt target of 60 percent of GDP by 2030, an entirely different exercise than that undertaken by the CBO or OMB. The implicit federal government debt under the illustrative Fiscal Monitor scenario would decline to about 54 percent of GDP by 2022. That scenario involves a larger improvement in the primary balance than the CBO or OMB forecasts, primarily because it targets a debt ratio that is much more ambitious than what comes out of the CBO/OMB projections.

  • Different macroeconomic outlook, different adjustment path. Both the OMB and CBO assume a relatively benign interest rate environment over the next 10 years, with interest rate–growth differentials (for the federal government) still close to zero by 2022.4 The Fiscal Monitor baseline assumes that interest rate–growth differentials (for the general government) will also remain relatively low for the next 5 years but will increase gradually thereafter as monetary policy is normalized and fundamentals, including pressure on the available saving pool and the risks implied by high debt levels, begin to have more weight. This implies an interest rate–growth differential of 1¼ percent by 2022 (Figure 5.1), above the pre-2007 long-term average (¾ percent) and reflecting much higher debt levels.5 Using instead the interest rate–growth differential projected by the CBO and OMB, the required adjustment to achieve the 60 percent general government debt target by 2030 would be reduced by about 1 percent of GDP relative to the Fiscal Monitor baseline (Table 5.1).

Figure 5.1.United States: Assumptions on Interest Rate–Growth Differential, 2012–22


Sources: U.S. Office of Management and Budget (2012); U.S. Congressional Budget Office (2013); and IMF staff estimates.

Note: The implied interest rate–growth differentials from the Office of Management and Budget and Congressional Budget Office are based on calculations using net interest and are converted to a calendar-year basis.

Table 5.1.United States: Key Results under the CBO, OMB, and Fiscal Monitor Adjustment Scenarios(Percent of GDP)
OMBCBOFiscal Monitor
Federal debt ratio in 20221757654
Headline primary balance adjustment, 2012–2227.53.510.2
Headline primary balance adjustment, 2012–22:
Scenario with OMB/CBO interest rate–growth differential assumptions9.0
Sources: U.S. Office of Management and Budget (OMB) (2012); U.S. Congressional Budget Office (CBO) (2013); and IMF staff estimates.

OMB and CBO debt ratios refer to federal government. Fiscal Monitor baseline scenario shows a general government debt of about 87 percent of GDP by 2022. Assuming state, local, and federal trust funds amount to 33 percent of GDP, the implied federal debt is 54 percent of GDP.

For the CBO, the change in the primary balance refers to the period 2012–23.

Sources: U.S. Office of Management and Budget (OMB) (2012); U.S. Congressional Budget Office (CBO) (2013); and IMF staff estimates.

OMB and CBO debt ratios refer to federal government. Fiscal Monitor baseline scenario shows a general government debt of about 87 percent of GDP by 2022. Assuming state, local, and federal trust funds amount to 33 percent of GDP, the implied federal debt is 54 percent of GDP.

For the CBO, the change in the primary balance refers to the period 2012–23.

1 The latest OMB analysis is contained in the Mid-Session Review of July 2012; the CBO released “The Budget and Economic Outlook for Years 2013–2023” in February 2013. Both institutions also produce 75-year fiscal forecasts annually.2 This includes debt held by the Federal Reserve but excludes federal trust funds and other government accounts. The concept of general government debt used in the Fiscal Monitor adds other liabilities of the federal government and the gross liabilities of state and local governments to the “debt held by the public” and subtracts the cross-holdings.3 This number is consistent with that reported in Statistical Table 13a, which shows the adjustment in the cyclically adjusted primary balance between 2013 and 2020. For comparison with the CBO/OMB numbers, the change in the headline primary balance between 2012 and 2022 is reported here.4 Owing to data availability, the interest rate–growth differential reported here for the CBO and OMB is based on net, rather than gross, interest payments. The Fiscal Monitor uses the latter concept. The implied interest rate–growth differential could differ by an average of about 0.5 percent because of this.5 The average interest rate–growth differential in the United States over 1991–2008 was 0.7 (Escolano, 2010).

Box 6.Public Debt Dynamics and Fiscal Adjustment in Low-Income Countries in Sub-Saharan Africa

In most low-income countries in sub-Saharan Africa,1 public debt–to–GDP ratios—particularly those involving external debt—declined significantly throughout the early 2000s (Figure 6.1, left panel). Although external debt relief under the Heavily Indebted Poor Countries Initiative and the Multilateral Debt Relief Initiative played a key role in the reduction, additional factors (such as, in particular, faster output growth and exchange rate appreciation) also helped (Figure 6.1, right panel). For oil exporters, the favorable terms-of-trade shock in 2005–06 improved fiscal balances and boosted growth, significantly alleviating the debt burden. For low-income countries, apart from debt relief (reported as part of “Other contributions” in Figure 6.1, right panel), the most crucial factor was GDP growth, with no significant impact from real exchange rate appreciation. Among fragile states, a few still exhibit high debt-to-GDP ratios, but most have experienced sharp declines (e.g., Burundi, the Central African Republic, the Democratic Republic of the Congo, and Liberia).

Figure 6.1.Sub-Saharan Africa: Public Debt–to–GDP Ratio and Debt Accumulation Decomposition

(Percent of GDP; median)

Sources: IMF staff estimates and projections.

Note: Fragile: fragile countries; LICs: low-income countries; LMICs: lower-middle-income countries; Oil: oil exporters; SSA: sub-Saharan Africa.

Since debt relief, improved macroeconomic policies and generally strong growth have kept debt ratios stable on average. However, a few countries (like Ghana and Senegal) have registered sizable increases, largely on account of rapid growth in spending, including for infrastructure. In addition, fiscal deficits widened in most countries in the aftermath of the Great Recession, and have narrowed little since then.

1This box focuses on sub-Saharan African countries currently eligible for concessional financing from the IMF under the Poverty Reduction and Growth Trust. A number of these countries are lower middle income.

    Other Resources Citing This Publication