Fiscal Politics
Chapter

Chapter 12. Expenditure Rules: Effective Tools for Sound Fiscal Policy?

Author(s):
Vitor Gaspar, Sanjeev Gupta, and Carlos Mulas-Granados
Published Date:
April 2017
Share
  • ShareShare
Show Summary Details
Author(s)
Till Cordes, Tidiane Kinda, Muthoora Priscilla and Anke Weber 

Introduction

Discussions about the role of fiscal institutions in promoting sound public finances are increasingly at the forefront of the policy debate. This renewed focus on fiscal institutions comes against the backdrop of a marked deterioration in public finances across most advanced economies since the global financial crisis of 2008. However, it also reflects the need to address the root causes of an apparent deficit and debt bias—even before the crisis, the average debt ratio had reached its highest level since World War II, driven by a ubiquitous increase in the share of government spending in GDP (Cottarelli 2012; IMF 2014).

The previous chapters in this book provide new evidence suggesting that the interaction of politics with fiscal policy influences the timing, level, composition, and financing of spending, contributing in large measure to the observed deficit and debt bias. Fiscal institutions may help contain political pressures on tax and spending decisions by constraining the behavior of governments and opening their policy choices to scrutiny by financial markets and the public (Schuknecht 2004). Notably, procedural rules can be laid down to guide the preparation of annual budgets; numerical fiscal rules can be used to impose limits on the level or growth of key budgetary aggregates; and independent institutions can be mandated to prepare macroeconomic forecasts, analyze the effects of proposed government plans, make policy recommendations, or assess compliance with fiscal targets.

Numerical fiscal rules, particularly expenditure rules, have received increasing attention because of a number of these features.1 Expenditure rules are directly aimed at addressing the spending pressures often at the origin of excessive deficits. They are most closely related to the formulation of the annual budget, setting legally binding appropriations that in turn contribute to the rules’ enforceability. By the same token, expenditure rules can be used to guard against political failure in coalition governments. According to Hallerberg, Strauch, and von Hagen (2007) and Hallerberg, Strauch, and Yläoutinen (2010), countries with ideologically dispersed coalitions will not be as willing to delegate power to a finance minister representing only one of the parties in government. Instead, coalition governments are more likely to opt for multiyear expenditure targets, which they can credibly fix in their coalition agreement for their electoral period. In practice, several advanced economies where coalition governments tend to dominate the political landscape (Belgium, the Netherlands, Luxembourg, Sweden, and Finland, for example) indeed set expenditure targets in their coalition agreements. Expenditure rules fully accommodate revenue shortfalls resulting from adverse economic shocks, allowing for a stabilizing role for fiscal policy. Importantly, and unlike deficit caps, expenditure rules also help create buffers in good times, when revenue windfalls can make spending pressures difficult to resist. Thus, expenditure rules can mitigate the effect of economic and political cycles. Finally, expenditure rules are also transparent and generally easy to monitor because the requirements for timely data reporting, accounting, and forecasting are more easily met than for other budgetary aggregates (Ayuso-i-Casals 2012).

A lingering concern, however, remains about the effect of expenditure rules on the composition of spending. By reducing incentives for spending overruns, expenditure rules can lead to stricter prioritization and greater efficiency in spending. But the interaction of political economy considerations with a binding constraint on total spending may result in the crowding out of productive but electorally unappealing projects (Debrun 2014).

This chapter draws on the 2013 vintage of the fiscal rules database developed by Schaechter and others (2012) to analyze compliance with fiscal rules. It also investigates the effect of expenditure rules on the level and composition of public spending in advanced and emerging economies using a sample of 29 economies with expenditure rules between 1985 and 2013.

The chapter’s findings suggest that expenditure rules are associated with spending control, countercyclical fiscal policy, and improved fiscal discipline. Fiscal performance is better in countries where an expenditure rule exists. This outcome appears to be related to the properties of expenditure rules because compliance rates are generally higher than with other types of rules (on the budget balance or debt, for example). In particular, compliance with expenditure rules is higher if the expenditure target is directly under the control of the government and if the rule is not a mere political commitment but enshrined in law or in a coalition agreement.

Evidence of adverse side effects is mixed. The introduction of expenditure rules is associated with a decrease in public investment only in emerging economies. A possible explanation is that any adverse effects on public investment could have been mitigated in advanced economies by well-designed budgetary frameworks and procedures.

Instead, the empirical analysis points to two positive side effects. First, expenditure rules reduce the volatility of expenditure, thus imparting a degree of predictability to fiscal policy and making it less destabilizing. Importantly, the evidence suggests that expenditure rules have been effective in mitigating political cycles. In economies with expenditure rules, the holding of legislative or executive elections is not associated with higher primary spending relative to nonelection years. Second, expenditure rules are associated with higher public investment efficiency. However, these results need to be interpreted with care given the relatively small sample size.

The rest of the chapter is structured as follows. The second section provides stylized facts on expenditure rules. In particular, it analyzes who uses expenditure rules and the key characteristics of expenditure rules that are (or were) in place. The third section discusses why countries adopt and abandon expenditure rules. The fourth section analyzes whether expenditure rules have been effective in practice, shedding light on compliance with expenditure targets, investigating the behavior of primary balances around the introduction of the rules, and looking at the response of public investment spending and efficiency following their introduction. The final section concludes.

What Type of Expenditure Rules are in Place and what are their Design Features?

In this chapter, expenditure rules are defined to include both specific numerical targets fixed in legislation and expenditure ceilings for which the targets can be revised, but only on a low-frequency basis (for example, as part of the electoral cycle), as long as they are binding for a minimum of three years.2

In practice, expenditure rules typically take the form of a cap on nominal or real spending growth over the medium term (Figure 12.1). Ceilings on real spending growth are relatively more frequent in advanced economies, possibly reflecting better capacity to estimate outlays in real terms.

Figure 12.1.Types of Expenditure Rules in 2013

(Number of countries)

Source: FAD Fiscal Rules database.

Expenditure rules are often used in connection with other national rules. The combination of expenditure rules and budget balance rules is particularly common in advanced economies, whereas in emerging economies, they are often used in connection with debt rules (Figure 12.2).3 A possible explanation is that not all types of fiscal rules are equally apt to support sustainability, economic stabilization, and possibly size-of-government objectives, even when their design features are fine tuned. Using a combination of fiscal rules can help address the gaps. For example, an expenditure rule combined with a debt rule would assist policymakers with short- to medium-term operational decisions while allowing for some countercyclicality and provide a link to debt sustainability.

Figure 12.2.Combination of Expenditure Rules with Other National Rules in 2013

(Number of countries)

Sources: FAD Fiscal Rules database; and authors’ calculations.

Note: BBR = budget balance rule; DR = debt rule; ER = expenditure rules; RR = revenue rule.

Expenditure rules are more commonly established through statutory norms in emerging economies than in advanced economies. In advanced economies, expenditure rules tend to be more closely integrated into the medium-term expenditure frameworks, which are sometimes part of coalition agreements (Figure 12.3). Belgium, Finland, Luxembourg, the Netherlands, and Sweden, for example, set expenditure targets in their coalition agreements.

Figure 12.3.Legal Basis of Fiscal Rules in 2013

(Percent)

Source: FAD Fiscal Rules database.

Note: EMDCs = emerging market and developing countries.

The majority of expenditure rules cover the central government. Of the 23 countries that had expenditure rules in place in 2013, 15 had one at the central government level and 8 at the general government level.4 The greater prevalence of rules at the central government level may reflect autonomy and coordination issues with subnational governments.

With regard to economic coverage, a number of items are frequently excluded (Table 12.1). Fiscal sustainability considerations argue for more comprehensive coverage, but other competing objectives (such as improving the composition of spending) and controllability arguments are put forward to exclude certain items. Broad coverage aims at managing total revenue and expenditure and makes the target more transparent and easier to monitor. Nevertheless, it is sometimes seen as desirable to exclude, for example, capital expenditure since it is generally expected to positively contribute to long-term growth. However, this exclusion can cause problems because it weakens the link with gross debt. Moreover, not all capital expenditure is necessarily productive and, depending on country circumstances, other items such as health care and education expenditure may raise potential growth even more. Excluding interest payments and cyclically sensitive expenditure from target variables is also often discussed since they are not under the control of governments in the short run and require short-term adjustments in other expenditure categories, with capital spending often the easiest to cut. An argument for including cyclically sensitive expenditure is that most cyclical sensitivity is on the revenue side.

Table 12.1.Economic Coverage of Aggregate
Item Most Frequently ExcludedCountries Where Exclusions Apply
Interest PaymentsFinland, France, Japan, Spain
Cyclically Sensitive ExpenditureFinland, Poland, Spain, United States
Capital ExpenditureCroatia, Ecuador, Peru
Security-Related SpendingIsrael, Peru
Source: FAD Fiscal Rules database.
Source: FAD Fiscal Rules database.

Design features vary across countries. In advanced economies, expenditure rules are often used in connection with medium-term expenditure frameworks and compliance is monitored by an independent fiscal body, a so-called fiscal council, whereas in emerging economies, formal enforcement mechanisms are more frequent (Figure 12.4).5 Well-defined escape clauses are relatively rare in connection with expenditure rules. Contingencies tend to be handled by leaving a margin between the budget envelope and the expenditure ceiling (Ayuso-i-Casals 2012).

Figure 12.4.Design Features of Expenditure Rules in 2013

(Number of countries)

Sources: FAD Fiscal Rules database; and authors’ calculations.

Why do Countries Adopt and Abandon Expenditure Rules?

From 1985 to 2012, 27 economies introduced 31 different expenditure rules. The adoption of these rules may have been motivated by various reasons. First, governments may adopt expenditure rules to help them achieve other supranational fiscal rules. Second, coalition governments can use these rules to bind them to certain fiscal targets for the electoral period. Finally, especially in recent years, economies have adopted expenditure targets in bad economic times to rein in spending. Out of the 31 expenditure rules that have been introduced since 1985, 10 had already been abandoned by 2013, either because the country had never complied with the rule or because fiscal consolidation was so successful that the government did not want to be restricted by the rule in good economic times.

Although expenditure rules are typically used in connection with other fiscal rules, they are most often not adopted at the same time as these other rules (Figure 12.5). Only 6 of the 31 expenditure rules were introduced together with other rules. For instance, Argentina, Brazil, and Peru included expenditure rules in their fiscal responsibility laws as part of a wider set of reforms in 1999 and 2000. However, for 20 of the 31 cases, other fiscal rules were already in place when the expenditure rule was introduced. In particular, many European Union countries adopted national expenditure rules to help them achieve the Maastricht deficit and debt limits. Belgium, Denmark, Luxembourg, and the Netherlands introduced national expenditure rules in response to these new supranational fiscal rules after 1992. Sweden and Finland followed a couple of years later after becoming members of the European Union in 1995. The variation in European countries’ adoption of national expenditure rules can partly be explained by the different structure of governments. Minority governments, such as in Sweden in the 1990s, have also introduced expenditure rules to agree on fiscal targets for their electoral period with the opposition in parliament.

Figure 12.5.Adoption of Expenditure Rules

(Percent)

Source: FAD Fiscal Rules database.

Most of the expenditure rules were adopted in bad economic times. In all but six cases, the implementation of the expenditure rule was preceded by a negative change in the output gap. More than one-third of the expenditure rules were introduced since 2009, in response to the financial crisis. Earlier findings suggest that expenditure rules were mainly adopted following prior consolidation to lock in fiscal adjustment (IMF 2009). However, taking into account recent adoptions of fiscal rules in response to the crisis, less than a third of all expenditure rules were introduced after fiscal consolidation, defined as an improvement in the budget balance compared with the previous two or three years.

About a third (10 of the 31) of expenditure rules have already been given up for various political and economic reasons. In 6 of the 10 cases, the country did not comply with the rule in the year before giving it up. Once a rule has lost its credibility, countries might not see any benefit in upholding it. In Argentina, Iceland, and Kosovo, the rule was never observed, and all three countries abandoned their prevailing rule when they faced global market turmoil in 2008–09. Bulgaria and Japan also had problems complying with their rules during the financial crisis and decided to implement new expenditure rules. Political changes were more important in other countries. In Australia, the first political commitment to an expenditure rule was only made for the life of parliament. Following an early election in 1988, the new parliament did not specify a new rule.

In some countries, there was a perception that expenditure rules had fulfilled their purpose. Following successful consolidations in Belgium, Canada, and the United States in the 1990s, these countries saw no need to continue to follow their national expenditure rules. Especially in the United States, spending pressures increased during an economic boom at the end of the 1990s. More and more spending was made outside of the expenditure ceilings via an emergency spending category (CBO 2003). The ceilings were raised on an ad hoc basis several times in the two years preceding abandonment of the rule in 2002.

What is the Evidence?

Combining the 2013 vintage of the fiscal rules data set with indicators on budgetary outturns from various sources (IMF databases, quantitative and qualitative ex post budgetary assessments from fiscal councils) can help shed light on who complied with expenditure rules and when, on behavior of primary balances around the introduction of the rules, and on associated changes in public investment spending and efficiency. The results reported below need to be interpreted with caution, given that establishing causation between institutions and policy outcomes is a perennial challenge. For instance, it could be that expenditure rules are primarily adopted by countries with an intrinsically strong commitment to fiscal discipline, good public expenditure management practices, or good institutions, generally ex ante. In addition, the relatively small sample suggests that results could be affected by outliers.

Compliance

Previous studies on the impact of expenditure rules on fiscal performance do not analyze whether countries actually comply with the rules they have adopted (Debrun and others 2008; Wierts 2008; Turrini 2008; Holm-Hadulla, Hauptmeier, and Rother 2010; Nerlich and Reuter 2013). This section provides a compliance assessment for almost all countries with expenditure rules, covering 95 percent of the 217 country-years since 1985. Compliance is measured as a dummy variable (that is, it does not control for near misses) and is established using both quantitative and qualitative data from various sources such as the World Economic Outlook database, country budgets, and assessments by fiscal councils.6 Two points are worth highlighting: First, the assessment of compliance does not control for whether the rule was effectively binding. Put differently, it does not distinguish cases where the rule was met because countries really tried hard.7 Second, and related, the compliance rates may be exaggerated if countries abandon rules once they become binding. A couple of instances when this occurred are discussed below. Overall, the analysis shows that countries comply more often with expenditure rules than with other fiscal rules. In addition, the section points to two rule characteristics that are associated with higher compliance rates: First, countries comply more often if the expenditure target is directly under the government’s control. Second, compliance with expenditure rules is higher if the rule is enshrined in law or in a coalition agreement.

Comparing Expenditure Rules with Other Fiscal Rules

Countries have complied with expenditure rules more than two-thirds of the time. Figure 12.6 shows that expenditure rules have a better compliance record than budget balance and debt rules. The only exception is the high performance of emerging market economies with debt rules. However, this can be explained by European emerging economies’ very high compliance rates with the supranational Maastricht debt rule that was not effectively binding for most of these countries. In addition, the high rate of compliance with debt rules can be explained by the favorable impact of financial repression, which translated into persistent negative interest-growth differentials, on debt dynamics (Escolano, Shabunina, and Woo 2011). The higher rate of compliance with expenditure rules is consistent with the fact that these rules are easy to monitor and that they immediately map into an enforceable mechanism—the annual budget itself. Besides, expenditure rules are most directly connected to instruments that policymakers effectively control. By contrast, the budget balance, and even more so public debt, is more exposed to shocks, both positive and negative, that are out of the government’s control.

Figure 12.6.Compliance with Fiscal Rules, 1985–2012

Sources: FAD Fiscal Rules database; and authors’ calculations.

Note: The y-axis measures the average compliance rate with budget balance rules (BBR), expenditure rules (ER), and debt rules (DR) in all years in which an assessment could be made. BBRs and DRs include both national and supranational rules.

One of the desirable features of expenditure rules compared with other rules is that they are binding not only in bad but also in good economic times. The compliance rate in good economic times, defined as years with a negative change in the output gap, is at 72 percent—almost the same as in bad economic times at 68 percent. In contrast to other fiscal rules, countries also have incentives to break an expenditure rule in periods of high economic growth with increasing spending pressures. For instance, Iceland never managed to comply with its expenditure rule during its economic boom in the 2000s. The financial and economic crisis of 2009 would have been the first time that Iceland complied with its rule. In the United States, spending pressures also undermined the expenditure rule in good economic times. Belgium and Canada gave up their expenditure rules during periods of relatively high economic growth.

Which Rule Characteristics Are Associated with High Compliance Rates?

Although compliance with expenditure rules is high overall, there is still a large variance between different countries’ performance. Although correlations between certain rule characteristics and compliance should be interpreted with caution, two design features in particular are associated with higher compliance rates. First, compliance is higher if the government directly controls the expenditure target. Figure 12.7 shows the compliance rates for different types of expenditure targets. Specific ceilings have the best performance record. For these rules, governments specify nominal targets for each individual year, which gives a clear guideline for the budget process. Although some countries, such as Finland and the Netherlands, first set these targets in real terms, they convert them in the budget process to nominal terms (Ljungman 2008). The government controls these nominal expenditures directly. In contrast, the government does not have full control over expenditure targets if they are defined in relation to GDP or inflation. Evidence suggests that expenditure targets specified in levels of GDP have also had high compliance records. In many cases, however, this outcome occurs because targets were often set at very high levels so that they did not lead to a binding expenditure constraint. Once this target became binding, the lack of control over GDP or inflation became apparent. Botswana, for instance, exceeded its 40 percent expenditure-to-GDP target by more than 5 percentage points because of declining demand for diamonds (IMF 2012).

Figure 12.7.Compliance and Type of Rule

(Percent)

Sources: FAD Fiscal Rules database; and authors’ calculations.

Nominal expenditure growth rules have a poor performance record compared with GDP and real expenditure growth rules for two reasons. First, the government can aim to comply with the rule, but fail because of unexpected economic shocks. For instance, Iceland targeted transfer payment growth that varied strongly, from 5.9 percent in 2003 to -7.3 percent in 2004 (Gunnarsson 2011). One alternative is to use past macroeconomic indicators, as did Lithuania, which decided to use the average of revenue growth over the past five years as a benchmark instead of yearly changes, which provides a clear ex ante benchmark that does not vary too strongly. The government may also lose control over expenditure targets if these targets include local government spending. In Denmark, for instance, the rule was often broken because local governments overspent, thus breaking the overall rule for the general government (this situation has since been remedied by the adoption of binding expenditure ceilings that include the bulk of local government spending).

Second, the government has a lower incentive to comply with rules specified in relation to other macroeconomic factors because independent institutions, legislators, or the general public cannot hold the government directly accountable for noncompliance. In several countries, the rule was so vaguely specified that even independent institutions had problems checking compliance. For instance, in Luxembourg, the central bank tried to assess the compliance record, but had to make several assumptions to do so (Banque Centrale du Luxembourg 2005). In contrast, in countries with specific nominal ceilings, it is relatively easy for independent institutions to control compliance. For instance, the Swedish Fiscal Council and the U.S. Congressional Budget Office increased visibility of the rule and provided an independent assessment of compliance with the nominal ceiling.

In addition to the type of expenditure target, Figure 12.8 shows a correlation between the legal basis of the rule and average compliance rates. Political commitments have the poorest performance record. In contrast to coalition agreements and statutory rules, political commitments do not have any binding character.

Figure 12.8.Compliance and Legal Basis

(Percent)

Sources: FAD Fiscal Rules database; and authors’ calculations.

Performance in the six countries with coalition agreements is at least as good as in countries with statutory rules, with a high mean compliance rate of more than 80 percent in countries with coalition agreements. Coalition agreements set expenditure caps only for the lifetime of a coalition. By contrast, political commitments and statutory rules often cap expenditures beyond an electoral period. In several instances, new governments decided to abandon the rule, to not comply with it, or to make ad hoc changes in quantitative targets. For example, the 2007 expenditure targets in Israel were increased in an ad hoc manner once the new government was formed. In the United States, the legislature increased spending ceilings on an ad hoc basis for 2001 and 2002. Statutory laws cannot prevent these cases of noncompliance if the specific target and the coverage of the rule are not specified in the law.

Expenditure Rules and Long-Term Sustainability

Have expenditure rules been associated with better fiscal performance? A number of studies focusing on European countries have shown that the presence of expenditure rules could help mitigate spending and procyclical bias (Debrun and others 2008; Wierts 2008; Holm-Hadulla, Hauptmeier, and Rother 2010). Covering a larger sample of advanced and developing economies, new empirical analysis described here extends previous studies on expenditure rules and fiscal performance. It focuses on whether the rule has strengthened long-term sustainability, as reflected in a higher primary balance or lower primary spending after taking into account standard determinants of these variables. The underlying econometric model is attributable to Bohn (1998) and explains the primary balance or primary expenditure by its lagged term (to allow for persistence), lagged gross debt (to capture the long-term solvency constraint), and the output gap (to control for the cyclicality of fiscal policy). The regression model also takes into account the simultaneous existence of an expenditure rule and other rules (budget balance and debt rules). To reduce the selection bias that may be inherent when analyzing countries with expenditure rules exclusively, the econometric analysis relies on a broader and representative sample of 57 advanced and developing economies. In addition to countries with expenditure rules, this broader sample also includes comparable countries that have not introduced expenditure rules during the period of analysis (1985–2012).8

The results illustrate that countries with expenditure rules in addition to other rules exhibit, on average, higher primary balances (Table 12.2).9 Similarly, countries with expenditure rules also exhibit lower primary spending. The results are robust to the use of an alternative indicator for the presence of expenditure rules: the expenditure rule index. In addition to reporting whether a country has an expenditure rule in place, this index also captures the comprehensiveness of the rule. For instance, countries where the expenditure rule covers the general government instead of the central government will have a higher index.10 Moreover, the introduction of dummy variables to capture political pressure in election years suggests that expenditure rules can help preserve good fiscal performance. In particular, the estimated coefficients suggest that neither legislative nor executive elections have a significant effect on primary spending. Legislative elections, however, appear to worsen primary balances, probably through lower revenues. These results suggest two preliminary conclusions. First, expenditure rules appear to be effective at containing politically motivated spending pressures. Second, there is only partial evidence of electoral budget cycles, on the revenue side, once fiscal rules are controlled for.

Table 12.2.Expenditure Rules and Fiscal Performance
Dependent Variable
Primary Balance in Percent of GDPPrimary Expenditure in Percent of GDP
(1)(2)(3)(4)(5)(6)(7)(8)
Lag Dependent0.7910.7910.7960.7950.9030.9020.9030.902
(28.43)***(28.23)***(29.39)***(29.50)***(44.46)***(44.41)***(41.64)***(41.19)***
Lag Debt0.0130.0130.0130.013−0.011−0.012−0.011−0.012
(3.47)***(3.59)***(3.31)***(3.41)***(3.03)***(3.12)***(2.84)***(2.93)***
Output Gap0.0630.0610.0640.061−0.021−0.019−0.021−0.019
(2.36)**(2.28)**(2.46)**(2.39)**(0.73)(0.66)(0.77)(0.7)
ER Dummy0.8220.792−0.643−0.646
(2.75)***(2.55)**(2.07)**(1.95)*
ER Index0.2710.265−0.213−0.214
(2.50)**(2.34)**(1.87)*(1.73)*
Other Rules0.7810.640.7610.628−0.539−0.434−0.536−0.43
(3.35)***(3.14)***(3.53)***(3.14)***(2.17)**(1.98)**(2.34)**(2.03)**
LEGELEC−0.3−0.3070.0520.058
(2.08)**(2.13)**(0.34)(0.38)
EXELEC0.080.084−0.254−0.256
(0.34)(0.36)(1.03)(1.04)
Observations1,0851,0851,0781,0781,0851,0851,0781,078
Country5757575757575757
Source: FAD’s Fiscal Rules database; IMF World Economic Outlook database; and World Bank’s Political Institutions database. Note: The least squared dummy variable is corrected. Bootstrapped t-statistics are in parentheses. LEGELEC and EXELEC are dummy variables that take a value of 1 in years when legislative or executive elections were held. ER = expenditure rule.*p < .1; **p < .05; ***p < .01.
Source: FAD’s Fiscal Rules database; IMF World Economic Outlook database; and World Bank’s Political Institutions database. Note: The least squared dummy variable is corrected. Bootstrapped t-statistics are in parentheses. LEGELEC and EXELEC are dummy variables that take a value of 1 in years when legislative or executive elections were held. ER = expenditure rule.*p < .1; **p < .05; ***p < .01.

Event studies, which normalize the implementation date of each countrys expenditure rule to year t and use a simple measure of fiscal impulse to gauge the cyclical stance of fiscal policy, illustrate that fiscal policy tended to be countercyclical in the years following the introduction of an expenditure rule (Figure 12.9). For emerging markets, this sharply contrasts with the years preceding the introduction of a rule, when fiscal policy was procyclical on average. These results confirm, for a broader sample of advanced and emerging economies, existing findings for European economies (Wierts 2008; Holm-Hadulla, Hauptmeier, and Rother 2010). The cyclicality indicator in Figure 12.9 averages fiscal impulses, with procyclical impulses entering with a positive value and countercyclical impulses with a negative value. Specifically, procyclical impulses enter the indicator as improvements in the primary balance during bad times (when growth is below potential) and the negative of deteriorations in the primary balance during good times (when growth is above potential).

Figure 12.9.Fiscal Impulse

(Percent of GDP)

Sources: FAD Fiscal Rules database; and authors’ calculations.

Note: Procyclical fiscal impulses are measured by the improvement in the primary balance during good times. A higher value in the figure corresponds to a more procyclical policy. A negative value corresponds to countercyclical policy. The variable t indicates the year the rule was implemented.

Other Implications of Expenditure Rules

Expenditure Rules and Public Investment

Even the best-designed fiscal rules can have undesirable side effects. The most common relates to the risk that policymakers seek to achieve compliance by compressing certain high-quality discretionary items, such as public investment (see Blanchard and Giavazzi 2004). Although this may be an argument for excluding public investment from expenditure rules, there are potential drawbacks to limiting the rule’s coverage because it weakens the link with debt sustainability and opens the door to reclassification of spending items.11

Expenditure rules are associated with a significant decrease in investment in emerging economies only. Event studies similar to those described above show that, on average, investment spending falls across countries following the implementation of an expenditure rule (panel 1 of Figure 12.10). However, the result only passes the test of a panel regression for emerging economies (panel 2 of Figure 12.10).12 The presence of well-designed medium-term budgetary frameworks, which may be more common in advanced economies, could be a mitigating factor and ensure that capital spending is not cut merely to comply with expenditure ceilings in the short term.

Figure 12.10.Expenditure Rules and Spending Composition

Sources: FAD Fiscal Rules database; and authors’ calculations.

Note: Panel 1 shows an average across economies with expenditure rules in place. The variable f is for the first year the rule was implemented. Panel 2 shows coefficients from regressions of the investment share of spending on an expenditure rule dummy and other control variables, which include political variables from the World Bank database.

*p < .1; **p < .05.

Implications for Government Size and Efficiency

The primary objective of expenditure rules is to enhance fiscal sustainability, but there may be other side effects. These potential repercussions include a reduction in the size of government—which might in fact be the intended objective, particularly in some advanced economies—but also a reduction in the volatility of government expenditure, as result of a more medium-term orientation of the budget under expenditure rules. Lower volatility improves the predictability (and credibility) of policy and directly contributes to macroeconomic stability. Finally, expenditure rules may promote greater efficiency.

The data provide some evidence of possible implications for government size and efficiency. Event studies illustrate that the introduction of expenditure rules is indeed followed by smaller governments in both advanced and emerging economies (panel 1 of Figure 12.11).13

Figure 12.11.Expenditure Rules, Efficiency, and Government Size

Sources: Dabla-Norris and others 2012; FAD Fiscal Rules database; and authors’ calculations. Note: In panel 1, volatility is calculated as the absolute value of the percentage change in the deviation of expenditure from its trend, as calculated by the Hodrick-Prescott filter. In panel 2, the public investment efficiency score refers to the efficiency scores based on quantitative output estimated using frontier analysis in IMF 2015, which covers 132 economies across income groups. ER = expenditure rule.

The volatility of government spending is also found to decline after the introduction of an expenditure rule.14 The distribution of public investment efficiency scores15 has a higher average level and less dispersion in countries with expenditure rules (panel 2 of Figure 12.11). This outcome could be due to investment projects being prioritized more carefully relative to when there is no binding constraint on investment. However, this result needs to be interpreted with caution given the small share of countries (about 20 percent) with expenditure rules in the sample.

Expenditure Rules and Medium-Term Budgetary Frameworks

Finally, one last desirable side effect of expenditure rules is that they could encourage or foster desirable and complementary public financial management (PFM) reforms, including the introduction of a genuine medium-term budgetary framework (MTBF), and the strengthening of budget procedures, such as the adoption of top-down budgeting (Ayuso-i-Casals 2012). The data indeed suggest that the majority of countries strengthened their medium-term fiscal frameworks either when the expenditure rule was introduced or afterward (Figure 12.12).

Figure 12.12.Expenditure Rules and Medium-Term Fiscal Frameworks

(Number of countries)

Sources: Grigoli and others 2012; and FAD Fiscal Rules database.

Note: Strengthening the medium-term fiscal framework (MTFF) implies that a country moved from an MTFF to a medium-term budget framework (MTBF) or from an MTBF to a medium-term performance framework. ER = expenditure rule.

Conclusions

This chapter examines the effectiveness of expenditure rules using the fiscal rules data set developed by Schaechter and others (2012). Overall, its findings lend support to the view that expenditure rules can foster better spending behavior if sound PFM systems are in place. Specifically, the analysis shows the following:

  • The compliance rate for expenditure rules is greater than that for budget balance rules, particularly if the expenditure rule is directly under the control of the government and the rule is enshrined in law or in a coalition agreement.

  • The presence of expenditure rules is associated with stronger fiscal performance, that is, a higher primary balance—after taking into account conventional determinants—and countercyclical policies.

  • Expenditure rules are associated with lower levels of public investment in emerging market economies, where weaker PFM systems may be less effective at preventing policymakers from deferring high-quality discretionary spending for the sake of complying with the rule.

These results need to be interpreted with caution given the relatively limited experience with expenditure rules (33 rules during 1985–2013). Moreover, in most cases, expenditure rules are generally used in conjunction with budget balance or debt rules, or both. While these combinations are needed to link the rule to debt sustainability, such simultaneity complicates identification of their impact.

Annex 12.1. Economy List for Regressions on Expenditure Rules and Fiscal Performance
Table 12.1.1.Economies Included in Regressions
Advanced EconomiesEmerging Market and Developing Economies
AustraliaItalyArgentinaMalaysia
AustriaJapanBrazilMexico
BelgiumKoreaBulgariaMoldova
CanadaNetherlandsChileNicaragua
CyprusNew ZealandChinaPeru
Czech RepublicNorwayColombiaPhilippines
DenmarkPortugalEgyptPoland
EstoniaSingaporeHaitiRomania
FinlandSlovak RepublicHondurasRussia
FranceSloveniaHungarySouth Africa
GermanySpainIndiaThailand
GreeceSwedenIndonesiaTurkey
Hong Kong SARSwitzerlandLatviaUkraine
IrelandUnited KingdomLithuania
IsraelUnited States
Source: IMF, World Economic Outlook database.
Source: IMF, World Economic Outlook database.
Annex 12.2. Additional Regression Results
Table 12.2.1.Fixed Effects Panel Regressions 1980–2010, Capital Spending as a Share of Total Spending as the Dependent Variable1(Percent of GDP)
All EconomiesAdvanced EconomiesEmerging Market Economies
Expenditure Rule Dummy−1.8−0.01−4.9
(1.0)*(0.8)(2.3)**
EXELEC0.60.21.3
(1.4)(1.4)(2.9)
LEGELEC−0.10.4−1.0
(0.9)(0.7)(2.5)
Constant8.42.717.4
(0.6)***(0.6)***(1.4)***
R2 (overall)0.020.000.01
Observations490274176
Number of Countries251311
Expenditure Rule Dummy−1.8−0.01−4.9
Sources: FAD Fiscal Rules database; IMF, World Economic Outlook database; and World Bank database of Political Institutions.Note: LEGELEC = 1 if there was a legislative election in this year. EXELEC = 1 if there was an executive election in this year. Robust standard errors are in parentheses.

A Hausman (1978) test was conducted to check whether a fixed-effects model is preferable to a random-effects model. The hypothesis that the individual-level effects are adequately captured by a random effects model can be rejected at the 1 percent level of significance.

*p < .1; **p < .05; ***p < .01.

Sources: FAD Fiscal Rules database; IMF, World Economic Outlook database; and World Bank database of Political Institutions.Note: LEGELEC = 1 if there was a legislative election in this year. EXELEC = 1 if there was an executive election in this year. Robust standard errors are in parentheses.

A Hausman (1978) test was conducted to check whether a fixed-effects model is preferable to a random-effects model. The hypothesis that the individual-level effects are adequately captured by a random effects model can be rejected at the 1 percent level of significance.

*p < .1; **p < .05; ***p < .01.

Annex 12.3. Expenditure Rules Details
Annex Table 12.3.1.Expenditure Rules Details
CountryStatutory BaseCoverageTarget/ConstraintDescriptionTime FrameEscape ClauseIndependent Body Monitors ImplementationFRL
ArgentinaLGGChange to GDPPrimary expenditure cannot grow more than nominal GDP or, at most, stay constant in periods of negative nominal GDP growth.2000−08NoYesYes
AustraliaLCGReal expenditure growth rateReal growth in spending is constrained to 2 percent a year once the economy recovers and grows above trend. Once the budget returns to surplus, and while the economy is growing at or above trend, the government will maintain expenditure restraint by retaining a 2 percent annual cap on real spending growth, on average, until surpluses are at least 1 percent of GDP.2009−NoNoYes
AustraliaPCCGChange to GDPGovernment expenditure cannot be raised as a proportion of GDP in 1985–86 and over the life of the parliament.1986−88NoNoNo
BelgiumCACGReal expenditure growth rateReal growth of primary expenditure of CG ought to be equal or be less than 0 percent.1993−97NoYesNo
BotswanaLCGLevel to GDPThe ceiling on the expenditure-to-GDP ratio is 40 percent. Also, 30 percent of total expenditures should be directed toward development spending, which includes all capital spending and the recurrent spending for health and education.2003−NoNoNo
BrazilLGGLevel to GDP[1) Personnel expenditure is limited to 50 percent of net current revenue for the federal government, and 60 percent for states and municipalities; (2) permanent spending mandates cannot be created without permanent revenue increases or spending cuts; (3) the government sets numerical multiyear targets for expenditures (for the current year and indicative targets for the next two years).2001−YesNoYes
BulgariaLGGLevel to GDPThe ceiling on the expenditure-to-GDP ratio is 40 percent.2012−NoNoNo
BulgariaPCGGLevel to GDPThe ceiling on the expenditure-to-GDP ratio is 40 percent.2006−09NoNoNo
CanadaLCGSpecific ceilingFederal spending control act sets clear nominal expenditure limits from FY91/92 to FY95/96.1992−96NoYesNo
CroatiaLGGChange to GDPThe temporary rule calls for general government expenditure cuts of 1 percent of GDP a year until at least a primary balance of zero is achieved in nominal terms.2012−YesYesYes
DenmarkPCGGReal expenditure growth rateReal growth in public expenditures cannot exceed potential GDP growth, which is a (rough) measure of structural development in the tax base. If growth in expenditures increases beyond potential GDP growth, it must be financed by specific discretionary measures, which increase revenues.2012−NoNoNo
DenmarkPCGGLevel to GDPPublic consumption as a share of cyclically adjusted GDP should be reduced to 26.5 percent by 2015.2009−11NoNoNo
DenmarkPCGGReal expenditure growth rateTarget of public consumption as a percentage of cyclically adjusted GDP and real growth in public consumption.2007−08NoNoNo
DenmarkPCGGReal expenditure growth rateReal public consumption growth capped at 0.5 percent per year, 1.0 percent during 2002-05.1994−2006NoNoNo
EcuadorLGGOtherPermanent expenditure cannot be higher than permanent revenue, though both are unclearly defined.2011−NoNoYes
FinlandCACGSpecific ceilingAnnual limits to government expenditure for the 4-year terms of office of the government. Limits are set in real terms for primary noncyclical expenditures (about 75 percent of total central government spending, about 37 percent of total general government spending).2003−NoNoNo
FranceLC2011−); PC (1998−2010)CGReal expenditure growth rateTargeted increase of expenditure in real terms, or targeted increase of expenditure excluding interest payments and pensions in nominal terms. The stricter provision applies.1998−NoYesNo
HungaryLGGReal expenditure growth rateCap real expenditure growth.2010−11NoYesYes
IcelandPCCGReal expenditure growth rateThe real expenditure growth limit of the central government (2 percent for public consumption and 2.5 percent for transfers).2004−08NoNoNo
IsraelLCGReal expenditure growth rateThe provision for limiting real growth of the central government fiscal expenditure is 1.7 percent from 2007. For the biannual budget, adopted July 2009, the rules were relaxed to allow a real growth of expenditure of 3 percent for 2009. The Deficit Reduction and Budgetary Expenditure Limitation Laws (2010) make spending growth a function of public debt—rising, as the gap falls between actual debt and the objective of reducing it to 60 percent of GDP; and rising with trend GDP—measured as a 10-year moving average—and with projected inflation. This formula caps real spending growth in 2011 at 2.6 percent.2005−NoYes (since 2009)No
JapanPCCGSpecific ceilingThe “overall expenditure limit” (the amount of the general account expenditure, excluding debt repayment and interest payment) should not exceed that of the previous fiscal year. Reconstruction-related expenditures shall be managed separately from other expenditures, accompanied with their financial resources (cutting other expenditures, nontax revenues including sales of government’s assets, and tax revenues by special taxes for reconstruction).2011−NoNoNo
JapanPCCGSpecific ceilingIn 2006, the government set numerical targets (cabinet decision) by spending category (for example, public investment, social security, and so on). The 2006 targets were intended to be valid through FY11/12 and indeed were valid for FY07/08 and FY08/09 budgets. But the targets were abandoned for FY09/10 due to the financial crisis.2007−09NoNoNo
KosovoPCGGReal expenditure growth rateCeiling on current expenditure growth of 0.5 percent per year in real terms.2006−08NoNoNo
LithuaniaLGGOtherIf the GG budgets recorded a deficit on average over the past 5 years, the annual growth of the budget appropriations may not exceed one half of (or 0.5 times) the average growth rate of the budget revenue of those 5 years.2009-NoNoNo
LuxembourgCACGChange to GDPIn the course of the legislative period (per coalition agreement), public expenditure growth is maintained at a rate compatible with the medium-term economic growth prospects, which are quantified. Since 2010, the target is to bring expenditure growth back to the medium-term growth prospects, once the countercyclical response to the crisis has been phased out.1995−NoNoNo
MongoliaLCGChange to GDPThe expenditure growth cannot exceed the growth of nonmineral GDP from 2013.2013−YesNoYes
NamibiaCACGLevel to GDPPublic expenditure levels below 30 percent of GDP2011−NoNoNo
NetherlandsCAGGSpecific ceilingThe real expenditure ceilings are fixed for total expenditure (covering CG, health care, and social security; covers about 90 percent of GG expenditure) and sectoral expenditure for each year of government’s four-year office term. Coverage of expenditure was changed in recent years: from 2007-10 interest payments were excluded; since 2009, expenditure is defined in net terms, that is, gross expenditure minus nontax revenues, from 2009-10 expenditure excluded unemployment and social assistance benefits. If overruns are forecast, the Minister of Finance proposes corrective action.1994−NoNoNo
PeruLCGReal expenditure growth rateThe real growth current expenditure ceiling was 2 percent (2000-02), 3 percent (2003-08), and 4 percent since 2009. Since April 2012, infrastructure maintenance is excluded from the expenditure cap and current expenditures associated with some social programs and equipment for military and police forces.2000−YesNoYes
PolandLCGReal expenditure growth rate Change to GDPThe overall increase in CG discretionary spending and all newly enacted spending cannot exceed 1 pps in real terms, based on CPI inflation (defined in the Public Finance Act as a temporary rule but envisaged to be replaced by a permanent rule once the excessive deficit procedure has been abrogated).2011–NoNoNo
RomaniaLGGReal expenditure growth rate Change to GDPThe total GG expenditure growth should not exceed projected nominal GDP for next three years until budget balance is in surplus. Moreover, personnel expenditure limits are binding for two years as set out in MTBF.2010–YesYesYes
RussiaLCGOtherThe rule sets a ceiling on expenditures (oil revenue at the “base” oil price, plus all nonoil revenues, plus a net borrowing limit of 1 percent of GDP). Oil revenues above the “base” oil price need to be saved in the Reserve Fund until it reaches 7 percent of GDP (though there are some allowable exceptions to this under the law).2013-YesYesYes
SpainLCGChange to GDPThe nominal expenditure growth for central and local governments shall not exceed Spain’s nominal medium-term GDP growth. Interest and nondiscretional expenditure on unemployment benefits are excluded.2011-NoNoYes
SwedenL(2010-); CA (1997-2009)CG+SSSpecific ceilingThe nominal expenditure ceiling for CG and the pension system is set for a 3-year period with the outer year added annually. Ceilings cannot be adjusted except when there are technical issues. A budgetary margin is used as a buffer. Interest expenditure is excluded from the ceiling.1997-NoYes (since 2007)No
United StatesLCGSpecific ceilingIn August 2011, Congress enacted discretionary spending caps, saving about $900 billion over the next decade. As a result of the failure to adopt a medium-term comprehensive deficit reduction plan, additional spending cuts (the so-called sequester) came into effect in March 2013. These additional cuts, if not repealed by Congress, will produce savings of US$1.2 trillion over a decade, with one-half of the savings coming from defense spending and the other half from domestic programs (excluding Social Security, Medicaid, parts of Medicare, and certain other entitlement programs).2011-NoYesNo
United StatesLCGSpecific ceilingThe annual appropriations limit adopted under the Budget Enforcement Act of 1990 for discretionary spending was allowed to lapse at the end of FY02/03.1991-2002NoYesNo
Source: FAD Fiscal Rules database 2013.Notes: CA = coalition agreement; CG = central government; CPI = consumer price index; FRL = Fiscal Responsibility Law; GG = general government; L= legal; PC = political commitment; pps = percentage points; SS = social security
Source: FAD Fiscal Rules database 2013.Notes: CA = coalition agreement; CG = central government; CPI = consumer price index; FRL = Fiscal Responsibility Law; GG = general government; L= legal; PC = political commitment; pps = percentage points; SS = social security
References

    Ayuso-i-CasalsJ.2012. “National Expenditure Rules: Why, How and When.Economic Papers473European CommissionBrussels.

    Banque Centraledu Luxembourg. 2005. “La situation économique et financiére.BCL Bulletin 2005/3Luxembourg CityLuxembourg.

    BlanchardO. J. and F.Giavazzi. 2004. “Improving the SGP through a Proper Accounting of Public Investment.Discussion Paper 4220Centre for Economic Policy ResearchLondon.

    BohnH.1998. “The Behavior of U.S. Public Debt and Deficits.Quarterly Journal of Economics113 (3): 94963.

    Congressional Budget Office (CBO). 2003. “An Analysis of the President’s Budgetary Proposal for Fiscal Year 2004.CBOWashington, DC. http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/41xx/doc4129/03–31-analysispresidentbudget-final.pdf.

    CottarelliC.2012. “Beyond the Austerity Debate: The Deficit Bias in the Post-Bretton Woods Era.iMF Direct Blog PostMay21. https://blog-imfdirect.imf.org/2012/05/21/beyond-the-austerity-debate-the-deficit-bias-in-the-post-bretton-woods-era/.

    Dabla-NorrisE. J.A.BrumbyZ. M.Kyobe and C.Papageorgiou. 2012. “Investing in Public Investment: An Index of Public Investment Efficiency.Journal of Economic Growth17 (3): 23566.

    DebrunX.2014. “How Expenditure Rules Can Help Get Public Spending Right.Public Financial Management Blog PostJune30. http://blog-pfm.imf.org/pfmblog/2014/06/how-expenditure-rules-can-help-public-spending-right.html.

    DebrunX.L.MoulinA.TurriniJ.Ayuso-i-Casals and M. S.Kumar. 2008. “Tied to the Mast? The Role of National Fiscal Rules in the European Union.Economic Policy23 (54): 297362.

    EscolanoJ.A.Shabunina and J.Woo. 2011. “The Puzzle of Persistently Negative Interest Rate-Growth Differentials: Financial Repression or Income Catch-Up?Working Paper 11/260International Monetary FundWashington, DC.

    GrigoliF.Z.MillsM.Verhoeven and R.Vlaicu. 2012. “MTEFs and Fiscal Performance.” Policy Research Paper 6186 World Bank Washington DC. Gunnarsson G. 2011. “Reforming Iceland’s Fiscal Framework.Unpublished paper Banca d’ItaltiaRome, Italy. https://www.bancaditalia.it/pubblicazioni/altri-atti-convegni/2011-rules-institution/Gunnarsson.pdf?language_id=1.

    HallerbergM.R.Strauch and J.von Hagen. 2007. “The Design of Fiscal Rules and Forms of Governance in European Union Countries.European Journal of Political Economy23 (2): 33859.

    HallerbergM.R.Strauch and S.Yläoutinen. 2010. “Political Power, Fiscal Institutions and Budgetary Outcomes in Central and Eastern Europe.Journal of Public Policy30 (1): 4562.

    HausmanJ.1978. “Specification Tests in Econometrics.Econometrica46 (6): 125171.

    Holm-HadullaF.S.Hauptmeier and P.Rother. 2010. “The Impact of Numerical Expenditure Rules on Budgetary Discipline over the Cycle.Working Paper Series 1169European Central BankBrussels.

    International Monetary Fund (IMF). 2009. “Fiscal Rules—Anchoring Expectations for Sustainable Public Finances.IMFWashington, DC. https://www.imf.org/external/np/pp/eng/2009/121609.pdf.

    International Monetary Fund (IMF). 2012. “Botswana: 2012 Article IV Consultation.Washington, DC.

    International Monetary Fund (IMF). 2014. Fiscal Monitor-—Public Expenditure Reforms: Making Difficult Choices.Washington, DCOctober.

    International Monetary Fund (IMF). 2015. “Making Public Investment More Efficient.Policy PaperIMFWashington, DC.

    LjungmanG.2008. “Expenditure Ceilings: A Survey.Working Paper 08/282IMFWashington, DC.

    NerlichC. and W. H.Reuter. 2013. “The Design of National Fiscal Frameworks and Their Budgetary Impact.Working Paper 1588European Central BankFrankfurt.

    SchaechterA.T.KindaN.Budina and A.Weber. 2012. “Fiscal Rules in Response to the Crisis—Toward the ‘Next-Generation’ Rules. A New Dataset.Working Paper 12/187IMFWashington, DC.

    SchuknechtL.2004. “EU Fiscal Rules: Issues and Lessons from Political Economy.Working Paper 421European Central BankFrankfurt.

    TurriniA.2008. “Fiscal Policy and the Cycle in the Euro Area: The Role of Government Revenue and Expenditure.European Economy Economic Papers 323 Directorate-General for Economic and Financial Affairs European Commission Brussels.

    WiertsP.2008. “How Do Expenditure Rules Affect Fiscal Behaviour?Working Paper 166 De Nederlandsche Bank Amsterdam.

The authors thank Nate Arnold, Nina Budina, Xavier Debrun, Julio Escolano, Johannes Eugster, Nan Geng, Sanjeev Gupta, Michael Gorbanyov, Martine Guerguil, Carlos Mulas-Granados, Joana Pereira, Marta Ruiz-Arranz, Svetlana Vtyurina, and participants at the FAD seminar series for helpful comments and discussions. The authors are also grateful to Andrea Schaechter for her encouragements and initial guidance. The main results of this paper were published in the IMF’s April 2014 Fiscal Monitor and in an IMF working paper in 2015. Ethan Alt and Nathalie Carcenac provided excellent research assistance.

In the European Union, for example, national expenditure rules have been reinforced through inclusion in the “Six-Pack” of an expenditure benchmark to reinforce the preventive arm of the Stability and Growth Pact. Under the expenditure benchmark rule, public spending is not allowed to increase faster than medium-term potential GDP growth unless it is matched by adequate revenues.

Using the same definition, a total of 33 expenditure rules were identified as being in place in 29 countries for some or all of the years during the period 1985–2013. Out of the 33, 10 had been dropped by 2013 and two were adopted in 2013. Annex 12.3 provides an overview of each expenditure rule.

In the EU-27 countries, the supranational Maastricht budget balance and debt rules also apply.

The expenditure rule in Sweden covers the central government and the pension system.

These include formal sanctions, which are often part of fiscal responsibility laws and automatic correction mechanisms.

Details on compliance scores are available from the authors upon request.

The question of fiscal rules and incentives for better fiscal performance is taken up in the next section.

Annex Table 12.1.1 in Annex 12.1 provides the list of countries included in the regressions in Table 12.2.

As mentioned above, the empirical analysis is subject to a caveat that applies to any empirical study of the impact of institutions on policies: reverse causality. In the absence of convincing instruments, we cannot exclude the possibility that the presence of fiscal rules reflects deep social preferences that would be the true cause of strong outcomes.

The expenditure rule index captures the comprehensiveness of expenditure rules by aggregating their key features such as coverage, legal basis, and formal enforcement procedure. See Schaechter and others (2012) for details on the methodology for constructing a similar fiscal rules index.

Including through the creation of new expenditure categories.

Details of the panel regression can be found in Annex 12.2.

Given that most expenditure rules were implemented in the context of and perhaps part of the response to “bad” times (see Figure 12.5), those patterns may not only be due to the introduction of an expenditure rule. They could also reflect that over the course of the recovery, the share of government spending in GDP typically falls and volatility declines. Nonetheless, introducing an expenditure rule following a rise in spending during recessions can ensure that the expenditure-to-GDP ratio does not stay at an elevated level because of political pressures during the recovery.

Following Grigoli and others (2012), spending volatility is calculated as the absolute value of the percentage change in the deviation of expenditure from its trend as calculated by the Hodrick-Prescott filter.

The scores are obtained from frontier method estimates based on physical output (IMF 2015). Specifically, an output-oriented data envelopment analysis is used to gauge the efficiency in transforming public investment into physical infrastructure. The data coverage is for 132 countries from 2000 onward.

    Other Resources Citing This Publication