Fiscal Politics

Chapter 16. Fiscal Watchdogs and Sound Fiscal Policy: Is the Barking Loud Enough to Tame Politicians?

Vitor Gaspar, Sanjeev Gupta, and Carlos Mulas-Granados
Published Date:
April 2017
  • ShareShare
Show Summary Details
Xavier Debrun, Marc Gérard and Jason Harris 

This chapter revisits the potential contribution of politically independent fiscal watchdogs (fiscal councils, or FCs) to improving fiscal performance. A simple theoretical model first illustrates that FCs cannot credibly exert a direct constraint on day-to-day policy choices. However, by contributing to the broader public debate on fiscal policy—through the provision of unbiased quantitative and qualitative analysis, forecasts, and possibly recommendations—these institutions can reduce informational asymmetries hindering voters’ ability to reward good policies and penalize bad ones. The chapter explores the empirical relevance of this argument by looking at the media impact of FCs in relation to real-time fiscal developments. It appears that FCs’ activity and media impact increase in times of budget slippages or relative fiscal activism, a necessary condition for the validity of the theory. However, FCs’ media impact is only weakly correlated with subsequent policy changes.


The fiscal legacy of the 2008–09 economic and financial crisis brought to the fore serious concerns about the capacity of governments to meet their obligations in full. Several vulnerable governments came under severe market pressure, including in countries considered to be safe at the time. In the euro area, repeated and sometimes acute bouts of panic seized sovereign debt markets, raising the specter of widespread self-fulfilling solvency crises in countries with otherwise manageable public debt dynamics.

The need for improved fiscal governance quickly emerged as an essential part of any solution to mitigate elevated risks of fiscal crisis. A sound budgeting framework indeed conveys useful information about the likely range of future fiscal outcomes. Hence, by making bad states of public finances less likely in the medium term, markets are less easily spooked by adverse fiscal or macroeconomic shocks, while policymakers can respond more flexibly to such shocks. That is the magic bullet of fiscal credibility.

A considerable literature confirms the robust association between strong fiscal performance and good fiscal institutions. If the past is a good predictor of the future, institutional reforms could anchor a credible commitment to sound fiscal policies and reduce the short-term risk of a bad market equilibrium. Better fiscal governance comes at an additional premium in the euro area, where the risk sharing implied by crisis management measures calls for safeguards against moral hazard.

Fiscal institutions conducive to macroeconomic stability often come in the form of fiscal policy rules. Quantitative limits on debts, deficits, or spending have long been used to contain fiscal profligacy (for example, Fabrizio and Mody 2006; Debrun and others 2008). Yet experience has revealed serious limitations often related to the rules’ inflexibility in the face of adverse or unusual circumstances, the lack of supportive budget procedures, or weak political commitment to effectively enforce them. Inflexibility ultimately threatens the credibility of the rule itself because the pressure of events can quickly lead to its suspension or even elimination.

Since the mid-1990s, a growing literature has argued that nonpartisan watchdogs—often dubbed fiscal councils or independent fiscal institutions—could shape policymakers’ incentives in a more credible and effective way than numerical limits on budgetary aggregates. Through independent analysis, assessments, and forecasts, such bodies would raise voters’ awareness of the consequences of certain policy paths, helping them reward desirable actions and sanction toxic ones (Kopits 2013; Debrun and others 2013). By barking loud enough in the face of undesirable behavior, fiscal watchdogs could foster democratic accountability and fiscal soundness. And instead of rigid rules, they could help devise an adequate policy response in most circumstances without undermining confidence in governments’ ability to keep public finances on a sustainable track. The argument gains particular traction in times of extreme shocks and crises, when policy flexibility and credibility are both highly valued.

Establishing an independent FC now figures prominently in characterizations of sound fiscal policy frameworks well beyond the narrow circle of public financial management experts (Deutsche Bank 2016). Recent reforms of fiscal governance in the European Union (EU) now mandate independent institutions to assess the quality of budgetary forecasts and to monitor compliance with national fiscal rules. At the EU level, an Advisory Fiscal Board has been created to monitor implementation of the EU fiscal framework, advise the European Commission on the euro-area-wide fiscal stance, and facilitate coordination among national independent fiscal institutions.

Despite a long-standing and active debate in academic and policy circles, only recently has the economic analysis of FCs developed beyond sketchy and largely informal policy papers and opinion pieces (Calmfors 2010; Calmfors and Wren-Lewis 2011; Debrun, Hauner, and Kumar 2009; Kopits 2013; Debrun and others 2013). Economic theory discussing the desirability and effectiveness of such institutions remains in its infancy (Debrun 2011; Beetsma and Debrun 2016a, 2016b), and systematic empirical evidence exploring the link between these institutions and fiscal behavior is limited by the very short lifespan of most FCs (Debrun and others 2013). As a result, no definitive consensus exists with regard to the tasks FCs should be assigned, what institutional form they should take, and their complementarity with or instead substitutability for rules-based frameworks.

This chapter is an effort to fill some of those gaps. The second section sketches a highly stylized and purely illustrative model to anchor the debate on politically independent fiscal agencies. An important aspect is that the effectiveness of FCs rests on their ability to address the root cause rather than the symptoms of deficit bias. The third section gathers empirical evidence on the effectiveness of existing FCs, looking at two testable hypotheses emanating from theory. First, councils can only credibly affect policymakers’ incentives and actions if they systematically influence the public debate on fiscal policy. In other words, does the watchdog bark when the risk of fiscal misbehavior looms large? Using data originally collected for case studies in Curristine, Harris, and Seiwald (2013), the chapter assesses the media impact of FCs at times when we would expect them to speak out, that is, in the aftermath of budgetary slippages or policy shifts. Second, the analysis tests whether desirable policy changes follow peaks in media impact, which would be consistent with FCs’ ability to encourage better fiscal behavior. In other words, are political decision makers sufficiently impressed by the barking to correct undesirable actions? Finally, the chapter emulates other recent pieces—especially Calmfors and Wren-Lewis (2011)—by gathering more forensic evidence on what FCs do and achieve. Concluding remarks form the fourth and last section of the chapter.

Fiscal Policy, Fiscal Councils, and Democratic Accountability

Since the late 1980s, a large literature has explored the reasons macroeconomic policies tend to deviate from a well-defined social optimum (see Kumar and Ter-Minassian [2007] for an extensive survey). On the fiscal side, the hypothesis of a deeply entrenched deficit bias emerged from the seemingly inexorable rise in debt-to-GDP ratios since the mid-1970s. Today’s public debt levels are often unprecedented for peacetimes. While theories of deficit bias abound, empirical analyses have failed to identify a dominant explanation.1 Because this literature has been surveyed elsewhere (for example, Calmfors 2010; Debrun and others 2008; Hagemann 2010), this chapter is limited to presenting a highly stylized two-period model of fiscal policy whose main virtue is to illustrate how FCs can help improve fiscal performance without assuming the delegation of policy instruments inherent to the theory of central bank independence (Wyplosz 2005).

A Simple Illustrative Model

Consider the two-period model of Alesina and Tabellini (1990), assuming constant and deterministic income levels. Identical private agents maximize utility U, which is separable over time and types of goods (private and public):

where c denotes the per capita consumption of private goods and q the per capita consumption of public goods. The term E0 symbolizes the expectations operator conditional on information available at the beginning of period 1 (time 0), and β is a subjective discount factor. Assuming a constant proportional income tax rate t, resource constraints simply write:

with R being the interest factor and l the stock of net private liabilities at the end of period 1.

Elected officials decide on public goods provision. They belong to one of two political parties (C or L) indexed by Q. Preferences are identical across political parties and to those of the population, but officials only value public goods when in office. These assumptions avoid the needless complexity of a partisan cycle in the conduct of fiscal policy, leading to a simple and well-defined deficit bias:

with vC(qt) = 0 if party L is in office, and vL(qt) = 0 if party C is in office. Elections with uncertain outcome take place at the end of period 1, and the parameter ρt–1 captures the probability of the incumbent party being in office at period t.

The government’s resource constraints determine the amount of public goods per capita delivered in each period:

with b denoting the overall deficit at the end of period 1 (or equivalently, the principal of the debt to be repaid in period 2). The resource constraints are subject to random shocks affecting government efficiency. For a given amount of resources (tax revenue and borrowing), a positive realization of δ negatively affects public goods delivery. Concretely, δ could capture resource diversion by corrupt civil servants, the effect of poor administrative capacities, or unforeseeable policy mistakes. Of course, good surprises can also occur (more public goods being delivered with the same budgetary envelope). Hence, the shocks are nonserially correlated, with zero mean and finite variance: δtN(0,σδ2). Also, we assume that fiscal policy decisions have no impact on income and, thereby, on private consumption since R is given and independent of b. Hence, these shocks capture the effect of good (or bad) luck on policymakers’ performance.

In this model, a deficit bias2 emerges because elected officials are uncertain about reelection. This result is immediately clear from a comparison of equations (16.1) and (16.2): any ρ<1 entails policymakers’ myopia (a policymaker discounts the future at a higher rate than a representative agent). While uncertainty is often exogenously given (for example, Beetsma and Debrun 2007), we argue that informational asymmetries between voters and policymakers can be instrumental in making election outcomes uncertain and are therefore at the root of the deficit bias. Specifically, this chapter makes the plausible assumption that voters cannot know for sure whether a given outcome for public goods delivery reflects the intrinsic competence of the elected policymaker or an exogenous event outside the policymaker’s control affecting efficiency in public goods delivery (luck). The unobservable shocks affecting public goods delivery can thus lead voters to vote out (punish) competent officials or reelect (reward) undeserving individuals.

Characterizing the Deficit Bias

The socially optimal solution results from direct maximization of the representative citizen’s utility (1) by a benevolent “social planner.” To economize on notation, we set β = R = 1 (discount and real interest rates are equal to zero) and assume quadratic utility functions u(x)=v(x)=(xx˜)2. Decision makers dislike deviations from predetermined objectives, denoted by a tilde. The Euler equation under the social planner yields a balanced budget:

Before deriving the political equilibrium—that is, elected policymakers’ choice—let us clarify the sequence of moves. First, “Nature” draws the governing party (C by assumption here). Second, party C officials prepare a budget, setting the deficit for period 1, and by extension, the expected time path of public consumption over the two periods. Third, an efficiency shock materializes during period 1, and finally, elections take place. In period 2, all debts are paid off after a new shock occurs. Solving this problem by backward induction rules out time inconsistency.

As noted above, the probability of reelection is a central determinant of the budget deficit in the political equilibrium, denoted by a ** superscript:

Certainty about election outcomes defines two boundary cases. Certain reelection (ρ = 1) eliminates myopia, leading party C officials to opt for a balanced budget: b**|ρ=1 = 0 = b*. By contrast, certain defeat maximizes myopia to the point that party C is not bound by the intertemporal budget constraint and chooses a level of public spending consistent with the expected delivery of q˜ in period 1. The corresponding budget deficit is b**|ρ=0=b˜. All other solutions fall in the [0,b˜] interval. Myopic policymakers generate a deficit bias only if b˜>0, which requires that the appetite for delivering public goods (parameterized by q˜) exceed available tax money. This condition can be interpreted as the common pool problem inherent to budgetary decisions so that b** > b*. It is important to note that because voters’ behavior remains purely exogenous, this deficit bias emerges in a political setting without formal democratic accountability.

Solving the Deficit Bias: Fiscal Rules versus Fiscal Councils

This section compares how fiscal rules and fiscal councils can affect policy outcomes.

Balanced Budget Rule

A straightforward solution to the deficit bias could be a balanced budget requirement. A fiscal rule affects policymakers’ decisions to the extent that violating it entails a certain cost measurable in terms of utility. The costs can be merely reputational or result from a formal enforcement procedure with explicit sanctions (Beetsma and Debrun 2007). Under a fiscal rule, the “constrained” utility of the elected official would be

Policymakers now maximize VC instead of UC, as they internalize some costs of exceeding the deficit ceiling b*. The optimal fiscal rule is such that f*=(1ρ)b˜ (the marginal cost of deviations), which implements b* in the political equilibrium (b** = b*). The optimal rule thus imposes higher costs of breaching the deficit cap in countries where political instability (lower ρ) and the common pool problem (higher b˜) are more severe.

Of course, showing that an optimal rule exists does not mean that policymakers have any incentive to set it up in the first place.3 In fact, it is straightforward to establish that the rule (f*, b*) violates the participation constraints of the policymakers as E0VC(q*)<E0UC(q**). Hence, even if policymakers inherit the rule from benevolent founding fathers, they will have an incentive to make it irrelevant or to scrap it altogether. Time inconsistency destroys the credibility of the rule (Debrun and Kumar 2009), which explains in part why these arrangements periodically come under intense pressure and are eliminated, substantially modified, circumvented, or temporarily ignored.4

Note that the time-inconsistency problem as characterized here may, a priori, be less severe in the face of supranational fiscal rules because such rules require international coordination to be changed. However, the experience with the Stability and Growth Pact in the EU suggests that supranational rules are not immune to change and circumvention. The pact has been thoroughly amended twice (in 2005 and 2010) since its introduction in 1997; its implementation has been uneven across countries, with large and influential member states seemingly benefiting from greater leniency;5 and it has encouraged certain countries to mask the true state of public finances through sometimes egregious accounting manipulations.

Fiscal Councils

The main lesson from the above exercise is that any mechanism aimed at directly constraining fiscal discretion is bound to be resisted by policymakers and is therefore at high risk of being weakened or dismantled as soon as the opportunity arises. Now the question is whether nonpartisan fiscal agencies could at least appeal to policymakers (that is, satisfy their participation constraint) and help correct the deficit bias inherent to the political equilibrium.

The answer is arguably positive if we think of the FC as an institution that can induce meaningful rewards for policymakers who pursue policies closer to the social optimum. To illustrate this, the basic requirement that elected officials be accountable to their principal (the voters) is incorporated in the model. An immediate task is therefore to show that the deficit bias persists even if voters can credibly sanction inadequate policies and vote out an incumbent perceived to be unable to deliver enough public goods given available resources.

The failure of accountable governments to deliver the social optimum can be linked to informational asymmetries discussed earlier: voters’ preferences may be hard to read, and the abilities and true agenda of those running for office are unobservable to voters. As a result, policy mistakes (or successes) can be difficult to detect, interpret, and adequately sanction (or reward) because they are indistinguishable from the effect of luck.

Thus, for the sake of the argument, two basic assumptions are made. First, voters cannot observe the intrinsic competence of policymakers nor the shocks to public goods delivery. Second, opaque public accounts prevent voters from assessing whether observable outcomes are the result of either luck or competent policymaking. Formally, this means that the efficiency shocks δt and the true level of debt (deficit) are unobservable after the fact. Only tax revenues and actual output of public goods are perfectly observed. (Voters only know for sure what they pay and what they get.) The combination of unobservable competence and observable outcomes implies that voters will use the size of q1 as a signal of competence.6 In the stylized setup in this analysis, it effectively means that if they could observe δ1, voters would interpret it as the incumbent’s capacity to deliver public goods given fixed budgetary resources.

Rational voters can thus only observe a “subjective” budget balance—that is, the difference between the public goods they get and the tax money they pay. That measure reflects a shock (δ1) and noise due to imperfect fiscal transparency (ξ):

where q1C denotes voters’ perceived value of the public goods they consume—which also reflects the true budget balance. To infer the incumbent’s efficiency or competence (which boils down to δ1), voters solve a basic signal-extraction problem at the end of period 1:

Because voters’ best guess of δ1 reflects their assessment of the incumbent’s competence, voters will form beliefs about the incumbent’s competence on the basis of E11], which amounts to treating δ as persistent: E12|Q = C] = λE11] with 0 < λ ≤ 1, while E12|Q = L] = 0.

The incumbent will be reelected if voters expect her to deliver more public goods given the remaining budgetary resources (which are the same for both parties). Formally, reelection occurs if E1[q2|Q = C] − E1[q2|Q = L] > 0 . At time 0 (budget preparation stage for period 1), the perceived probability of reelection is therefore ρ=Pr(λσδ2σδ2+σξ2(q1Cτy)>0).

This shows that political incentives leading to a deficit bias remain even after explicitly introducing democratic accountability in the model. The difference with the section “Characterizing the Deficit Bias” above is that because the probability of reelection depends positively on the (ex post) realization of the budget balance, policymakers now face an opportunistic (ex ante) motive to run a deficit. Indeed, a deficit reduces the likelihood that voters will detect adverse efficiency realizations at the end of period 1, which they would interpret as a lack of competence, motivating them to vote against the incumbent.

What could be the role of an independent fiscal agency? The analysis above suggests that an institution able to clearly inform voters of whether the incumbent behaved consistently with an ex ante intent to produce q1* would solve the kind of asymmetric information problem described previously, and as a result, would contribute to eliminating the deficit bias that such asymmetry entails. For example, providing quantitative and qualitative analyses of fiscal policy (ex ante and in real time) could in principle help voters see through policymakers’ incentives and fiscal accounts so as to better grasp the adequacy of the observed policy stance. By definition, such an institution should be politically independent and have a broad remit that could include elements of judgment on the nature of shocks to the budget. If voters are better equipped to distinguish between bad luck and bad intentions, they can make better decisions, thereby eliminating opportunistic motives and directly rewarding competent governments with greater reelection chances.7

In sum, a well-functioning FC, by becoming the main source of information on the underlying quality of fiscal policy, would allow democratic accountability to play its role. Of course, placing such a considerable amount of trust in an FC would require strict guarantees of independence from partisan influences, a clear definition of the policy objectives under the council’s scrutiny (for example, fiscal sustainability), and a modus operandi genuinely “owned” by voters.

Fiscal Councils, the Public Debate, and Fiscal Policy

This section gathers empirical evidence on the effectiveness of FCs with a deliberate focus on the watchdog function, that is, the ability of the FC to signal fiscal misconduct. The focus is on two empirical questions: Does the watchdog bark when it should? And is the barking sufficiently impressive to deter misconduct and improve policy outcomes? To measure the impact of fiscal watchdogs in the public debate, this analysis uses the data set on the media impact of FCs compiled for the case studies in Curristine, Harris, and Seiwald (2013). This chapter thus complements the existing empirical literature, which relies mainly on thorough country case studies (for example, Calmfors and Wren-Lewis 2011; Coene and Langenus 2011) and largely illustrative correlations between outcomes and specific features of FCs (Debrun and others 2013).

The Approach

The approach is twofold. First, we build on the theoretical sketch developed above to explore the influence of FCs on the public debate through direct measures of their media impact. Second, two short case studies analyze the media impact in greater details. The use of higher-frequency data (monthly instead of annual) allows for a more refined assessment of the noise-to-signal ratio of media reports.

The media impact variable is based on the number of times the official name of the FC appears in a country’s national press (either in English or in national languages). Simple panel regressions are used to detect a relationship between the intensity of media reports referring to the FC and two real-time fiscal policy indicators expected to be of interest to the FC. The first policy indicator is the planned change in the cyclically adjusted budget balance (CAB) at the beginning of the year, which captures the degree of fiscal activism planned in the budget (stimulus or consolidation). The second is the first estimate of the deviation in the CAB with respect to plans. This variable is interpreted as a symptom (see Beetsma, Giuliodori, and Wierts 2009) of slippages (or overperformance) during the budget year, although its signaling power is affected by possibly large revisions in estimated output gaps.

Clearly, the results reported in the remainder of this section only constitute a first exploration, a potential enticement for further research that will be needed when the large number of new FCs will have gathered sufficiently long experience. One source of concern is the reliability of the underlying media data, which could create unwanted noise in the statistical analysis. For instance, it is difficult to assess whether the comprehensiveness of press coverage is comparable across countries. Another potential issue is that the straightforward measure of media impact that is used8 could capture events only loosely related to the national budgetary debate, for example, political bickering about the appointment of a new FC member.

Data, Notation, and Testable Hypotheses

Data Sources and Notation

Real-time fiscal and economic data are collected from the European Commission’s assessments of Stability and Convergence Programs (SCPs)—typically discussed in early spring on the basis of data transmitted by national authorities. To save on notation, country indices are omitted and time superscripts are used to denote the vintage of the commission’s assessment of SCPs. As a result, Xtt is the “realtime” forecast of Xt at the beginning of year t, and Xtt+1 is the first estimate of the realization of Xt. The Δ operator symbolizes the first-difference of a variable over time. Hence, ΔXtt=XttXt1t measures the planned “real-time” variation in X between t and t – 1, while ΔXtt+1=Xtt+1Xt1t+1 is the “first estimate” of the actual change. Finally, the “forecast error” of X is defined as ERRXtt+1=Xtt+1Xtt.

Because of severe data limitations in the media data set, the sample only includes seven continental European countries—Austria, Belgium, Germany, Denmark, Netherlands, Sweden, and Slovenia—over the period 2003–10—years with comprehensive assessments of SCPs by the European Commission. To maximize the time dimension, the main fiscal policy indicator is the CAB instead of the more conventional cyclically adjusted primary balance, which was not consistently reported in SCPs by all countries in the early years of the sample. The set of control variables is necessarily parsimonious given the small size of the sample. It includes the gross public debt (DEBT) and the output gap (OG). The key indicator of media impact, NEWS, is the standardized number (z-score) of written press articles quoting the national FC (in English and national languages). It reflects the intensity of the news coverage of the FC’s activity in each country and is thus taken as a proxy of its participation in the public debate on fiscal matters.

Testable Hypotheses

Theory points to two testable hypotheses. First, a necessary condition for the effectiveness of an FC is that it publicly reacts to fiscal developments, either because they signal a policy shift with a notable bearing on the realization of policy objectives or because undesirable deviations from planned outcomes call for an analysis of the causes—bad luck or bad budget execution. To the extent that these reactions contribute to the public debate, the FC will enhance democratic controls. In that case, a change in the NEWS variable follows variations in fiscal indicators. Second, an effective FC would be expected to make a difference in the conduct of fiscal policy, which would imply that changes in fiscal indicators would tend to follow peaks in FCs’ media activity. The analysis focuses on the planned change in the CAB as a proxy for deliberate policy action, and on the forecast error, which could reflect unexpected shocks affecting the budget or policy slippages.


We first document systematic differences in key fiscal dimensions between the seven countries that have FCs (hereafter FC7) considered in the analysis and the others, that is, countries that do not have FCs and those that might have introduced an FC more recently. The FC7 exhibit, on average, stronger fiscal balances than the rest of the EU (Figure 16.1). By contrast, while public debt was higher—by about 10 percentage points of GDP—in the FC7 before the crisis, it is slightly lower, on average, by the end of 2014. Because the stronger fiscal positions may in part reflect greater concerns about debt stabilization, market perceptions of sovereign risk for the two groups are also compared (Figure 16.2). Again, the FC7 enjoy much lower spreads, on average, than other EU countries, even after excluding the euro area periphery.

Figure 16.1.Fiscal Perfomance in EU Countries with Fiscal Councils (FC7) and other EU Member States

(Percent of GDP)

Sources: International Monetary Fund, World Economic Outlook database, and authors’ calculations.

Note: EU = European Union; FC7 = European countries with fiscal councils: Austria, Belgium, Germany, Denmark, Netherlands, Sweden, and Slovenia.

Figure 16.2.CDS Spreads (Five-Year) in EU Countries with Fiscal Councils (FC7) and the Other EU Member States

(Basis points)

Sources: Markit; and authors’ calculations.

Note: CDS = credit default swap; EU = European Union; FC7 = European countries with fiscal councils: Austria, Belgium, Germany, Denmark, Netherlands, Sweden, and Slovenia.

Do the Watchdogs Bark When Needed?

Using the seven-country sample over the period 2003–10, the following model is estimated: NEWSt=c+dt+αFISCtt+i+ɛt, where FISC is a fiscal indicator, dt is time dummies, and c indicates the use of country fixed effects. The fiscal indicator is either the planned variation in the CAB at the beginning of year t (so i = 0) or the first estimate (in early t + 1) of the forecast error (i = 1), while NEWS captures the intensity of citations in the press during the entire year t.

The regressions reported in Table 16.1 detect a statistically significant correlation between the two fiscal indicators and the media impact of FCs. On average, the fiscal watchdogs seem to bark louder when large fiscal policy changes are planned, and they seem to be heard in the press. The FCs in the sample also tend to be more present in the media during years when forecast errors end up being greater, suggesting that they publicly reveal policy slippages or express concerns when unexpected shocks send budget plans off track. Interestingly, the negative signs obtained for the first differenced CAB and the raw forecast error suggest that deteriorations in the CAB lead to increased media attention to FCs’ messages. This finding is consistent with FCs’ primary role of promoting fiscal responsibility.

Table 16.1.Media Impact of Fiscal Councils and Magnitude of Fiscal Developments(Dependent variable: NEWS)
Fiscal Councils−0.392**«0.721***−0.217*0.453**
Adjusted R20.
Source: European Commission’s Assessments of Stability Programs; and authors’ calculations.Note: All regressions include country fixed effects; robust t-statistics are in parentheses. CAB is the cyclically-adjusted balance, and ERRCAB represents the forecast errors of the cyclically-adjusted balance as defined in the text.*p < .1; **p < .05; ***p < .01.
Source: European Commission’s Assessments of Stability Programs; and authors’ calculations.Note: All regressions include country fixed effects; robust t-statistics are in parentheses. CAB is the cyclically-adjusted balance, and ERRCAB represents the forecast errors of the cyclically-adjusted balance as defined in the text.*p < .1; **p < .05; ***p < .01.

Although these results provide an indication that FCs may be effective in sending the right messages at the right time, the limited size of the sample and the potentially weak signaling power of media activity call for caution. In particular, fiscal developments explain only a small fraction of the quantity of news reports referring to the FCs. Also, the introduction of time dummies severely reduces the precision in the estimated impact of fiscal indicators on FCs’ presence in the media, suggesting that the results in Table 16.1 may capture developments—such as the global financial crisis—that jointly affect the budget and FCs’ reported activity.

Do FCs Influence the Conduct of Fiscal Policy?

Following the earlier discussion, the underlying regression is as follows: |ΔCABtt|=c+dt+δNEWSt1+β1DEBTt1t+β2OGt1t+ɛt. In contrast to the first set of regressions, several control variables known to be related to fiscal developments are introduced, namely the cyclical position of the economy (lagged output gap) and the (lagged) public debt level. With regard to the left-hand-side variable, meaningful correlations could be detected only between the absolute value of the planned change in the CAB for year t and FCs’ media presence in year t – 1.

As Table 16.2 suggests, stronger presence of the FC in the media in any given year is correlated with greater planned “fiscal activism” for the following year, regardless of whether the plans envisage a more ambitious fiscal consolidation or a greater stimulus. The results are fairly consistent across alternative regressions (fixed effects or pooled) and with the inclusion of the only two statistically significant time dummies (2007 and 2008). The control variables display the expected signs: a reduction in the output gap (less positive or more negative) encourages fiscal activism, whereas high public debt is less conducive to activism.

It may be tempting to conclude that more intense FC involvement in the public debate pushes governments to do the right thing; however, the paucity and quality of the data and the fragility of the results call for caution. In addition, simultaneity problems loom large, given that the plans for year t are prepared and discussed intensively in the second part of the preceding year. The results may thus reflect the fact that FCs are simply taking an active part in the debate surrounding budget preparation, particularly when significant policy shifts are discussed. While this would be welcome and supportive of the evidence in Table 16.1, it might not necessarily reflect the council’s actual influence on policy choices. Taking two lags to the NEWS variable did not help, suggesting that the distance between the council’s message and actual decision making would then be too large.

Table 16.2.Do Fiscal Councils Shape Policy Outcomes?(Dependent variable: Absolute value of the one-year-ahead ΔCAB)
Adjusted R20.610.570.560.550.56
Source: Factiva; European Commission’s Assessments of Stability Programs and Spring Forecasts; and authors’ calculations.Note: Robust t-statistics are in parentheses. NEWS is the standardized number (z-score) of written press articles quoting the national fiscal council. OG and DEBT are the first estimates of the output gap and general government debt, respectively. LSDV = least squares dummy variable; OLS = ordinary least square.*p < .1; **p < .05; ***p < .01.
Source: Factiva; European Commission’s Assessments of Stability Programs and Spring Forecasts; and authors’ calculations.Note: Robust t-statistics are in parentheses. NEWS is the standardized number (z-score) of written press articles quoting the national fiscal council. OG and DEBT are the first estimates of the output gap and general government debt, respectively. LSDV = least squares dummy variable; OLS = ordinary least square.*p < .1; **p < .05; ***p < .01.

Going Forensic: Fiscal Councils and the Budget Process

Although promising, the preliminary evidence on a link between FCs’ media presence and fiscal indicators—which only conveys a clear policy signal at annual frequency—calls for a much more detailed analysis that goes beyond the scope of the present chapter. Insights can nevertheless be gained from a more “forensic” analysis involving higher-frequency data. In contrast to more comprehensive case studies of FCs (for example, the analysis of Sweden and the United Kingdom by Calmfors and Wren-Lewis [2011]), this analysis focuses on the link between certain fiscal episodes and FCs’ media visibility.

Two case studies of long-lasting FCs are interesting in this regard: the Netherlands’ Bureau for Economic Analysis, better known by its Dutch acronym CPB, and the Public Sector Borrowing Requirement Section of Belgium’s High Council of Finance. The selection of two well-established institutions offers some guarantees that media quotes capture the actual impact of their routine work on the public debate. Analyzing new or failed institutions could distort the measure of media presence because the press could report on debates about the council itself—for example, public discussions about individual appointments or reports about personal or political conflicts preceding the FC’s demise—rather than fiscal policy. Belgium and the Netherlands also provide the advantage of operating in contexts in which rules and numerical norms have typically played an important role. This facilitates the identification of specific “stress” episodes—for example, clear threat of breaking a rule or an official numerical target—during which the FC would be expected to communicate.

Case Study 1: The Netherlands

The Netherlands Bureau for Economic Policy Analysis—known by its familiar acronym of CPB, for Central Planning Bureau—is a veteran among FCs, beginning operations in 1945. It has been preparing economic forecasts since the 1950s, publishing reports on the state of Dutch public finances since the 1960s, undertaking election-commitment costing since the 1980s, and preparing long-term scenario analyses since the 1990s. The CPB also has a number of other roles, including cost-benefit analysis, and a fairly broad research agenda.

The CPB is formally a branch of the civil service within the Ministry of Economic Affairs and it is funded from the budget. Despite lacking a separate status for staff and being financially dependent on the budget, several elements contribute to the independence of the CPB, including a supervisory committee whose members are appointed by the cabinet for five years, and five-yearly independent evaluations (usually by academics, many of whom are not Dutch nationals).

Like most FCs, the CPB cannot bite: it has no explicit mandate or instrument that can directly influence fiscal policy. Also, it does not provide strictly normative assessments, not even against the government’s own commitments and objectives. However, the CPB quantitatively evaluates the effect of government policies on public finances. Clearly, a positive statement saying that the continuation of a certain policy is bound to threaten fiscal sustainability is not materially different from a normative judgment that such policy should be changed.

Another positive way for the CPB to enter into normative territory at politically critical times relates to its role of costing electoral platforms before, and coalition agreements after, elections. That function has arguably improved the quality of public information and influenced the debate around elections, encouraging parties to tighten up their commitments in advance (so that costing can be undertaken) and to avoid making commitments to unaffordable policies (Bos and Teulings 2011).

The CPB is an insider working within the budget process. It provides the macroeconomic forecasts used for the budget. Those forecasts are produced behind closed doors and discussed with the government (initially through the Ministry of Finance, and later through the cabinet) before being released. While this means that the budget relies on independent forecasts, private discussions with the cabinet provide an opportunity for cabinet members to put pressure on the CPB (Bos and Teulings 2011). For the same reason, the CPB provides little public information about the evolution of the fiscal outlook during the year, allowing the government to change course, if necessary, without raising public awareness. Following meetings with the cabinet, the CPB publishes its forecasts four times a year, with more extensive updates in March and September than in June and December.

Overall, this set of attributes, combined with a long history of well-regarded analysis and forecasts, has given the CPB a considerable degree of public credibility. Although its specific contribution is difficult to identify, the Netherlands’ fiscal performance since the mid-1990s has been relatively sound. In particular, there has been no apparent bias in macroeconomic forecasts, and ex post compliance with ex ante targets has been satisfactory.

Let us look at each of these three aspects in greater detail. Using data from the SCPs, we calculate the real GDP growth forecasting errors (at one-, two-, and three-year horizons) across countries over the period 2000–14. In the Netherlands, the average errors for the budget year have been, if anything, pessimistic; and while they appear slightly optimistic over the out-years, they remain well within one standard deviation of GDP growth in the sample. The same is true for the other countries whose FCs have within their remit providing independent macroeconomic forecasts for the budget (Austria and Belgium). This finding is in line with earlier results by Jonung and Larch (2006) showing that independent forecasting helps prevent an optimistic bias in GDP forecasts (Figure 16.3).

The discussion now turns to compliance with ex ante targets set over the political cycle. The Dutch budget system is governed by coalition agreements reached at the beginning of each parliamentary term. These agreements specify, among other things, a deficit ceiling for the remainder of the term (bold black lines in Figure 16.4). Figure 16.4 also displays yearly forecasts for the budget balance (green line) and the final budget outcome (red line).

Figure 16.3.Real GDP Forecast Errors (Actual Minus Forecast), 2000–14

(Percentage points)

Sources: European Commission’s Assessments of the Stability and Convergence Programs; and authors’ calculations.

Note: The variable t indicates year.

Figure 16.4.The Netherlands: Fiscal Targets, Forecasts, and Outcomes

(Percent of GDP)

Sources: Netherlands Bureau for Economic Policy Analysis; and authors’ calculations.

This comparison allows identification of episodes of interest characterized as follows: (1) either the budget forecast or fiscal outturn is exceeding the deficit ceiling, or (2) the fiscal outturn is significantly worse than the budget forecast. In either of these cases, the FC would be expected to raise awareness of these outcomes and make recommendations or suggestions to rectify the situation. Using the same media information as above (now at monthly frequency), we examine whether this did, in fact, occur.

The first episode was in 2002, when the fiscal outturn came out significantly worse than the budget forecast, and was in fact approaching the deficit ceiling set in the coalition agreement. The second episode occurred in 2003, when the outturn was again significantly worse than forecast, but this time actually breached the deficit ceiling. The third and final episode was in 2005, when the budget forecast exceeded the ceiling, even though the outturn ended up comfortably inside the ceiling. The events related to the global financial crisis are excluded because they represent a more significant, exogenous shock, with less control available to policymakers.

These episodes are now mapped in Figure 16.5, compiling the CPB’s media presence calculated as the number of newspaper articles containing a reference to the CPB. We have adjusted and corrected that series for trend, looking at the number of articles relative to a four-year rolling average. The vertical bars in red refer to the months in which the CPB releases its public forecasts and fiscal assessments. Some of the spikes observed in the series refer to election periods, where the CPB featured heavily in the news.

Figure 16.5.Media Presence of the CPB

(Number of press articles: relative deviations from four-year rolling average)

Sources: Factiva; and authors’ calculations.

Note: CPB = Netherlands Bureau for Economic Policy Analysis (Centraal Planbureau).

The three episodes identified above coincide with a marked increase in media reports. It is useful to note that the major spikes, particularly in 2003 and 2005, relate to the timing of the CPB’s release of its public assessments.

An examination of the contents of CPB reports at those times confirms this interpretation. In 2002, the slippage became apparent toward the end of the year, and the CPB noted in its December report that because of the worsening economy, the small surplus achieved in 2001 would turn to a deficit of 0.8 percent of GDP in 2002. In 2003, this continued deterioration was noted in the March report, where the projections in the Central Economic Plan of that year were revised downward and the CPB noted that “additional deficit-reducing measures are necessary to comply with the rules from the SGP [Stability and Growth Pact].” In the December 2003 report, the deficit forecasts were further revised upward, with the CPB noting that “even though the 3.35% deficit will be just above the 3% SGP ceiling, this does not mean that Brussels will determine this deficit to be ‘excessive.’ Netherlands will probably be able to call on exceptional economic conditions.” In 2005, the forecast for the year exceeded the agreed-upon deficit ceiling, which prompted the CPB to warn in its December 2004 report that, “Regardless of the austerity measures of the government, the [2005] deficit stays dangerously close to the 3% SGP deficit ceiling.”

Beyond individual episodes, the CPB’s ability to communicate to the public when it has something important to say and when it matters most during the budget process is important every year. Indeed, an FC constantly out in the public with a running commentary, disconnected from important parliamentary budget preparation deadlines, would raise doubts about its ability to add something important to the debate. We can assess this “noise-to-signal” ratio by plotting the average monthly seasonal patterns of media presence of the council (Figure 16.6). The spikes suggests that the CPB times its media interventions in a fairly systematic way, taking full advantage of the publication of its most extensive reports in March and September to influence the public debate.

Figure 16.6.Average Seasonal Pattern of the CPB Media Presence

(Percent of annual hits)

Sources: Factiva; and authors’ calculations.

Note: The red bars correspond to the publication dates of the main reports. CPB = Netherlands Bureau for Economic Policy Analysis (Centraal Planbureau).

The message emanating from this analysis is consistent with the broader correlations detected earlier in the panel of seven countries. Although it is difficult to identify a robust causal link between the activities of the CPB and improved fiscal performance in the Netherlands, the CPB’s media visibility supports the idea that it actively contributes to the quality of the public debate and reduces information asymmetries between decision makers and the public. There is indeed no significant forecast bias, the messages of the council appear relevant and those messages are communicated in an effective and timely manner.

Case Study 2: Belgium

Two bodies provide independent fiscal inputs into the budgetary process in Belgium. Their responsibilities are split between the normative recommendations and assessments of fiscal policy from the High Council of Finance (HCF, and more specifically the so-called Public Sector Borrowing Requirement Section), and the positive, forecast-input role of the Federal Planning Bureau (under the umbrella of the National Accounts Institute)—see Lebrun (2009). The case of Belgium is of particular interest because it is one of only a few countries where an FC makes normative recommendations. Over the past two decades, the HCF has gone through periods of being highly influential, followed by periods where that influence has waned (Coene and Langenus 2011).

The HCF dates back to the 1930s, when it was formed to advise the Ministry of Finance. It assumed its current role in 1989 when it started issuing recommendations on public sector borrowing requirements. The main objective of Belgian public finances at the time was to reduce the high budget deficit and public debt ratios, which stood at 7 percent and 125 percent of GDP, respectively, in 1988. Much of the HCF’s recommendations focused on monitoring and coordinating the fiscal effort across the different levels of government. Belgium’s commitment to meet the Maastricht criteria provided an opportunity to expand the HCF’s normative role with recommendations to comply with these criteria as well as with medium-term objectives.

Although the HCF does not have any direct instruments with which to shape budgetary choices, its recommendations, compiled in two reports, have often carried significant weight. The first report is published at the beginning of the budget process (June/July), when the HCF provides ex ante recommendations for fiscal targets. The second report provides an ex post assessment of the extent to which the government followed the recommendations and implemented the SCP of the previous year.

Unlike the Dutch CPB, the institutional setup governing the HCF is more that of an outsider looking into the budget. Its recommendations are developed in a fairly independent manner for the government to take on as it sees fit, rather than being worked out and negotiated behind closed doors. The HCF is staffed by experts from ministries, the National Bank, the Federal Planning Bureau, and academia. The HCF’s members are appointed by the Minister of Finance for a five-year renewable mandate running across electoral cycles, which helps preserve independence. The chairperson is often an academic, although ties with political parties often exist. In addition, the political principal has at times let the HCF’s influence wane. Over the mid-2000s, for instance, after a number of unfavorable assessments, the Minister of Finance let the chairman’s mandate lapse without appointing a successor. Although this hiatus did not obviously reflect political interference—the composition of the HCF is subject to many constraints, which can cause delays in appointments—it prevented the body from providing fiscal recommendations, resulting in a significant loss of influence (Coene and Langenus 2011).

Overall, the performance of the HCF can be broken down into two periods: pre– and post–euro adoption (Coene and Langenus 2011). In the pre–euro adoption period, the Maastricht convergence criteria led to a good alignment of policymakers’ objectives with the HCF’s recommendations, which were largely followed. During that period, the structural primary surplus increased by 5 percent of GDP, public debt started falling, and the budget moved toward balance. After euro adoption, the weight of convergence criteria in fiscal policy formulation was lost, and the appeal of the HCF’s recommendations eroded. The primary balance deteriorated sharply because most of the savings on the interest bill were used to fund tax cuts and expenditure growth.

Because of data constraints, only the post–euro adoption period can be subject to an analysis similar to the one performed for the CPB. In Figure 16.7, the HCF’s budget year recommendations for the fiscal balance are displayed in black, the budget forecast in green, and the outturn in red. A number of interesting episodes stand out. The first is the two years that follow adoption of the euro in 1999, when outcomes were broadly in line with the HCF’s recommendations. The second episode is in 2004, when increasing divergences with the HCF’s recommendations led to critical assessments of government policies. Finally, in 2005 and 2006, the HCF could not produce reports or recommendations because its chair was left vacant.

Figure 16.7.High Council of Finance Budget Balance Recommendations, Forecasts, and Outturns

(Percent of GDP)

Sources: High Council of Finance; and authors’ calculations.

These episodes are now mapped into the measure of the HCF’s media presence (Figure 16.8). In the initial period leading up to and during adoption of the euro, coverage of the HCF in the Belgian press was strong and sustained. Coverage tailed off as the 2000s progressed and the HCF clearly began losing its influence. In 2004, the HCF turned openly critical of the government’s policies,9 leading to a surge in media reports. Finally, during the period when the HCF had no formal head and stopped reporting, its media footprint shrank substantially, despite emerging discrepancies between plans and outcomes in 2006.

Figure 16.8.Media Presence of the High Council of Finance

(Number of press articles)

Sources: Factiva; and authors’ calculations.

Again, a more detailed analysis of the HCF’s reports and press citations supports our reading of the data. In the early 2000s, when public finances were on track, press articles were relatively benign and the HCF reports noted with a fairly neutral tone minor deviations with respect to recommendations. For instance, the October 1999 report observed that “The deficit of the central government for 2000 is projected at 1 percent of GDP, which is 0.2 percent higher than recommended by the HCF.” The tone of the HCF’s reports changed drastically in 2004. In July of that year, it launched a stern warning in these terms: “The High Council considers these fiscal scenarios unrealistic in the medium-term,” and “The High Council wants to point out that the deteriorating fiscal position is [not only] due to cyclical factors but also—and mainly—due to discretionary measures such as lowering taxes and a major growth in government spending.” Of course, there is no report to quote for the last episode.

The monthly pattern of media reporting on the HCF is less clear-cut and incisive than in the Dutch case, pointing to a weaker ability to shape the public debate on fiscal policy (Figure 16.9). In particular, media presence is spread out more evenly over the year and is less concentrated around the times when the HCF published its key fiscal reports (in March and July). Such a pattern could also reflect a greater persistence of the HCF’s message, for instance, because the media use the recommendations as a constant benchmark when discussing budgetary matters. Supporting that conjecture is the fact that despite less-pronounced monthly spikes than in the CPB case, a concentration of press articles cite the HCF in the latter half of the year, the period during which the budget is being negotiated and passed by parliament.

Figure 16.9.Average Seasonal Pattern of the High Council of Finance’s Media Presence

(Percent of annual hits)

Sources: Factiva; and authors’ calculations.

Note: The red bars correspond to the publication dates of the main reports.

Concluding Remarks

The chapter discusses how independent fiscal institutions, even though they have no direct policy levers, can influence fiscal performance. From a theoretical perspective, the effectiveness of FCs depends on their capacity to deal with the root cause of fiscal policy biases, and, in particular, informational asymmetries between voters—the only legitimate principal in the policy game—and politicians. FCs do this by performing tasks that improve the quality of the public debate on fiscal policy (by providing analysis, recommendations, assessments, forecasts). This output allows the general public to properly process and interpret the often complex, opaque, and ideologically tainted signals received on the performance of policymakers. This line of argument means that FCs should behave and be perceived as serious watchdogs that bark when needed and are listened to.

The chapter offers an empirical assessment of this watchdog role. In line with the conclusions of the theoretical model, the analysis tries to quantify FCs’ impact through their capacity to influence the public debate. We propose to proxy such influence with direct measures of media presence.

While this exploration of the statistical link between media presence and fiscal policy is promising, drawing robust conclusions would require a much broader sample. With only a handful of FCs with a sufficiently long lifetime, such exploration can only be considered a very first pass. Keeping these caveats in mind, it seems that, on average, fiscal watchdogs bark when expected. This is a necessary condition to ensure that FCs clarify signals about the appropriateness of fiscal policy, helping the general public reward good policies and sanction bad ones. If an FC can meaningfully improve policymakers’ incentives, it can credibly strengthen fiscal performance. It nevertheless proved more difficult to find evidence of systematic policy corrections after peaks in FC media presence. One possible reason for these mixed results is that the repercussions of FCs’ activity on the budget process might be more subtle and detectable only in specific inputs to the budget rather than in the aggregate fiscal stance itself. For instance, Debrun and Kinda (2014) show that the presence of an FC mandated to assess or produce budgetary forecasts leads to improvements in their quality (absence of bias and greater precision). Two case studies of well-established FCs lend support to the statistical findings.

An important avenue for future research would be to analyze FC design features that seem conducive to good capacity to improve the public’s information about the quality of fiscal policy. The somewhat contrasting experiences of Belgium and the Netherlands suggest that strict independence from politics and recognized expertise in economics and public finance are key for the barking to be heard by all and taken seriously by policymakers.


    AlesinaA. and G.Tabellini.1990. “A Positive Theory of Fiscal Deficits and Government Debt.Review of Economic Studies57 (3): 40314.

    BeetsmaR. and X.Debrun.2007. “The New Stability and Growth Pact: A First Assessment.European Economic Review51 (2): 45377.

    BeetsmaR. and X.Debrun.2016a. “Fiscal Councils: Rationale and Effectiveness.Working Paper 16/86International Monetary FundWashington, DC.

    BeetsmaR. and X.Debrun.2016b. “The Political Economy of Fiscal Transparency and Independent Fiscal Councils: Theory and Evidence.Unpublished report.

    BeetsmaR.M.Giuliodori and P.Wierts.2009. “Planning to Cheat: EU Fiscal Policy in Real Time.Economic Policy24: 753804.

    BosF. and C.Teulings.2011. “Evaluating Election Platforms: A Task for Fiscal Councils? Scope and Rules in Light of 25 Years of Dutch Practice.MPRA Paper 31536University Library of Munich.

    CalmforsL.2010. “The Role of Independent Fiscal Policy Institutions.Report to the Prime Minister’s OfficeFinland.

    CalmforsL. and S.Wren-Lewis.2011. “What Should Fiscal Councils Do?Economic Policy26 (68): 64995.

    CoeneL. and G.Langenus.2011. “Promoting Fiscal Discipline in a Federal Country: The Mixed Track Record of Belgium’s High Council of Finance.Paper presented at the Workshop “Fiscal Policy Councils—Why Do We Need Them and What Makes Them Effective?” organized by the Austrian Government Debt Committee ViennaJanuary31.

    CukiermanA. and A.Meltzer.1986. “A Theory of Ambiguity, Credibility, and Inflation under Discretion and Asymmetric Information.Econometrica54 (5): 1099128.

    CurristineT.J.Harris and J.Seiwald.2013. “Case Studies of Fiscal Councils—Functions and Impact.Policy PaperInternational Monetary FundWashington, DC.

    DebrunX.2011. “Democratic Accountability, Deficit Bias, and Independent Fiscal Agencies.Working Paper 11/173International Monetary FundWashington, DC.

    DebrunXD.Hauner and M.Kumar.2009. “Independent Fiscal Agencies.Journal of Economic Surveys23 (1): 4481.

    DebrunX.T.CurristineL.EyraudJ.Harris and J.Seiwald.2013. “The Functions and Impact of Fiscal Councils.Policy PaperInternational Monetary FundWashington, DC.

    DebrunX. and T.Kinda.2014. “Strengthening Post-Crisis Fiscal Credibility—Fiscal Councils on the Rise. A New Dataset.Working Paper 14/58International Monetary FundWashington, DC.

    DebrunX. and M.Kumar.2009. “The Discipline-Enhancing Role of Fiscal Institutions: Theory and Empirical Evidence.” In Policy Instruments for Sound Fiscal Policies; Fiscal Rules and Institutions edited by J.Ayuso-i-CasalsS.DerooseE.Flores and L.Moulin.London: Palgrave Macmillan.

    DebrunX.L.MoulinA.TurriniJ.Ayuso-i-Casals and M.Kumar.2008. “Tied to the Mast? The Role of National Fiscal Rules in the European Union.Economic Policy23 (54): 298362.

    Deutsche Bank. 2016. “Better Budgeting in Europe: What Can Fiscal Councils Contribute?EU Monitor (June7): 120.

    FabrizioS. and A.Mody.2006. “Can Budget Institutions Counteract Political Indiscipline?Economic Policy21: 689739.

    HagemannR.2010. “Improving Fiscal Performance through Fiscal Councils.Economics Department Working Paper 829Organisation for Economic Co-operation and DevelopmentParis.

    JensenH.1997. “Credibility of Optimal Monetary Delegation.American Economic Review87 (5): 91120.

    JonungL. and M.Larch.2006. “Improving Fiscal Policy in the EU: The Case for Independent Forecasts.Economic Policy21 (47): 491534.

    KopitsG. ed. 2013. Restoring Public Debt Sustainability: The Role of Independent Fiscal Institutions.Oxford, U.K.: Oxford University Press.

    KrogstrupS. and C.Wyplosz.2010. “A Common Pool Theory of Supranational Deficit Ceilings.European Economic Review54 (2): 26978.

    KumarM. S. and T.Ter-Minassian.2007. Promoting Fiscal Discipline.Washington, DC: International Monetary Fund.

    LebrunI.2009. “Fiscal Councils, Independent Forecasts, and the Budgetary Process: Lessons from the Belgian Case.” In Policy Instruments for Sound Fiscal Policies: Fiscal Rules and Institutions edited by J.Ayuso-i-CasalsS.DerooseE.Flores and L.Moulin320345. London: Palgrave Macmillan.

    McCallumB.1995. “Two Fallacies Concerning Central Bank Independence.American Economic Review85 (2): 20711.

    RogoffK. and A.Sibert.1988. “Elections and Macroeconomic Policy Cycles.Review of Economic Studies55 (1): 116.

    WyploszC.2005. “Fiscal Policy: Institutions versus Rules.National Institute Economic Review191 (1): 6478.

The authors are grateful to Petra Dacheva, Chifundo Moya, and Inna Remizova for outstanding research assistance, and to Bat-el Berger for help in the case studies. The views expressed in this chapter are those of the authors and do not necessarily represent those of the IMF or IMF policy.

Fragmented government coalitions (for example, Fabrizio and Mody 2006) and political instability (Debrun and Kumar 2009) emerge as more robust causes of excessive deficit. This may reflect common pool problems—that is, the failure to coordinate competing claims on finite budget resources—as well as the fear of not being reelected (implying myopia for politician); see, for example, Krogstrup and Wyplosz (2010).

The term deficit bias means that a utility-maximizing policymaker delivers a fiscal balance that is systematically weaker than one delivered by a representative agent who is directly in charge of fiscal policy.

This argument is analogous to McCallum’s (1995) second fallacy of central bank independence, stating that if governments have the discretion to set up an independent central bank with the right incentives, they also have the discretion to revert to a dependent central bank with inadequate incentives. Jensen (1997) formally demonstrates in the Barro-Gordon-Rogoff framework that delegation does not matter if the no-renegotiation assumption is lifted.

The problems for numerical fiscal rules have been documented and extensively discussed. However, in part because FCs can complicate attempts to circumvent rules—for example, by exposing accounting tricks and inadequate implementation of the rule—they are also exposed to attempts to reduce their influence. The abolition of the Hungarian Fiscal Council, created only two years earlier, is a vivid illustration of the inherent fragility of discipline-enhancing fiscal institutions. Beetsma and Debrun (2016b) formally model circumstances under which a country that has initially established an FC could eliminate it following elections.

See Chapter 17, in this volume, by Gaspar.

Cukierman and Meltzer (1986) and Rogoff and Sibert (1988) first introduced that type of conjecture in formal models of fiscal policy.

Beetsma and Debrun (2016a) demonstrate this in a simple Bayesian game with unobservable competence based on the same basic setup as in this chapter. They formally show that FCs contribute to improving average fiscal performance by (1) having competent governments elected more often and (2) encouraging incompetent governments to mimic competent ones.

All media data used in the case studies of Curristine, Harris, and Seiwald (2013) come from the Factiva database.

The large deficit in 2005 was almost entirely due to a one-off debt assumption, related to the restructuring of the national railway company, which was attributed after the event.

    Other Resources Citing This Publication