## wp17118

## Source details

**Canonical URL:** [wp17118](https://www.imf.org/-/media/files/publications/wp/2017/wp17118.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp17118.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp17118.pdf.json)

---

### Political economy framing and scope
- Euro area (EA) public-debt-to-GDP ratio rose from an average level of less than 60 percent of GDP in the early 1990s to more than 90 percent of GDP in 2015.
- Analysis covers 19 EA countries over 1999–2015 using AMECO ex post outturns and a real-time database built from vintages of Stability and Convergence Programs.
- Focus: document fiscal policy biases (excessive deficits, procyclicality, compositional distortions), assess compliance with SGP numerical rules, and evaluate political economy drivers at national and supranational levels.
- Scope limitations: does not assess optimal rule design or fiscal policy under an effective lower bound monetary policy.

### Evidence on procyclicality (planning versus implementation)
- Econometric specification: Δsb_{i,t} = α_i + β × Δgap_{i,t} + γ × debt_{i,t−1} + ρ_t + ε_{i,t}, with sb = structural-balance-to-potential-GDP ratio.
- Key estimation results (Fixed Effects Estimator, 19 countries, 1999–2015):
  - Plans (Δ(SB/GDP) t|t-1):
    - Initial Debt/GDP: 0.02
    - ΔOutput Gap: -0.12
    - Observations: 105; Number of Countries: 19; R2: 0.33
  - Actual, Real Time (Δ(SB/GDP) t|t+1):
    - Initial Debt/GDP: 0.05***
    - ΔOutput Gap: -0.27*
    - Observations: 102; Number of Countries: 19; R2: 0.65
  - Actual, Ex Post (Δ(SB/GDP) t|2016):
    - Initial Debt/GDP: 0.04***
    - ΔOutput Gap: -0.55**
    - Observations: 224; Number of Countries: 19; R2: 0.29
- Interpretation:
  - Planned fiscal impulses appear broadly acyclical (insignificant β in plans).
  - Real-time and ex post outcomes display procyclicality (negative and significant β).
  - Dynamic GMM and robustness checks (including political variables and exclusion of 2011–13) confirm procyclicality is not crisis-specific.

### Evidence on asymmetric response and deficit bias
- Asymmetry regression: Δsb_{i,t} = α_i + β_1 × Δgap_{i,t} × D(Δgap_{i,t} > 0) + β_2 × Δgap_{i,t} × D(Δgap_{i,t} < 0) + γ × debt_{i,t−1} + ρ_t + ε_{i,t}.
- Selected coefficients (Table 2; Fixed Effects Estimator):
  - Initial Debt/GDP:
    - Column (1) Plans: 0.01* [0.01]
    - Column (2) Actual, Real Time: 0.06*** [0.01]
    - Column (3) Actual, Ex Post: 0.04*** [0.01]
  - ΔOutput Gap*D(ΔOutput Gap>0):
    - Column (1): -0.45*** [0.07]
    - Column (2): -0.84*** [0.20]
    - Column (3): -0.31 [0.30]
  - ΔOutput Gap*D(ΔOutput Gap<0):
    - Column (1): 0.17 [0.26]
    - Column (2): -0.13 [0.18]
    - Column (3): -0.73*** [0.22]
  - Observations: Column (1) 105; Column (2) 102; Column (3) 224. Number of Countries: 19. R^2: 0.39, 0.67, 0.3 respectively.
- Findings:
  - Strong evidence of asymmetric response: lax policies in good times prevented buffer rebuilding.
  - Estimations corroborated by GMM and robust to inclusion of political-economy variables and exclusion of post-2009–10 consolidation periods.

### Fiscal composition and investment gap
- During 2010–13 adjustment, EA adjustments leaned more on revenue measures and cuts to public investment than current spending relative to the United States and other advanced economies.
- Rebased series (2007 = 100) indicate capital spending declined relative to wages, goods and services, interest, and other spending over 2007–2016 (index values between 70 and 130 in underlying figures).
- Investment gap (proxy): difference between observed public investment and investment that would stabilize the stock of public capital to GDP.
  - Capital-stabilizing net investment ratio = y/(1+y) × k, with y = real potential growth rate, k = initial stock of public capital/GDP.
  - Sample: 15 EA countries with capital stocks from IMF public debt database.
  - As of 2015, investment gap ranged between 0.35 percent and 3.7 percent of GDP across countries.
- Implication: steady depletion of public capital could slow potential and harm fiscal sustainability.

### Compliance with fiscal rules: incidence, drivers, and decomposition
- Simplified compliance metrics: 3 percent nominal deficit ceiling; 60 percent debt ceiling; MTO in structural terms; benchmark fiscal effort of at least 0.5 percent when below MTOs or under EDP.
- Compliance outcomes 1999–2015:
  - MTO violated in 80 percent of observations.
  - In more than three-quarters of total country-year observations the deficit exceeded the 3 percent ceiling.
  - In 16 of the 19 EA countries, the average deficit over 1999–2015 was positive (in the range of 1.7 to 7.9 percent of GDP, depending on the country).
  - Share of countries with debt ratio > 60 percent rose from 35 percent in 1999 to 75 percent in 2015.
- Decomposition of rule slippages (deficit example):
  - 푫푬푭풕|ퟐퟎퟏퟔ − ퟑ = (푫푬푭풕|풕−풊 − ퟑ) + (푫푬푭풕|ퟐퟎퟏퟔ − 푫푬푭풕|풕−풊)
  - Labels: Ex post compliance = left term; Ex ante compliance = middle term; Execution = right term.
- Empirical finding on slippages (medians, 19 countries, 1998–2013):
  - For both the 3 percent deficit and 60 percent debt rules, poor execution is the main culprit: non-compliers planned sizable improvements but execution slippages offset plans.
  - For the debt rule, non-compliers planned to reduce debt by about 10 percent of GDP (from about 80 to 70 percent of GDP on average) but the debt ratio remained broadly unchanged ex post.
- Possible causes of execution-stage deviations:
  - Unexpected nominal growth shortfall (e.g., negative inflation surprise).
  - Intentional biases (deliberate underestimation of deficits).
  - Valuation changes and statistical adjustments in fiscal aggregates.
- Policy practice: since 2005 reforms, assessments often estimate and subtract economic factors to isolate the policy component (“conditional compliance”).

### Country size, fiscal slippage, and EDP treatment
- Regression on execution-stage slippage:
  - Δob_{i,t|2016} − Δob_{i,t|t−1} = α_i + β×Size_{i,t} + γ×Controls_{i,t} + ρ_t + ε_{i,t}, with Size = country nominal GDP share of EA GDP.
- Main empirical pattern:
  - Coefficient on Size is negative and significant in many specifications: larger countries tend to underperform plans by a larger margin.
  - Example coefficients reported across specifications: -0.32**, -0.30*, -0.29*, -0.23, -0.43***, -0.30**, -0.32***, -0.17, -1.64**, -0.63*, -0.30*, -0.35***, -0.32**, -1.99* (standard errors shown in source).
- EDP recommendations:
  - Regression Effort_i = α + β×Size_i + γ×Controls_i + ε_i on 39 EDP episodes (2003–15) shows β not significant in most specifications.
  - Table 4 extracts: for 19 EA countries coefficients on Size include -0.022*, -0.021*, -0.01, 0.00, -0.01, 0.00; for 28 EU countries coefficients include -0.02, -0.02, -0.01, -0.01, -0.01, 0.00.
- Summary: evidence on biased application of rules by country size is mixed—larger countries deviate more from plans, but EDP recommended adjustment effort does not systematically differ by size once controls are included.

### Institutional reforms, limitations, and policy options
- Past SGP reform elements:
  - Reverse qualified majority voting (2011) to strengthen EC proposals unless a qualified majority blocks them.
  - Creation of the European Fiscal Board (operational October 2016) for independent evaluation.
  - Increased role for independent fiscal institutions (Fiscal Compact) and greater specificity for flexibility clauses (e.g., 2015 investment clause).
- Limitations of reforms:
  - Increased complexity and reduced transparency; fiscal councils not immune to political interference.
  - Reverse qualified majority may not fully insulate decisions from political considerations.
- Principles for future reforms:
  - Make sanctions more politically acceptable:
    - Consider emphasis on preventive-arm sanctions and nonpecuniary sanctions under corrective arm; adopt gradual and proportionate penalties for repeated/larger deviations.
    - Credible enforcement may require further political integration; sovereign sanctions problematic under international law.
  - Create tangible benefits for compliers:
    - Link structural funds or central stabilization capacity access to compliance; design incentives so national policymakers reap short-run benefits from compliance.
  - Simplify the framework:
    - Reduce complexity, consolidate preventive and corrective arms, eliminate redundant/ill-designed rules to enhance legitimacy and reduce scope for interpretation.
- Long-term models and trade-offs:
  - Decentralized model with stronger market discipline: rely on markets, orderly default mechanisms, enhanced fiscal transparency, credible joint backstops.
  - Centralized model with deeper fiscal and political integration: stronger central oversight, potential central fiscal capacity, greater democratic legitimacy through supranational institutions.
  - Current EA combines elements of both models without fully implementing either; enduring solutions likely require broader political choices and integration.

### Conclusion: synthesis and policy recommendations
- Key empirical conclusions:
  - Persistent fiscal policy biases in the EA include procyclicality, excessive deficits, and compositional distortions.
  - Main driver of non-compliance with SGP numerical rules has often been execution-stage slippages rather than lack of ambition in plans.
  - Investment gaps as of 2015 ranged between 0.35 percent and 3.7 percent of GDP for sampled countries.
- Policy implications:
  - Reforms must combine sound rule design with incentives that align national political incentives with supranational objectives.
  - Short- to medium-term options: gradual and proportionate sanctions, tangible rewards for compliers, simplification of the fiscal framework, strengthened independent monitoring.
  - Long-term options: clearer choice between stronger market discipline and deeper fiscal/political integration, accompanied by complementary banking and capital market union measures.
- Econometric and data notes:
  - Dependent variables are overall (rather than primary) structural balances where data availability dictates.
  - Robustness: results hold across panel OLS and dynamic GMM, and remain after excluding 2011–13 crisis years in sensitivity checks.

*IMF working paper wp17118, 2016.*

### References ___________________________________________________________________________________________ 39

### wp17118 - References ___________________________________________________________________________________________ 39

### Figures
- 1. National Political Economy _________________________________________________________________________ 4
- 2. Supranational Political Economy ___________________________________________________________________ 8
- 3. Relationship between Changes in Structural Balances and Output Gaps __________________________ 11
- 4. Reliance on Revenue and Expenditure Measures in Fiscal Adjustment ____________________________ 16
- 5. General Government Expenditure Composition ___________________________________________________ 16
- 6. Share of Non-Compliers in the Euro Area _________________________________________________________ 18
- 7. Decomposition of Deficit Rule Slippages __________________________________________________________ 19
- 8. Decomposition of Debt Rule Slippages ____________________________________________________________ 20
- 9. Relationship between Country Size and EDP Length ______________________________________________ 23
- 10. Past SGP Reforms and Policy Biases ______________________________________________________________ 25

### Tables
- 1. Fiscal Policy Procyclicality _________________________________________________________________________ 12
- 2. Asymmetric Policy Response and Deficit Bias______________________________________________________ 15
- 3. Country Size and Fiscal Plan Execution ____________________________________________________________ 22
- 4. Country Size and EDP Recommendations _________________________________________________________ 24

### Annex Tables
- 1. Baseline Specification: Fiscal Policy Procyclicality __________________________________________________ 33
- 2. Baseline Specification: Asymmetric Policy Response and Deficit Bias ______________________________ 34
- 3. Including Political Economy Variables: Fiscal Policy Procyclicality __________________________________ 35
- 4. Including Political Economy Variables Asymmetric Policy Response and Deficit Bias ______________ 36
- 5. Excluding Periods of Fiscal Consolidation during the Global Financial Crisis: Fiscal Policy 
  Procyclicality ___________________________________________________________________________________ 37
- 6. Excluding Periods of Fiscal Consolidation during the Global Financial Crisis: Asymmetric Policy 
  Response and and Deficit Bias _________________________________________________________________ 38

*Source: wp17118 - References ___________________________________________________________________________________________ 39*

### INTRODUCTION

### INTRODUCTION

### Overview
- Euro area (EA) public-debt-to-GDP ratio rose from an average level of less than 60 percent of GDP in the early 1990s to more than 90 percent of GDP in 2015.
- The surge in debt was particularly pronounced in 2009 and 2010 at the peak of the global financial crisis, with some countries priced out of the market and fiscal stress spilling over to others.
- Literature has emphasized institutional shortcomings, fiscal stimulus and contingent liabilities, and low nominal growth as drivers of deteriorating fiscal positions; less attention has been given to political economy considerations.

### Contributions and Scope
- Conceptual: Extends analysis of political economy factors and policy distortions from unitary countries to the supranational level.
- Empirical: Reviews a series of policy biases, including the relationship between country size and budget execution.
- Normative: Focuses on correcting incentives in fiscal governance reforms rather than only strengthening economic rationale for rules.
- Scope limitation: Documents fiscal policy biases and compliance with existing fiscal rules; whether the rules are well designed is beyond the paper’s scope. The conduct of national fiscal policies when common monetary policy is constrained by the effective lower bound is not discussed.

### Political Economy Perspective
- Political economy defined broadly as interactions between the political environment and the economic system, including cross-country policy coordination and effects of fiscal developments on public confidence.

H3: Policy Biases at the National Level
- Typical fiscal policy deviations: excessive deficits and procyclicality.
- Deficit bias often arises from failure to control spending increases and tax cuts in “good times,” closely linked to procyclicality.
- Political drivers described:
  - Electoral cycles and reelection concerns leading to short-term orientation and opportunistic preelection spending.
  - Partisanship and optimistic revenue forecasts or unrealistic spending estimates.
  - Time inconsistency when policymakers deviate from previous promises after agents adjust expectations.
  - Common pool problem: heterogeneous groups (line ministries, levels of government, coalition parties) failing to internalize the cost of competing claims.
  - Fiscal illusion: population’s imperfect understanding of tax and debt finance and misperception of government’s intertemporal budget constraint.
- Composition biases: in adjustment episodes, incentives may favor cutting public investment and raising taxes over current expenditure restraint; public investment cuts are less disruptive and tax increases may have lower short-term visibility.

H3: Policy Biases at the Supranational Level
- EU institutional specifics that affect fiscal policy:
  - Small EU budget and staff relative to federal budgets; limited risk sharing and redistribution; no ability to borrow for general budget or tax citizens directly.
  - Complex set of fiscal rules with weaker enforcement mechanisms than in federations; limited harmonization of spending and tax policies.
- Cross-country fiscal spillovers amplified by trade linkages, confidence effects, financial market integration, and single monetary policy. Spillovers can be negative (contagion via government yields) or positive (trade channel).
- Supranational political economy issues documented:
  - Supranational common pool problem: financial integration made national bonds quasi-perfect substitutes pre-crisis, creating a common pool of financing and conditions for localized credit booms; fragmentation occurred since the crisis but nonfundamental/global factors remain important drivers of bond yields.
  - Distorted incentives over fiscal stance due to spillovers: positive spillovers may induce overly restrictive fiscal policies; negative spillovers may induce excessive loosening; individual countries may not internalize costs of profligacy in a common bond market.
  - Biased enforcement of fiscal rules: enforcement responsibility shared between the European Commission (EC) and ECOFIN; ECOFIN’s political incentives and national accountability may lead to uneven or lenient enforcement and “sinners judging sinners” dynamics, exacerbated when the number of fiscal delinquents rises.
  - Deficit of democratic legitimacy and scapegoating of the center: limited remit of the European Parliament (EP), lack of transparency in intergovernmental decision making (for example, ECOFIN has no minutes), concentration of powers in ECOFIN, complexity of governance, and multiplicity of monitoring schemes—these factors undermine transparency, ownership of the Stability and Growth Pact (SGP), and public trust; national policymakers may attribute politically costly decisions to Brussels.

### Empirical Evidence and Methods
- Purpose: Assess whether the governance framework (SGP) has been effective in correcting policy biases.
- Data: Analysis relies on two databases covering 19 EA countries over 1999–2015:
  - AMECO database (EC’s Directorate General for Economic and Financial Affairs) for ex post outturns.
  - Real-time database built from vintages of Stability and Convergence Programs (plans and real-time realizations).
- SGP design: aims to prevent negative spillovers and help create buffers while leaving room for stabilization via automatic stabilizers and limited discretionary policy. Country-specific structural balance targets (medium-term objectives, MTOs) and numerical limits (benchmark MTO structural deficit ceiling of 0.5 percent of potential GDP and a 3 percent nominal deficit ceiling) are key features.

H3: Evidence on Procyclicality
- Descriptive evidence (scatterplot of changes in output gaps vs. changes in cyclically adjusted fiscal balances, 1999–2015, 19 countries) shows a negative association in more than half of observations—discretionary fiscal policy loosened in good times and contracted in downturns—consistent with procyclical fiscal stance in practice.
- Caveats: measurement errors, omitted factors, and inability of scatterplot to distinguish planning vs. implementation sources of procyclicality.
- Econometric approach: panel regression following literature on real-time analysis:
  - Equation estimated: Δsb_{i,t} = α_i + β × Δgap_{i,t} + γ × debt_{i,t−1} + ρ_t + ε_{i,t}, where sb is structural-balance-to-potential-GDP ratio, debt is debt-to-GDP ratio, and β measures procyclicality (negative = procyclical, positive = countercyclical, insignificant = acyclical).
  - Specifications use: one-year ahead planned fiscal impulse (t|t−1), real-time outturns (t|t+1), and ex post outturns (t|2016).
- Key estimation results (Table 1, Fixed Effects Estimator; sample includes 19 euro area countries covering the period 1999–2015):
  - Plans (Δ(SB/GDP) t|t-1)
    - Initial Debt/GDP: 0.02
    - ΔOutput Gap: -0.12
    - Constant: -0.93
    - Observations: 105
    - Number of Countries: 19
    - R2: 0.33
  - Actual, Real Time (Δ(SB/GDP) t|t+1)
    - Initial Debt/GDP: 0.05***
    - ΔOutput Gap: -0.27*
    - Constant: -3.88***
    - Observations: 102
    - Number of Countries: 19
    - R2: 0.65
  - Actual, Ex Post (Δ(SB/GDP) t|2016)
    - Initial Debt/GDP: 0.04***
    - ΔOutput Gap: -0.55**
    - Constant: -3.27***
    - Observations: 224
    - Number of Countries: 19
    - R2: 0.29
  - Notes: Columns (1) and (2) rely on stability program data; column (3) uses AMECO. In absence of structural balances in AMECO, cyclically adjusted balances used for pre-2011 period in column (3). Overall (rather than primary) structural balances used due to data availability. All regressions include time and country fixed effects. Significance: *** p < 0.01, ** p < 0.05, * p < 0.1.
- Interpretation:
  - Planned fiscal impulses appear broadly acyclical (insignificant β in plans), but real-time and ex post outcomes are procyclical (negative and significant β in actual series).
  - Endogeneity concerns (reverse causality) addressed with dynamic GMM using lagged variables and trade-weighted output gaps as an additional instrument; although reducing sample size, GMM confirms evidence of procyclicality.
  - Robustness checks:
    - Including political variables (dummies for legislative and presidential election years, political fractionalization) does not materially affect results (Annex Table 3).
    - Interaction terms of output gaps with political variables generally not significant—procyclicality in the whole sample does not appear primarily driven by national political factors, suggesting supranational factors may play a role.

*Source: INTRODUCTION, wp17118*

### 10. To test the sensitivity of the results, alternative estimations excluding

### 10. To test the sensitivity of the results, alternative estimations excluding

### Sensitivity checks
- Alternative estimations excluding 2011–13 are reported (Annex Table 5). Results are not materially affected, meaning that procyclicality is not a crisis-specific phenomenon.

### Evidence of deficit bias
- Question: does procyclicality occur throughout the cycle or only during upturns? If procyclicality happens in good times but policy is acyclical or countercyclical in bad times, the asymmetric reaction produces a “deficit bias” and excessive debt accumulation (Balassone and Francese 2004).
- SGP provisions intended to ensure symmetric policy response and mitigate deficit bias:
  - ceiling on the deficit ratio of 3 percent of GDP;
  - public debt anchor at 60 percent of GDP;
  - country-specific medium-term targets in cyclically adjusted terms (MTOs) and annual adjustments toward MTOs.
- Despite these provisions, deterioration continued:
  - For EA countries over 1998–2015, the deficit exceeded the 3 percent ceiling in more than three-quarters of total country-year observations.
  - In 16 of the 19 EA countries, the average deficit over 1999–2015 was positive (in the range of 1.7 to 7.9 percent of GDP, depending on the country), exceeding the “close to balance” medium-term objective.
  - Since 1999, fiscal deficits of EA countries and other advanced economies (excluding Japan) have fluctuated around 3 percent of GDP, with the EA posting lower deficits than its peers since the beginning of the global financial crisis.

### Testing asymmetric fiscal response (specification and interpretation)
- Regression used to test asymmetry over 1999–2015 for 19 EA countries:
  - Δsb_{i,t} = α_i + β_1 × Δgap_{i,t} × D(Δgap_{i,t} > 0) + β_2 × Δgap_{i,t} × D(Δgap_{i,t} < 0) + γ × debt_{i,t−1} + ρ_t + ε_{i,t}.
- Interpretation:
  - β_1 and β_2 measure whether discretionary fiscal policy was procyclical (negative), countercyclical (positive), or acyclical (insignificant).
  - A negative β_1: in good times (Δgap>0), an increase in output gap is associated with fiscal relaxation → procyclical.
  - A negative β_2: in bad times (Δgap<0), a decrease in output gap is associated with fiscal contraction → procyclical.

### Estimation results and robustness
- Table 2: Asymmetric Policy Response and Deficit Bias (Fixed Effects Estimator)
  - Dependent variables:
    - Column (1): Δ(SB/GDP) t|t-1 (Plans)
    - Column (2): Δ(SB/GDP) t|t+1 (Actual, Real Time)
    - Column (3): Δ(SB/GDP) t|2016 (Actual, Ex Post)
  - Coefficients (with reported significance and standard errors in brackets):
    - Initial Debt/GDP:
      - Column (1): 0.01* [0.01]
      - Column (2): 0.06*** [0.01]
      - Column (3): 0.04*** [0.01]
    - ΔOutput Gap*D(ΔOutput Gap>0):
      - Column (1): -0.45*** [0.07]
      - Column (2): -0.84*** [0.20]
      - Column (3): -0.31 [0.30]
    - ΔOutput Gap*D(ΔOutput Gap<0):
      - Column (1): 0.17 [0.26]
      - Column (2): -0.13 [0.18]
      - Column (3): -0.73*** [0.22]
    - Constant:
      - Column (1): -0.48 [0.66]
      - Column (2): -4.19*** [0.66]
      - Column (3): -3.17*** [0.94]
  - Sample and estimation details:
    - Observations: Column (1) 105; Column (2) 102; Column (3) 224.
    - Number of Countries: 19 in all columns.
    - R^2: Column (1) 0.39; Column (2) 0.67; Column (3) 0.3.
    - Sources: AMECO; Stability and Convergence Programs; authors’ estimates.
    - Note: Columns (1) and (2) rely on stability program data, column (3) uses AMECO. Overall structural balances used where available; panel OLS with time and country fixed effects. t|t−1 indicates plans for period t reported in period t−1; t|t+1 indicates outturns in period t reported in period t+1; t|2016 indicates outturns in period t reported in 2016. Significance: *** p < 0.01, ** p < 0.05, * p < 0.1.
- Key findings from estimations:
  - Estimations confirm asymmetric response to the cycle and provide evidence of a deficit bias.
  - Lax policies in good times have prevented countries from rebuilding buffers when economies grew above potential (columns (1) and (2)).
  - GMM regressions corroborate the finding while addressing endogeneity but with a large drop in observations (Annex Table 2).
  - Results are broadly robust to inclusion of political economy variables (Annex Table 4) and exclusion of periods of fiscal consolidation after 2009–10 (Annex Table 6).

### Evidence of suboptimal fiscal composition
- Political economy can produce suboptimal budget composition even with unchanged stance, notably during fiscal consolidation:
  - Tendency to preserve current spending and raise distortionary taxes, undermining growth potential (IMF 2015a).
- Policy response to protect public investment:
  - “Investment clause” in 2015 flexibility guidelines allows temporary deviation from MTOs to expand investment with positive, direct, and verifiable long-term budgetary effects on growth and fiscal sustainability.
  - Applicability limited by strict eligibility conditions (Melyn, Van Meensel, and Van Parys 2015).
- Empirical patterns during the global financial crisis:
  - Fiscal adjustment during 2010–13 in EA was tilted toward revenue measures compared with the United States and other advanced economies (Figure 4).
  - Expenditure composition shifts: brunt of adjustment fell on investment spending rather than current spending (Figure 5).
  - Rebased ratio to GDP series (2007 = 100) show capital spending declined relative to wages, goods and services, interest, and other spending (figure data in text: series from 2007–2016 with index values between 70 and 130).
- Possible legitimate economic drivers of composition changes noted (e.g., shift to private provision, protecting social transfers when consumers are credit constrained), but:
  - In the EA, case for lowering taxes on labor and capital and promoting public investment is compelling (IMF 2016b).

### Investment gap and public capital erosion
- Proxy for investment gap: difference between observed public investment and investment that would stabilize the ratio of the stock of public capital to GDP at its initial level.
- Capital-stabilizing net investment ratio estimated as y/(1+y) × k, where:
  - y = real potential growth rate;
  - k = initial stock of public capital as a ratio of GDP.
- Data and results:
  - Initial level uses 2013 stock of capital (most recent available).
  - Analysis conducted for 15 EA countries with capital stocks available from IMF’s public debt database.
  - As of 2015, the investment gap ranged between 0.35 percent and 3.7 percent of GDP, depending on the country.
- Implication: steady depletion of public capital stock could slow long-term potential and adversely affect fiscal sustainability.

### Compliance with fiscal rules: incidence and trend
- Simplified view: SGP treated as four main numerical rules for compliance analysis:
  - 3 percent of GDP nominal deficit ceiling;
  - 60 percent of GDP debt ceiling;
  - MTO in structural terms;
  - benchmark fiscal effort of at least 0.5 percent per year in structural terms when structural balances are below MTOs or under EDP.
- Caveats on the compliance exercise (Figure 6):
  - Based on ex post AMECO data and not correcting for ESA95 to ESA2010 classification changes.
  - Targets assumed similar across countries and constant over time.
  - Does not account for activation of escape clauses or other flexibility provisions.
  - Covers all 19 EA countries including those joining euro after 1999.
  - Numerical deviations may not equate to non-compliance due to EC economic judgment in preventive and corrective arms.
  - Includes crisis years when breaches were systematic.
- Compliance outcomes 1999–2015:
  - MTO violated in 80 percent of observations; almost two-thirds of countries exceeded MTOs in every single year.
  - In 2009: MTO violated by 90 percent of countries; debt ceiling violated by 50 percent; deficit ceiling violated by 85 percent; required fiscal effort violated by 75 percent.
  - Share of countries with debt ratio > 60 percent rose from 35 percent in 1999 to 75 percent in 2015.
  - Governance reforms over 2005–13 (increased flexibility, greater automaticity in enforcement, revisions in national legislation) have not had an evident impact on compliance (without correcting for other factors).

### Drivers of weak compliance: ex ante vs ex post
- Accounting decomposition:
  - [Outcome − Target] (Ex post compliance) = [Plan − Target] (Ex ante compliance) + [Outcome − Plan] (Execution).
- Analysis for EA “non-compliers” over 1998–2013 focused on 3 percent deficit and debt rules.
- Key finding for the 3 percent deficit rule (Figure 7, three-year planning horizon):
  - Main driver of poor ex post compliance was weak execution of plans (outcome − plan), not lack of ambition in plans.
  - Non-compliers consistently planned to reduce deficits below the 3 percent threshold in projected years, but execution slippages offset plans.
  - Median upward deviation from the 3 percent ceiling reached up to 2 percent of GDP at the end of the third year.
  - Given that the EC did not apply fines or sanctions in these cases, weak enforcement is implicated.

*Source: IMF authors’ estimates and analysis (content unit: wp17118).*

### 2016. Decomposition formula: 푫푬푭

### 2016. Decomposition formula: 푫푬푭

### Decomposition of Rule Slippages (deficit and debt)
- Decomposition formula for deficits (presented in the source):
  - 푫푬푭풕|ퟐퟎퟏퟔ − ퟑ = (푫푬푭풕|풕−풊 − ퟑ) + (푫푬푭풕|ퟐퟎퟏퟔ − 푫푬푭풕|풕−풊)
  - Labels: Ex post compliance (left term), Ex ante compliance (middle term), Execution (right term).
- Key observation:
  - The three components (ex post compliance, ex ante compliance, and execution) do not add up because of the use of medians.
- For the 60 percent debt-to-GDP rule the analogous decomposition is:
  - 푫푬푩푻풕|ퟐퟎퟏퟔ − ퟔퟎ = (푫푬푩푻풕|풕−풊 − ퟔퟎ) + (푫푬푩푻풕|ퟐퟎퟏퟔ − 푫푬푩푻풕|풕−풊)
- Empirical finding on debt rule slippages (medians, sample note included in source):
  - Poor execution is the main culprit: non-compliers planned to reduce debt by about 10 percent of GDP (from about 80 to 70 percent of GDP on average), but the debt ratio remained broadly unchanged ex post.
  - Note: Sample includes 19 euro area countries, over 1998-2013; reported are medians; sample includes country-years with debt/GDP>60 percent. In decomposition formula, t|t−i indicates plans for period t reported in period t−i and t|2016 indicates outturns in period t reported in 2016.
- Possible reasons for execution-stage deviations (as listed in the source):
  - Unexpected nominal growth shortfall (e.g., negative inflation surprise) altering outcomes from plans.
  - Intentional biases (e.g., deliberate underestimation of the fiscal deficit in the budget).
  - Valuation changes affecting the debt stock.
  - Statistical adjustments in the perimeter of the fiscal aggregate (also reflected in the “execution” term).
- Policy practice since 2005 reform of the SGP:
  - Assessment of effective action has focused on “conditional compliance” by estimating and subtracting the contribution of economic factors to isolate the policy part.

### Country Size and Fiscal Outcomes
- Research question: Is there a relationship between country size and deviation from fiscal plans?
- Regression (equation (3) in the source):
  - Δob_{i,t|2016} − Δob_{i,t|t−1} = α_i + β×Size_{i,t} + γ×Controls_{i,t} + ρ_t + ε_{i,t}
  - Dependent variable: difference between outcome and plan for the change in the overall-balance-to-GDP ratio.
  - Size measured as share of nominal GDP of an EA member to total EA GDP.
  - Controls: forecast errors in macro variables (nominal and real GDP growth, output gap, openness, borrowing costs), initial fiscal positions (debt ratio, distance from the MTO), election cycle (dummy in parliamentary election years), and whether the country is under EDP.
  - Estimation: panel ordinary least squares with time and country fixed effects; sample period 1999–2015; sample includes 19 euro area countries.
- Main empirical result:
  - The coefficient of Size is negative and significant in most specifications: overall balances in bigger countries tend to underperform relative to plan by a larger margin.
  - Example interpretation from source: column (1) shows that when the share of a country in the EA GDP increases by 1 percent, the fiscal slippage of the overall balance increases by about 0.3 percent of GDP.
- Caution on interpretation (from source):
  - Negative correlation could be driven by other factors (e.g., larger fiscal multipliers in bigger countries making adjustment costlier; deeper financial markets; greater ECB attention to large countries).
- Table 3 reported coefficients and statistics (selected highlights preserved exactly as in source):
  - Country Size (GDP as a Share of EA GDP, Percent) coefficients across columns include: -0.32**, -0.30*, -0.29*, -0.23, -0.43***, -0.30**, -0.32***, -0.17, -1.64**, -0.63*, -0.30*, -0.35***, -0.32**, -1.99*.
  - Standard errors shown in brackets in source (e.g., [0.13], [0.16], ...).
  - Observations and panels: e.g., Observations vary by column (187, 182, 154, 185, ...); Number of Countries = 19 in most specifications.
  - R^2 reported across columns: 0.20, 0.20, 0.20, 0.22, 0.20, 0.20, 0.33, 0.19, 0.35, 0.36, 0.21, 0.20, 0.20, 0.71 (as presented).
- Evidence on EDP modalities and country size:
  - Figure 9 (described in text) shows larger countries have tended to stay longer under EDP than smaller countries; data from EC (2015) on EDP length for 19 euro area countries, period 2003–16 (no EDP before 2003).
  - Regression for EDP recommended effort (equation (4) in the source):
    - Effort_i = α + β×Size_i + γ×Controls_i + ε_i
    - Effort = average required annual fiscal effort (in structural terms) for EDP countries; sample of 39 EDP episodes over 2003–15.
  - Main empirical result on EDP recommendations:
    - β coefficients are not significant in most regressions; negative coefficient of Size is marginally significant without controls but significance vanishes with controls and in the broader sample of 28 EU countries.
  - Table 4 summary highlights preserved exactly:
    - For 19 EA countries (columns (1)–(6)), coefficients on Size in source: -0.022*, -0.021*, -0.01, 0.00, -0.01, 0.00 (with bracketed standard errors [0.01], etc.).
    - For 28 EU countries (columns (7)–(12)), coefficients on Size in source: -0.02, -0.02, -0.01, -0.01, -0.01, 0.00 (with bracketed standard errors [0.01], etc.).
    - Other reported variables with significance include Overall Balance (Percent of GDP) with coefficients like -0.054***, -0.036*** and Real GDP Growth (Percent) with coefficients like -0.135***, -0.115**.
    - Observations in panels: 47 (EA columns) and 68 (EU columns); R^2 reported across specifications (0.03, 0.07, 0.26, 0.32, 0.38, 0.54 for EA; 0.02, 0.06, 0.18, 0.26, 0.39, 0.49 for EU).
- Summary conclusion from source:
  - Evidence on biased application of rules is mixed: bigger countries tend to deviate more from fiscal plans, but this could be unrelated to enforcement; smaller countries under EDP are not requested to adjust excessive deficits faster than bigger countries (caveat: small sample).

### Past Reforms and Policy Options for the Future
- Framing:
  - Political economy factors matter for fiscal policy outcomes; past SGP reforms improved technical design but paid less attention to aligning political incentives.
- A. Past Reforms of the European Fiscal Governance Framework (key elements listed in source)
  - Four ways successive SGP reforms tried to mitigate political economy factors:
    - Fostering evenhanded treatment of countries:
      - “Reverse qualified majority” voting system adopted in 2011 gives more power to the EC by ensuring its recommendations/proposals are approved by ECOFIN unless a qualified majority votes against.
      - Creation of the European Fiscal Board—operational since October 2016—to provide independent evaluation of execution of the fiscal governance framework.
    - Making deviations more costly (financially and politically):
      - Possibility of earlier and stronger sanctions for EA countries; reverse qualified majority may increase sanction application likelihood.
      - 2012 Fiscal Compact entrusted independent institutions (fiscal councils) with monitoring implementation of fiscal rules to raise political cost of deviations.
    - Bringing more specificity in application of rules:
      - Successive reforms improved specificity and criteria for use of flexibility within the framework.
    - Making the framework more legitimate and homegrown:
      - Fiscal Compact required countries to introduce structural balance rules (MTOs) in national legislation monitored by domestic independent institutions; supranational surveillance better integrated with national budget calendar.
  - Limitations of past reforms (as described):
    - Reforms increased complexity, reduced transparency, made communication difficult.
    - Fiscal councils help but are not immune to political interference and fiscal sustainability may not be central in public debate.
    - “Reverse qualified majority” may not fully guarantee evenhandedness; EC interactions with national authorities create risk political considerations are internalized; European Fiscal Board effectiveness depends on independence and stature.
- B. Taking Into Account Political Economy Factors within the Existing Framework (recommendations and principles)
  - Two guiding principles for future reforms (from source):
    - Make sanctions more politically acceptable:
      - Current sanctions lack credibility: financial sanctions in EDP can exacerbate distress; elevated sanctions carry stigma making application unlikely.
      - Consider emphasis on sanctions under the preventive arm; consider nonpecuniary sanctions under corrective arm (e.g., administrative sanctions as in federations).
      - Consider a more gradual system: initial/small deviations entail small financial costs; repeated/larger deviations penalized more heavily.
      - Note: credible enforcement likely requires further political integration; sovereign sanctions problematic in international law.
    - Create tangible benefits for compliers:
      - Current framework is heavily tilted toward negative incentives; short-run individual benefits of complying not always clear.
      - Options: link structural funds and other EU subsidies to compliance; condition access to a central stabilization capacity on past compliance (as discussed in IMF 2016b); design the framework so policymakers can link compliance to a well-functioning EMU delivering price stability and robust growth.
  - Simplify the framework:
    - Successive reforms made SGP increasingly complex; simplifying would enhance democratic legitimacy, transparency, reduce loopholes, and limit scope for interpretation.
    - May require rethinking overall structure, consolidating preventive and corrective arms, and eliminating redundant/ill-designed rules (Eyraud and Wu 2015).
- C. Toward a More Comprehensive Approach (long-term options and trade-offs)
  - Argument: SGP reforms alone unlikely to correct policy biases rooted in EA architecture (fiscal, monetary, banking, financial interplay).
  - Two polar models along a continuum (as framed in source):
    - Decentralized model with stronger market discipline:
      - Leave large autonomy to member states, reinforce market discipline; examples: Canada, United States.
      - Prerequisites: mechanisms to manage state defaults orderly; avoid unconditional bailouts; enhance fiscal transparency and disclosure of government balance sheets; credible joint fiscal backstop (as part of banking union) to weaken sovereign-bank links.
      - Note: subcentral fiscal rules self-imposed to signal commitment to markets in successful federations.
    - Centralized model with deeper fiscal and political integration:
      - Stronger central oversight and control over regional policies; would require member states to surrender some fiscal sovereignty.
      - Potential benefits: better management of EA aggregate fiscal stance; mitigate common pool problem; improve distribution of fiscal effort; use fiscal space for aggregate public investment; better policy mix reducing burden on ECB.
      - Political integration required to underpin legitimacy and accountability of stronger fiscal center (Issing 2016): stronger democratic feedback mechanisms, greater role for directly elected institutions, potential establishment of central executive authority (e.g., Minister of Finance of EMU).
  - Current EA stance (as described):
    - EA combines elements of both models without fully achieving either.
    - Maastricht “no-bailout” clause lacked credibility; market discipline did not work properly in EMU.
    - SGP shares some centralized features but member states retain most fiscal sovereignty and EC enforcement tools are limited compared with other federations.
  - Political economy implications:
    - Greater political integration could help address democratic deficit and make decisions that reflect Europe-wide interests; ECOFIN composition and national-interest guidance can create issues including stronger say for big countries and potentially more lenient treatment.
    - Recent institutions (European Stability Mechanism, banking union) increased risk sharing but more central authority could further support coordination.

*Italic source attribution: IMF working paper wp17118, 2016. Decomposition formula and related sections as provided in the source PDF.*

### CONCLUSION

### CONCLUSION

### Key findings on fiscal policy biases and incentives
- Some fiscal challenges in the EA are rooted in distorted political incentives at both the national and supranational levels.
- Evidence from real-time data in stability programs shows a range of fiscal policy biases, including:
  - procyclicality,
  - excessive deficits,
  - compositional distortions.
- The presence of national and supranational fiscal rules has not successfully alleviated these biases; they have continued to prevail following recent reforms of the fiscal framework.
- Evidence of bias in applying the rules (as a function of country size) is not as conclusive.
- The sample and data context used in the analysis include "19 euro area countries covering the period 1999–2015" and robustness checks excluding "2011–13".

### Analysis of political economy implications
- Distorted incentives are a simple result with far-reaching implications that have not been fully acknowledged in fiscal institution reform.
- A dangerous manifestation of these distortions is the perception among many analysts that European fiscal governance releases countries from their national responsibilities—whereas fiscal policy remains primarily a national responsibility.
- The combination of this misperception with the predominance of the national dimension of politics constitutes a dangerous mix.
- Further political integration could take the form of a supranational political union or tighter intergovernmental coordination.

### Policy recommendations and institutional design
- Reform efforts should continue on two fronts: right design combined with right incentives.
- Strengthening the economic basis of the rules has made significant progress, but will not bear fruit unless reforms garner political and public support.
- It is possible and desirable to have a stronger system of incentives, including:
  - gradual and proportionate sanctions,
  - clear benefits for compliers.
- In the longer term, a lasting solution must combine market discipline and stronger fiscal governance.
- Fiscal union, if it happens, should be part of a comprehensive architecture accompanying bank and capital markets unions and would reflect political choices in Europe.

### Annex and econometric context (high-level)
- Annex econometric results reported use stability program data and AMECO data; where structural balances (SB) are absent in AMECO, cyclically adjusted balances are used for the pre-2011 period.
- Dependent variables in the econometric analysis are overall (rather than primary) structural balances, with estimations relying on panel ordinary least squares and dynamic panel GMM estimators; regressions include time and country fixed effects.
- Notable sample and estimation details preserved in the Annex:
  - Sample: 19 euro area countries, period 1999–2015.
  - Alternative specifications and robustness checks reported (including exclusion of 2011–13).
  - Political-economy variables sourced from the Database of Political Institutions; definitions include Legislative election (dummy), Presidential elections (dummy), Fractionalization (continuous between 0 and 1).
  - Significance notation used in tables: *** p < 0.01, ** p < 0.05, * p < 0.1.

*Source: wp17118 - CONCLUSION (https://www.imf.org/-/media/files/publications/wp/2017/wp17118.pdf).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp17118.pdf_
