## Europe’s Debt (Un)Sustainability: Looking Through Bohn’s Magnifying Glass (WP/25/70, April 2025)

## Source details

**Canonical URL:** [Europe’s Debt (Un)Sustainability: Looking Through Bohn’s Magnifying Glass (WP/25/70, April 2025)](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025070-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025070-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025070-print-pdf.pdf.json)

---

### Overview and main conclusions
- Authors: Aleš Bulíř and Khyati Chauhan.
- Main empirical verdicts:
  - "No large European countries and only few small ones have met the so-called Bohn rule during the past 40 years or so."
  - Post-1980 European cyclically-adjusted primary balances are driven by spending growth and consumption smoothing.
  - Empirical results "change little between periods before and after the global financial crisis."

### Compliance with the Bohn rule — cross-country evidence
- Broad rejection of Bohn-type fiscal reaction:
  - For 24 out of 27 sample countries the null that debt changes are followed by primary balance changes is rejected.
  - Only countries with positive and statistically significant lagged-debt parameters in baseline estimates:
    - Belgium (Group 3): Δb_{t−1} = 0.21 (3.8)
    - Denmark (Group 2): b_{t−2} coefficient = 0.05 (2.2) and EC term = -0.57 (-3.9)
    - Slovenia (Group 1): Δb_{t−1} = 0.04 (2.1)
  - Additional positive but smaller or sample-dependent findings:
    - Hungary: Δb_{t−1} = 0.16 (1.6) in baseline (pre-GFC larger: 0.54 (3.9))
    - Norway: Δb_{t−1} = 0.076 (0.9) (Group 3)
- Speed of adjustment when present:
  - Estimated adjustment magnitudes are small (less than 0.1 for Denmark, Slovenia, Spain).
  - Example: Slovenia’s point estimate of 0.05 implies, all else equal, about 20 years to offset an unexpected increase in the debt-to-GDP ratio.

### Dominant drivers of fiscal balances (empirical regularities)
- Fiscal balances primarily driven by:
  - Cyclical spending growth — g_t coefficients largely negative and significant (examples: many countries g_t ≈ -0.7 to -1.75 with strong t-statistics).
  - Business-cycle smoothing — x_t coefficients consistently positive and often significant (examples: x_t ≈ 0.15 to 0.75 across countries).
- Fiscal reaction to debt rare; where observed, point estimates are small and/or statistically fragile.

### Sample, data and stationarity findings (1980–2022)
- Sample composition:
  - Unbalanced panel: 26 European countries (OECD members) plus the United States.
  - Primary data vintages: IMF WEO and IMF Global Debt Database, April 2024 vintages.
- Key sample statistics:
  - Sample mean of the 2018–2022 debt-to-GDP ratios: 65 percent.
  - Cluster changes in debt-to-GDP (average change between 2018–2022 and 1995–1999):
    - Cluster 1: lowered debt-to-GDP on average by 20 percentage points (Nordic countries, Belgium, Ireland, Netherlands).
    - Cluster 2: modest increase on average by about 15 percent.
    - Cluster 3: increased on average by about 60 percent (France, Greece, Slovenia, Spain, Portugal, UK, US).
- Stationarity and cointegration diagnostics:
  - Public debt is nonstationary (I(1)) in almost all countries; only Norway and Poland show strictly stationary debt in levels.
  - Primary balance is nonstationary in about one-half of countries.
  - Output gap and cyclical spending are stationary by construction.
  - Johansen tests: 17 countries exhibit some long-run relationships, but many cointegrating vectors are not statistically significant and almost half have negative signs (opposite the debt-stability implication).

### r − g differentials and market discipline
- Effective interest-growth differentials:
  - Many new EU members and small countries show large negative r − g differentials (to the tune of 2 to 4 percentage points) in historical periods.
  - Differential positive and larger than one percentage point for Italy, Denmark, Belgium, the US, Finland, France, and UK.
  - Germany had a positive, if small, differential.
- Market discipline:
  - No inverse relationship between r − g differential and change in public debt across the sample; correlation between change in debt and average effective interest differential vis-à-vis Germany is close to zero.
  - Exceptions with clearer market discipline: Belgium, Denmark, and Sweden.

### Regression methodology and group-specific estimators
- Empirical strategy:
  - Respect integration/cointegration: use Group 1 regressions (s_t on Δb_{t−1}) when debt I(1) and s_t I(0); Group 2 ECM when both I(1) and cointegrated; Group 3 Δs_t on Δb_{t−1} when both I(1) and not cointegrated.
  - Controls included: output gap x_t and cyclical government spending g_t; tests for structural break at the global financial crisis (pre-GFC 1980–2008 vs post-GFC 2009–2022).
  - Effective nominal yields measured as general government interest payments in domestic currency divided by previous-year stock of debt in domestic currency times 100.
- Representative regression diagnostics (selected examples):
  - Austria (Group 1): Δb_{t−1} = 0.01 (4.2); x_t = 0.54 (4.6); g_t = -0.56 (-4.2); R2 = 0.74.
  - Germany (Group 1): Δb_{t−1} = -0.10 (1.4); g_t = -1.01 (5.1); R2 = 0.67.
  - Belgium (Group 3): Δb_{t−1} = 0.21 (3.8); g_t = -1.25 (-5.1); R2 = 0.46.
  - Denmark (Group 2 ECM): EC term = -0.57 (-3.9); b_{t−2} = 0.05 (2.2); R2 = 0.77.

### Robustness checks and sample splits
- Euro adoption split (17 euro adopters): reject hypothesis that euro adoption changed fiscal response for all 17 countries; regression results identical to baseline. Belgium: pre- and post-euro responses both positive and statistically significant with identical point estimate 0.2.
- Adding effective interest rate differential vis-à-vis Germany:
  - Coefficient on risk premium statistically indistinguishable from zero in almost all countries.
  - Inclusion made the Bohn null not rejectable in Hungary, Lithuania, and Switzerland by reducing standard errors; left Belgium unchanged; Denmark and Slovenia lost significance on the lagged-debt parameter.
- Using WEO output-gap estimates instead of HP-filtered gaps: no change to baseline findings.
- Pre-GFC / Post-GFC stability:
  - Coefficients broadly stable across splits; selected pre- vs post-GFC lagged-debt estimates reported (examples):
    - United States: Pre-GFC -0.00 (-0.1); Post-GFC -0.08 (-1.0).
    - Belgium: Pre-GFC 0.16 (1.6); Post-GFC 0.24 (3.3).
    - Hungary: Pre-GFC 0.54 (3.9); Post-GFC 0.03 (0.4).
    - Denmark: Pre-GFC 0.03 (0.5); Post-GFC 0.05 (2.4).
    - Netherlands: Pre-GFC 0.08 (1.0); Post-GFC -0.17 (-2.9).

### Policy implications
- Fiscal risks and policy challenges:
  - IMF estimate noted in the paper: global public debt will approach 100 percent of GDP by 2030 (noted as subject to optimism bias).
  - Main concern: widespread inability or political unwillingness to stabilize debt through planned, systematic discretionary fiscal consolidations (sequences of cyclically-adjusted primary surpluses).
  - Alternative channels for debt reduction (faster growth, surprise inflation, financial repression) are unlikely near-term drivers for the sample countries.
  - Large countries (France, Italy, Spain, UK, and the US) have been accumulating debt without corresponding discretionary consolidations; Germany shows broadly stable/stationary debt over the last ~20 years.
  - Consequences: large countries’ borrowing needs may crowd out investment of smaller countries; continued lack of Bohn-type adjustment by large countries risks violating new EU fiscal rules and incentivizing rule breaches by others.
- Practical takeaway for policymakers:
  - Reliance on cyclical recovery or favorable r − g dynamics is not a robust substitute for policy-controlled primary surplus adjustments.
  - Where fiscal reaction exists, the speed is slow (example: 20 years for Slovenia’s implied offset), implying that prompt, sustained policy action would be required to stabilize debt ratios meaningfully.

### Methodological strengths and caveats
- Strengths:
  - Careful separation of I(0) and I(1) series, cointegration testing, and use of ECMs or first-difference specifications as appropriate.
  - Extensive robustness checks (euro adoption split, inclusion of effective interest differentials, alternative output-gap measures).
- Caveats highlighted in the paper:
  - Point estimates of lagged-debt response are often small and sensitive to specification choices and sample splits.
  - Cointegrating relationships, where found, are sometimes weak or sign-inconsistent with debt-stabilizing interpretation.

*Source: Annex 2. Detailed Regression Results, IMF Working Paper WP/25/70 (wpiea2025070-print-pdf).*

### 2024. Aleš

### Europe’s Debt (Un)Sustainability: Looking Through Bohn’s Magnifying Glass

### Overview
- Authors: Aleš Bulíř and Khyati Chauhan.
- IMF Working Paper WP/25/70; April 2025.
- Abstract summary:
  - "No large European countries and only few small ones have met the so-called Bohn rule during the past 40 years or so."
  - The Bohn rule specifies that "past increases of public debt need to be systematically compensated with current and future fiscal surpluses to stabilize debt at some steady-state level."
  - Main empirical finding: "post-1980 European fiscal primary balances have been driven by spending growth and consumption smoothing."
  - Comparative finding: "The results change little between periods before and after the global financial crisis."

### Key findings (from the abstract and front matter)
- Compliance with the Bohn rule:
  - No large European countries met the Bohn rule during the past 40 years.
  - Only a few small European countries met the Bohn rule during the past 40 years.
- Primary balance drivers:
  - Post-1980 European fiscal primary balances are driven by spending growth.
  - Post-1980 European fiscal primary balances are driven by consumption smoothing.
- Stability across periods:
  - Empirical results are broadly stable; they "change little between periods before and after the global financial crisis."

### Analytical framework and literature (sections listed)
- Main analytical components (section headings in the paper):
  - I. Introduction
  - II. The Analytical Framework and Survey of the Literature
    - A. Debt Dynamics
    - B. The Bohn Rule
    - C. A Selective Review of the Literature
  - III. Data and Stationarity Tests
    - A. Data Sources
    - B. Data Transformations
    - C. Stationarity Tests
    - D. Cointegration
  - IV. The Estimates of the Bohn Rule
    - A. Regression models
    - B. Regression Results for the Baseline Specification
    - C. Robustness Checks
    - Euro Adoption
  - V. Policy Implications
  - VI. Conclusions

### Data and empirical approach (as described in contents)
- Data and testing components included in the paper:
  - Data sources and country-specific variables (Annex 1).
  - Data transformations applied to the raw series.
  - Stationarity tests and cointegration analysis.
  - Regression modelling to estimate the Bohn rule, including baseline specification and robustness checks.
  - Consideration of Euro adoption effects.

### Policy implications and conclusions (sections present)
- The paper contains a dedicated section V. Policy Implications and a section VI. Conclusions.
- The abstract signals policy-relevant conclusions by linking the lack of widespread compliance with the Bohn rule to observed fiscal dynamics (spending growth and consumption smoothing), and by noting persistence of results across pre- and post-global financial crisis periods.

*Prepared by Aleš Bulíř and Khyati Chauhan; IMF Working Paper WP/25/70, April 2025.*

### Annex 2. Detailed Regression Results ...................................................................................

### Annex 2. Detailed Regression Results

### I. Introduction — main conclusions
- No evidence of a debt-stabilizing mechanism (Bohn rule) in large European countries; only a few small countries exhibit such mechanisms.
- Even where a stabilizing mechanism is estimated, the speed of adjustment is very slow: it would take up to 20 years of primary balance surpluses to offset the debt shocks observed during the global financial crisis or the COVID pandemic.
- Fiscal balances are typically driven by cyclical spending growth and cyclical-smoothing objectives rather than by corrective responses to debt.
- No statistically significant corrective mechanism of risk premiums was found—countries with much higher government paper yields than Germany behaved similarly to countries with little or no premium.
- The global financial crisis shock does not appear to have materially altered fiscal behavior; results are similar in pre- and post-global financial crisis samples and across euro adopters/non-adopters.
- Introduction of the fiscal Maastricht criteria or euro adoption/non-adoption had no measurable effect on primary balances in the sample.

### II. Conceptual framework and empirical strategy
- The Bohn rule: regress the cyclically-adjusted primary balance (s_t) on lagged public debt (b_{t-1}) and conditioning variables (output gap x_t and deviations of real government spending from trend g_t). A positive and statistically significant coefficient on lagged debt implies a fiscal response sufficient for debt stationarity.
- Standard debt-dynamics identity (variables in percent of GDP except r and g):
  - ∆b_t = s_t + ((r - g)/(1 + g)) * b_{t-1} + z_t
- Interpretation of components:
  - s_t: discretionary cyclically-adjusted primary balance (policy-controlled).
  - ((r - g)/(1 + g)) * b_{t-1}: autonomous dynamics driven by real interest rate-growth differential and exchange-rate valuation effects.
  - z_t: stock-flow adjustments and contingent liabilities (privatization receipts, valuation changes, guarantees/bailouts) not typically under fiscal control.
- Empirical approach emphasized stationarity testing and cointegration analysis, estimating the Bohn rule in an error-correction framework or in first differences if no cointegration is found, to avoid spurious regressions and inconsistent estimates when mixing I(0) and I(1) series.

### III. Literature context — selective review
- Empirical literature is inconclusive and sensitive to:
  - Sample period (papers including post-2000 or post-global financial crisis data tend to find positive but insignificant or negative debt coefficients).
  - Panel vs. single-country estimation (stability of debt-parameter estimates in panels is questionable).
  - Estimation technique (OLS, ECM, ARDL, Bayesian time-varying coefficients, Kalman filter, regime switching)—no consistent pattern across methods.
- Early results (Bohn) suggesting U.S. debt sustainability from 1916 to 2000 were not robustly replicated when post-2000 data are included.
- Studies find parameter instability, regime-switching behavior, and limited impact of the Maastricht 60-percent debt ceiling on fiscal behavior.
- For emerging economies, some evidence of stronger primary-surplus response to debt than in industrial economies, but industrial economies did not display a robust positive response beyond certain thresholds.

### IV. Data, sample, and stationarity findings
- Sample:
  - Unbalanced dataset of 26 European countries (OECD members) plus the United States.
  - Primary source: 1980–2022 annual dataset from the IMF World Economic Outlook, April 2024 vintage for fiscal balances, government spending, and real GDP.
  - Public debt ratios from the IMF Global Debt Database, April 2024 vintage.
  - Former transition economies’ series often start in the mid-1990s.
- Key sample statistics and patterns:
  - Sample mean of the 2018–2022 debt-to-GDP ratios: 65 percent.
  - Cluster findings (change in debt-to-GDP ratio between average 2018–2022 and average 1995–1999):
    - Cluster 1 (red): countries that lowered debt-to-GDP ratios between 2018–2022 and 1995–1999, on average by 20 percentage points; examples include Nordic countries, Belgium, Ireland, Netherlands.
    - Cluster 2 (blue): countries where debt increased modestly, on average by about 15 percent; mixed primary balance outcomes.
    - Cluster 3 (green): countries where debt levels increased on average by about 60 percent; examples include France, Greece, Slovenia, Spain, Portugal, and also the UK and the US.
  - Many countries ran negative cyclically-adjusted primary balances during the COVID period and immediately prior to it (2018–2022); Baltics, Slovenia, and Slovakia experienced notably large discretionary deficits in that period.
  - Nordic countries (Denmark, Norway, Sweden), Germany, Netherlands, and Luxembourg recorded either balanced primary budgets or small surpluses in recent years.
  - Government size trend: all sample countries posted positive average growth of real fiscal spending.
- Stationarity properties:
  - During 1980–2022, public debt is nonstationary, or I(1), in almost all countries.
  - Primary balance is nonstationary in one-half of the countries in the sample.
  - Mixing stationary and nonstationary variables without appropriate cointegration/error-correction treatment risks spurious regressions, inconsistent estimates, and incorrect inference.

### V. Empirical results summary
- Overall rejection of the Bohn rule for most countries in the European sample.
- Only a handful of small countries show discretionary fiscal policy responding to debt movements.
- Dominant drivers of fiscal balances are cyclical spending growth and business-cycle smoothing.
- No significant effect detected from:
  - Country-specific risk premiums vis-à-vis German sovereign yields.
  - Euro adoption or the Maastricht criteria.
  - The global financial crisis shock (pre- vs post-crisis samples produce very similar results).

### VI. Data notes and measurement choices
- Effective nominal yields: measured as general government interest payments in domestic currency divided by the previous-year stock of debt in domestic currency times 100—this captures what governments effectively pay on past borrowings and can differ from current-period sovereign yields or policy rates.
- Observed large negative r - g differentials (to the tune of 2 to 4 percentage points) in many new EU member states and several small countries, reflecting factors such as sizable concessional borrowing from European lenders in the 1990s and early 2000s.
- The COVID-related debt increases in 2020–2021 were partly offset by unexpected inflation in 2021–2022 that increased the GDP denominator and thus moderated debt-to-GDP ratios.

*Source: Annex 2. Detailed Regression Results, IMF Working Paper (wpiea2025070-print-pdf).*

### 4.3 percent, and Germany paid 4.6 percent!

### wpiea2025070-print-pdf - 4.3 percent, and Germany paid 4.6 percent!

### Key findings on r - g differentials and debt dynamics
- The difference between the average effective interest rate on government debt (푟) and the economy growth rate (푔) is presented; no large advanced country (except perhaps Netherlands and Spain) experienced an autonomous decline in debt driven by a negative 푟െ푔 differential.
- Differentials were positive and larger than one percentage point for Italy, Denmark, Belgium, the US, Finland, France, and UK.
- The differential was positive, if small, even in Germany.
- There is no inverse relationship between the 푟െ푔 differential and change in public debt across the sample (Figure 4): countries with 푟െ푔൏െ1 (Estonia to Czechia) increased their debt on average by about 10 percentage points; many countries with positive 푟െ푔 differentials borrowed more—some doubling (U.K.) or even quadrupling (France) their debt-to-GDP ratios.
- Bond market pressure appears not to have disciplined fiscal behavior: the correlation between change in debt and the average effective interest differential vis-à-vis Germany is close to zero (Figure 5). Exceptions with clear market discipline are Belgium, Denmark, and Sweden.

### Data, variable construction, and transformations
- Core variables: cyclically-adjusted primary fiscal balance (푠
௧
), public debt (푏
௧
), cyclical primary government spending (푔
௧
), and cyclical position/output gap (푥
௧
).
- All data are annual; UK and US fiscal-year data were converted to calendar year by weighted averages.
- Output gap 푥
௧
 = (푌
௧
−푌
௧
∗)/푌
௧
∗ × 100, with potential GDP 푌
௧
∗ computed using the Hodrick-Prescott filter with 휆 = 100.
- Cyclically-adjusted primary balance: (1) logs of primary revenue and primary expenditure; (2) calculate cyclically-adjusted revenues and expenditures; (3) subtract adjusted expenditures from adjusted revenues.
- Cyclical adjustment follows the OECD approach: 휏̃ = 휏 + (푦
∗
െ푦)휀
ఛ,௬ and 푔̃ = 푔 + (푦
∗
െ푦)휀
௚,௬ where elasticities 휀
ఛ,௬ and 휀
௚,௬ are country-specific (from Table 8 in Price, Dang, and Botev (2015)); Lithuania elasticities estimated as average of Latvia and Estonia.
- Cyclical real public spending 푔̃
௧
 computed both via OECD elasticity approach and as HP-filter gap; series co-move despite different magnitudes.

### Stationarity and cointegration diagnostics
- Extensive ADF and KPSS tests run on output gap (푥
௧
), cyclically-adjusted government spending (푔
௧
), cyclically-adjusted primary balance (푠
௧
), and public debt (푏
௧
); tests with constant and with constant+trend.
- Findings:
  - Almost all sample debt series are nonstationary; only Norway and Poland show strictly stationary debt in levels.
  - For more than two-thirds of the sample, debt is strictly nonstationary; five others are trend-stationary.
  - Primary balance stationarity is mixed: in two-fifths of the sample unit root cannot be rejected; other countries show stationarity. Nonstationary series are I(1) and become stationary after first-differencing.
  - Output gap and cyclical component of real spending are stationary in levels by definition and corroborated by tests.
- Johansen cointegration tests between debt-to-GDP (푏
௧
) and cyclically-adjusted primary balance (푠
௧
):
  - 17 countries exhibit some long-run relationships; many estimated cointegrating vectors are not statistically significant and almost half have negative signs (opposite the debt sustainability implication).
- Countries grouped by stationarity/cointegration for appropriate regression models (Table 5):
  - Group 1: Debt I(1), primary balance I(0): regress 푠
௧
 on ∆푏
௧ିଵ
.
  - Group 2: Both I(1) and cointegrated: estimate ECM—∆푠
௧
 on error-correction term and short-term dynamics.
  - Group 3: Both I(1) and not cointegrated: regress ∆푠
௧
 on ∆푏
௧ିଵ
.
  - Country assignments to groups are provided (e.g., Austria in Group 1; Denmark, France, Ireland, Italy, Spain, Switzerland, United Kingdom in Group 2; Belgium, Hungary, Norway in Group 3).

### Regression specifications and sample splits
- Baseline specifications respect integration/cointegration tests and include controls for output gap and cyclical spending.
- Structural break tested at the global financial crisis (GFC) of 2008; sample split into pre-GFC (1980–2008) and post-GFC (2009–2022) with interactive dummies. Rationale: separates low-volatility era and ECB policy regimes; post-GFC unweighted average policy rate was -0.14 percent.
- Group-specific baseline equations provided explicitly:
  - Group 1 baseline: 푠
௧
 = 훽 + 훼
ଵ
푥
௧
 + 훼
ଶ
푔
௧
 + 훼
ଷ
∆푏
௧ିଵ
 + 푢
௧
  (and version with time dummies d1,d2).
  - Group 2 ECM baseline: detailed equation with ∆푠
௧
 regressed on lags, ∆푏 terms, and error-correction term (specified).
  - Group 3 baseline: ∆푠
௧
 = 훽 + 훼
ଵ
푥
௧
 + 훼
ଶ
푔
௧
 + 훼
ଷ
∆푏
௧ିଵ
 + 푢
௧
  (and version with time dummies).

### Main regression results (Bohn rule estimates)
- Three main findings from full-sample baseline:
  - For the vast majority of countries, the expected positive adjustment of the primary balance to past debt (Bohn rule) is not found. For 24 out of 27 sample countries the null that debt changes are followed by primary balance changes is rejected.
  - Only Belgium (Group 3), Denmark (Group 2), and Slovenia (Group 1) show positive and statistically significant point estimates on the lagged-debt parameter.
  - Stability tests using pre- and post-GFC splits do not support a systemic change in fiscal behavior after the GFC—the coefficients are broadly stable. Positive and statistically significant pre-GFC parameters found in Hungary and Spain; post-GFC in Belgium and Slovenia.
- Magnitudes:
  - Estimated adjustments (parameters on 푏
௧ିଵ
) are very small (less than 0.1) for Denmark, Slovenia, and Spain.
  - Example: Slovenia’s point estimate of 0.05 implies, all else equal, about 20 years to offset an unexpected increase in the debt-to-GDP ratio.
  - Only Belgium and Hungary (pre-GFC) show point estimates meaningfully large and comparable to Bohn and early literature.

### Robustness checks
- Robustness modifications tested:
  1. Splitting samples before and after euro adoption for the 17 euro-adopting countries (ERM/pre-euro vs post-adoption interaction dummies).
     - Result: reject hypothesis that euro adoption changed fiscal response for all 17 countries; regression results identical to baseline. Belgium marginally differs: both pre- and post-euro fiscal responses positive and statistically significant with identical point estimate 0.2.
  2. Adding effective interest rate differential vis-à-vis Germany (contemporaneous and lagged) to baseline regressions.
     - Result: coefficient on risk premium statistically indistinguishable from zero in almost all countries.
     - Including the differential reduced standard error (increased t-value) of the lagged-debt parameter only in Hungary, Lithuania, and Switzerland—making the Bohn null not rejectable in these additional countries.
     - For Belgium, Denmark, and Slovenia (baseline nonrejections of Bohn), inclusion leaves Belgium unchanged, but Denmark and Slovenia lose statistical significance on 훼
ଷ
.
     - Explanations offered: muted price signals for most countries; effective interest differentials stable for euro adopters; ECB/Draghi policies narrowed spreads; post-GFC ECB policy rate near zero lower bound for most of 2009–2022.
  3. Re-estimating with WEO output-gap estimates instead of HP-filtered gaps.
     - Result: no change to baseline findings.
- Detailed regression results available on demand.

### Policy implications
- IMF estimated that global public debt will approach 100 percent of GDP by 2030 (noted as subject to optimism bias in projections).
- Main policy concern: widespread inability or political unwillingness to stabilize debt through planned, systematic discretionary fiscal consolidations (sequences of cyclically-adjusted primary surpluses).
- Alternative debt-reduction channels (faster growth, surprise inflation, financial repression) are unlikely near-term drivers in sample countries.
- Bohn-type backward-looking tests indicate public debt is not Bohn-sustainable in the sample (exceptions: Belgium, and to a lesser extent Denmark).
- Four countries (Ireland, Netherlands, Sweden, Switzerland) lowered or kept low debt despite insignificant 훼
ଷ
; some new member states still show debt increases from ~10 percent in the 1990s to above 40 percent.
- Worry: large countries (France, Italy, Spain, UK, and the US) have been accumulating debt without corresponding discretionary consolidations; only Germany shows broadly stable/stationary debt in the last ~20 years.
  - Consequences: large countries’ borrowing needs may crowd out investment of smaller countries; continued lack of Bohn-type adjustment by large countries risks violating new EU fiscal rules and incentivizing rule breaches by others.

### Conclusions
- On a post-1980 sample of European countries and the US, estimates of the Bohn rule do not support a systematic debt-to-primary-balance stabilizing mechanism.
- Only a handful of small countries exhibit discretionary fiscal responses to past debt movements.
- Dominant drivers of fiscal balances are cyclical spending growth and business-cycle smoothing.
- No detectable impact from: (1) effective interest rate differential vis-à-vis Germany; (2) euro introduction; or (3) the global financial crisis shock.
- Methodological strengths: careful separation of I(0) and I(1) series, cointegration testing, and regression specifications respecting series properties; results robust to additional variables.

*IMF WORKING PAPERS Europe’s Debt (Un)Sustainability, INTERNATIONAL MONETARY FUND*

### References

### wpiea2025070-print-pdf - References

### References
- Aldama, Pierre, and Jérôme Creel. 2019. "Fiscal Policy in the US: Sustainable After All?" Economic Modelling 81: 471–479. https://hal.science/hal-03384685/document.
- Balasundharam, Vybhavi, Olivier Basdevant, Dalmacio Benicio, Andrew Ceber, Yujin Kim, Luca Mazzone, Hoda Selim, and Yongzheng Yang. 2023. "Fiscal Consolidation: Taking Stock of Success Factors, Impact, and Design." IMF Working Paper No. 2023/63. International Monetary Fund, Washington, DC. https://www.imf.org/-/media/Files/Publications/WP/2023/English/wpiea2023063-print-pdf.ashx.
- Blanchard, Olivier. 2022. Fiscal Policy under Low Interest Rates. Cambridge, Massachusetts: The MIT Press.
- Bohn, Henning. 1998. "The Behavior of U.S. Public Debt and Deficits." Quarterly Journal of Economics 113: 949–963. Available at: https://www.jstor.org/stable/2586878.
- Bohn, Henning. 2008. "The Sustainability of Fiscal Policy in the United States." In Sustainability of Public Debt, edited by R. Neck and J. Sturm. 15–49. MIT Press.
- Bulíř, Aleš, and Jaromír Hurník. 2009. "Inflation Convergence in the Euro Area: Just Another Gimmick?" Journal of Financial Economic Policy 1 (4): 355–369.
- Buti, Marco, and Vitor Gaspar. 2021. "Maastricht Values." VoxEU Blog, July 8, 2021. Available at: https://cepr.org/voxeu/columns/maastricht-values.
- Cao, Yongquan, Era Dabla-Norris, and Enrico Di Gregorio. 2024. "Fiscal Discourse and Fiscal Policy." IMF Working Paper No. 2024/194. International Monetary Fund, Washington, DC. https://www.imf.org/-/media/Files/Publications/WP/2024/English/wpiea2024194-print-pdf.ashx.
- Carrière-Swallow, Yan, and José Marzluf. 2023. "Macrofinancial Causes of Optimism in Growth Forecasts," IMF Economic Review, vol. 71 (2): 509–537.
- Cassou, Steven P., Hedieh Shadmani, and Jesús Vázquez. 2017. "Fiscal Policy Asymmetries and the Sustainability of US Government Debt Revisited." Empirical Economics 53: 1193–1215.
- Chalk, A. Nigel, and Richard Hemming. 2000. "Assessing Fiscal Sustainability in Theory and Practice." IMF Working Paper No. 2000/81. International Monetary Fund, Washington, DC. https://www.imf.org/en/Publications/WP/Issues/2016/12/30/Assessing-Fiscal-Sustainability-in-Theory-and-Practice-3561.
- Checherita-Westphal, Cristina, and Václav Žďárek. 2017. "Fiscal Reaction Function and Fiscal Fatigue: Evidence for the Euro Area." ECB Working Paper Series No. 2036.
- Collignon, Stefan. 2012. "Fiscal Policy Rules and the Sustainability of Public Debt in Europe." International Economic Review 53 (2): 539–567. https://www.jstor.org/stable/23251598.
- Darvas, Zsolt, Lennard Welslau, and Jeromin Zettelmeyer. 2024. "The Implications of the European Union’s New Fiscal Rules." Bruegel Policy Brief, June 2024. Brussels: Bruegel. https://www.bruegel.org/policy-brief/implications-european-unions-new-fiscal-rules.
- Draghi, Mario. 2012. Speech by Mario Draghi, President of the European Central Bank at the Global Investment Conference in London, July 26, 2012. https://www.ecb.europa.eu/press/key/date/2012/html/sp120726.en.html.
- Ghosh, Atish R., Jun I. Kim, Enrique G. Mendoza, Jonathan D. Ostry, and Mahvash S. Qureshi. 2013. "Fiscal Fatigue, Fiscal Space and Debt Sustainability in Advanced Economies." The Economic Journal 123: F4–F30.
- International Monetary Fund. 2002. "Assessing Sustainability." IMF Policy Paper, Policy Development and Review Department. https://www.imf.org/external/np/pdr/sus/2002/eng/052802.pdf.
- International Monetary Fund. 2003. "Public Debt in Emerging Markets: Is It Too High?" World Economic Outlook, September 2003, Chapter III, International Monetary Fund, Washington, DC. https://www.imf.org/~/media/Websites/IMF/imported-flagship-issues/external/pubs/ft/weo/2003/02/pdf/_chapter3pdf.ashx.
- International Monetary Fund. 2021. "Review of the Debt Sustainability Framework for Market Access Countries." IMF Policy Paper No. 2021/003. Policy Development and Review Department, International Monetary Fund, Washington, DC. https://www.imf.org/-/media/Files/Publications/PP/2021/English/PPEA2021003.ashx.
- International Monetary Fund. 2022. "Staff Guidance Note on the Sovereign Risk and Debt Sustainability Framework for Market Access Countries,” International Monetary Fund, Washington, DC. https://www.imf.org/en/Publications/DSA/sovereign-risk-and-debt-sustainability-analysis-for-market-access-countries.
- International Monetary Fund. 2024. "Putting a Lid on Public Debt." IMF Fiscal Monitor. International Monetary Fund, Washington, DC. https://www.imf.org/-/media/Files/Publications/fiscal-monitor/2024/October/English/ch1.ashx.
- Koen, Vincent and Paul van den Noord. 2005. "Fiscal Gimmickry in Europe: One-off Measures and Creative Accounting." Economics Department Working Papers, No. 417, OECD, Paris. https://www.oecd-ilibrary.org/economics/fiscal-gimmickry-in-europe_237714513517.
- Mauro, Paolo, Rafael Romeu, Ariel Binder, and Asad Zaman. 2013. "A Modern History of Fiscal Prudence and Profligacy," IMF Working Paper No. 2013/5. International Monetary Fund, Washington, DC. https://www.imf.org/external/pubs/ft/wp/2013/wp1305.pdf.
- Mendoza, G. Enrique, and Jonathan D. Ostry. 2008. "International Evidence on Fiscal Solvency: Is Fiscal Policy 'Responsible'?" Journal of Monetary Economics 55: 1081–1093. https://www.sciencedirect.com/science/article/abs/pii/S0304393208000779.
- Paniagua, Jordi, Juan Sapena, and Cecilio Tamarit. 2017. "Fiscal Sustainability in EMU Countries: A Continued Fiscal Commitment?" Journal of International Financial Markets, Institutions & Money 50: 85–97.
- Papadopoulos, Athanasios, and Moïse Sidiropoulos. 1999. "The Sustainability of Fiscal Policies in the European Union." International Advances in Economic Research, International Atlantic Economic Society 5 (3): 289–307.
- Plödt, Martin, and Claire A. Reicher. 2015. "Estimating Fiscal Policy Reaction Functions: The Role of Model Specification." Journal of Macroeconomics 46: 113–128. https://doi.org/10.1016/j.jmacro.2015.08.005.
- Price, Robert, Thai-Thanh Dang, and Jarmila Botev. 2015. "Adjusting Fiscal Balances for the Business Cycle: New Tax and Expenditure Elasticity Estimates for OECD Countries." OECD Economics Department Working Papers No. 1275. Paris: OECD Publishing. https://doi.org/10.1787/5jrp1g3282d7-en.
- Reinhart, Carmen M., Jacob F. Kirkegaard, and M. Belen Sbrancia. 2011. "Financial Repression Redux." Finance & Development, June 2011: 22–26. https://www.imf.org/external/pubs/ft/fandd/2011/06/pdf/reinhart.pdf.

### Annex 1. Country Specific Variables — overview
- The annex graphically presents the five sets of variables used in the paper for all 27 sample countries during 1980–2022 or whenever the series starts:
  - (1) the output gap,
  - (2) cyclical government spending,
  - (3) the cyclically-adjusted primary balance,
  - (4) the debt-to-GDP ratio, and
  - (5) effective interest rate differential vis-à-vis Germany.

### Figures and data sources (figure notes and specifics)
- Figure A1.1 Output Gap in Percent of Potential GDP:
  - Source: Authors’ calculations, and IMF World Economic Outlook Database (April 2024).
  - Notes: The blue line represents output gap calculated by the authors using the HP filter and the dotted red line reports the output gap from the World Economic Outlook Database (April 2024).
  - Country coverage displayed: Austria; Belgium; Czechia; Denmark; Estonia; Finland; France; Germany; Greece; Hungary; Iceland; Ireland; Italy; Latvia; Lithuania; Luxembourg; Netherlands; Norway; Poland; Portugal; Slovakia; Slovenia; Spain; Sweden; Switzerland; United Kingdom; United States.
- Figure A1.2 Cyclical Component of Real Government Spending in percent of GDP:
  - Source: Authors’ calculations.
  - Notes: Country panels for the same set of sample countries as above; data plotted on a -10 to 10 scale for percent of GDP.
- Figure A1.3 Cyclically-Adjusted Primary Balance in Percent of Potential GDP:
  - Source: Authors’ calculations, and IMF World Economic Outlook Database (April 2024).
  - Notes: The blue line represents estimates based on authors’ calculation of the output gap and the dashed red line represents the estimates from the WEO Database (April 2024). * WEO does not publish the cyclically-adjusted primary balance for Norway.
- Figure A1.4 Debt-to-GDP Ratio in Percent:
  - Source: IMF Global Debt Database (April 2024).
  - Notes: Country panels show debt-to-GDP from 1980 to 2022 on a 0–140 percent scale.
- Figure A1.5 Effective Interest Rate Differential vis-à-vis Germany In Percent:
  - Source: Authors’ calculations.
  - Notes: The blue line is the ex-post effective real interest rate; the dashed red line is the rate of growth of real GDP; and the green dashed line measures the difference between the domestic effective interest rate and the effective interest rate of Germany. Country panels cover the full sample of 27 countries and plot r, g, and r - g differential on a -10 to 10 scale (with some country panels showing -8 to 8).

*Source: wpiea2025070-print-pdf - References, Annex 1 (IMF Working Paper).*

### Annex 2. Detailed Regression Results

### Annex 2. Detailed Regression Results

### Baseline Results for Group 1 (Debt is I(1) and primary balance is I(0); regress s_t on Δb_{t−1})
- Austria
  - Δb_{t−1}: 0.01 (4.2)
  - x_t: 0.54 (4.6)
  - g_t: -0.56 (-4.2)
  - Const.: -0.99 (-1.5)
  - AR(1): 0.71 (4.2)
  - R2: 0.74
  - F-statistic: 20.65
  - DW: 1.79
- Czechia
  - Δb_{t−1}: -0.01 (-0.1)
  - x_t: 0.39 (5.3)
  - g_t: -0.66 (-6.3)
  - Const.: -1.69 (-1.6)
  - AR(1): 0.82 (6.8)
  - R2: 0.88
  - F-statistic: 35.29
  - DW: 1.75
- Estonia
  - Δb_{t−1}: 0.04 (0.4)
  - x_t: 0.18 (3.4)
  - g_t: 0.18 (-4.8)
  - Const.: -0.62 (-1.4)
  - AR(1): 0.44 (2.4)
  - R2: 0.66
  - F-statistic: 9.55
  - DW: 2.41
- Finland
  - Δb_{t−1}: 0.02 (0.4)
  - x_t: 0.60 (6.4)
  - g_t: -0.87 (-3.6)
  - Const.: -0.22 (-0.2)
  - AR(1): 0.81 (7.6)
  - R2: 0.91
  - F-statistic: 90.40
  - DW: 2.45
- Germany
  - Δb_{t−1}: -0.10 (1.4)
  - x_t: 0.15 (1.1)
  - g_t: -1.01 (5.1)
  - Const.: 0.13 (0.6)
  - AR(1): (not included due to multicollinearity)
  - R2: 0.67
  - F-statistic: 19.38
  - DW: 0.71
- Greece
  - Δb_{t−1}: -0.08 (1.7)
  - x_t: -0.07 (0.7)
  - g_t: -0.74 (4.3)
  - Const.: -0.98 (2.2)
  - AR(1): (not included due to multicollinearity)
  - R2: 0.48
  - F-statistic: 11.41
  - DW: 0.87
- Iceland
  - Δb_{t−1}: -0.12 (-2.3)
  - x_t: 0.41 (2.7)
  - g_t: -0.95 (-5.9)
  - Const.: 0.12 (0.2)
  - AR(1): 0.16 (0.9)
  - R2: 0.62
  - F-statistic: 14.19
  - DW: 2.00
- Latvia
  - Δb_{t−1}: -0.01 (-0.2)
  - x_t: 0.22 (4.5)
  - g_t: -0.78 (-5.5)
  - Const.: -1.65 (-2.9)
  - AR(1): 0.64 (4.3)
  - R2: 0.84
  - F-statistic: 24.62
  - DW: 1.45
- Lithuania
  - Δb_{t−1}: -0.01 (-0.1)
  - x_t: 0.25 (3.3)
  - g_t: -0.74 (-5.4)
  - Const.: -1.13 (-1.8)
  - AR(1): 0.51 (2.6)
  - R2: 0.73
  - F-statistic: 13.64
  - DW: 1.66
- Luxembourg
  - Δb_{t−1}: 0.03 (0.3)
  - x_t: 0.34 (3.1)
  - g_t: -0.77 (-4.3)
  - Const.: 0.77 (1.2)
  - AR(1): 0.69 (4.2)
  - R2: 0.77
  - F-statistic: 17.92
  - DW: 2.06
- Netherlands
  - Δb_{t−1}: -0.02 (-0.7)
  - x_t: 0.38 (4.4)
  - g_t: -0.99 (-9.6)
  - Const.: 0.04 (0.1)
  - AR(1): 0.68 (5.4)
  - R2: 0.86
  - F-statistic: 54.89
  - DW: 1.84
- Poland
  - Δb_{t−1}: -0.02 (-0.4)
  - x_t: 0.41 (3.3)
  - g_t: -0.67 (-4.3)
  - Const.: -1.69 (-4.1)
  - AR(1): 0.49 (2.7)
  - R2: 0.66
  - F-statistic: 10.63
  - DW: 1.82
- Portugal
  - Δb_{t−1}: 0.03 (1.2)
  - x_t: 0.41 (5.7)
  - g_t: -1.09 (-8.8)
  - Const.: -0.20 (-0.2)
  - AR(1): 0.85 (9.4)
  - R2: 0.85
  - F-statistic: 48.03
  - DW: 1.56
- Slovakia
  - Δb_{t−1}: -0.01 (-0.2)
  - x_t: 0.23 (2.7)
  - g_t: -0.78 (-8.5)
  - Const.: -2.69 (-3.2)
  - AR(1): 0.72 (5.1)
  - R2: 0.82
  - F-statistic: 24.53
  - DW: 2.38
- Slovenia
  - Δb_{t−1}: 0.04 (2.1)
  - x_t: 0.40 (8.2)
  - g_t: -0.97 (-19.8)
  - Const.: -1.87 (-2.8)
  - AR(1): 0.81 (5.7)
  - R2: 0.96
  - F-statistic: 132.05
  - DW: 1.65
- Sweden
  - Δb_{t−1}: -0.07 (-1.3)
  - x_t: 0.60 (5.9)
  - g_t: -1.24 (-6.0)
  - Const.: 0.74 (1.6)
  - AR(1): 0.64 (4.5)
  - R2: 0.92
  - F-statistic: 98.24
  - DW: 1.72
- United States
  - Δb_{t−1}: -0.01 (-0.2)
  - x_t: 0.41 (3.7)
  - g_t: -0.94 (-7.5)
  - Const.: -3.72 (-1.0)
  - AR(1): 0.95 (16.2)
  - R2: 0.94
  - F-statistic: 130.75
  - DW: 1.06

Notes for Group 1
- Point estimates significant at the 5 percent level are indicated in bold, estimates significant at the 10 percent level are indicated in italics. t-statistics in parentheses. Only positive and statistically significant estimates indicate Bohn-type fiscal sustainability.
- The regressions for Germany and Greece do not include an AR(1) term due to multicollinearity.

### Baseline Results for Group 2 (Both debt and primary balance are I(1) & cointegrated; estimate ECM: regress Δs_t on γ(s_{t−1} − αΔb_{t−2} − β))
- Denmark
  - Δb_{t−1}: 0.02 (0.4)
  - x_t: 0.75 (4.9)
  - g_t: -0.43 (-1.6)
  - Δb_{t−2}: -0.04 (-0.7)
  - x_{t−1}: -0.25 (-1.5)
  - g_{t−1}: 0.31 (1.0)
  - EC term: -0.57 (-3.9)
  - b_{t−2}: 0.05 (2.2)
  - Cons.: -0.62 (-0.5)
  - AR(1): 0.77
  - R2: 0.77
  - F-statistic: 10.94
  - DW: 1.74
- France
  - Δb_{t−1}: -0.01 (-0.2)
  - x_t: 0.52 (8.2)
  - g_t: -0.82 (-2.8)
  - Δb_{t−2}: 0.01 (0.1)
  - x_{t−1}: -0.47 (-4.7)
  - g_{t−1}: 1.21 (4.6)
  - EC term: -0.23 (0.1)
  - b_{t−2}: -0.03 (-1.6)
  - Cons.: 0.30 (0.2)
  - AR(1): 0.88
  - R2: 0.88
  - F-statistic: 25.00
  - DW: 2.36
- Ireland
  - Δb_{t−1}: 0.02 (0.4)
  - x_t: 0.24 (2.4)
  - g_t: -0.90 (-9.1)
  - Δb_{t−2}: -0.01 (-0.2)
  - x_{t−1}: -0.13 (-1.3)
  - g_{t−1}: 0.73 (4.8)
  - EC term: -0.20 (-1.9)
  - b_{t−2}: 0.19 (1.6)
  - Cons.: -12.46 (-1.5)
  - AR(1): 0.87
  - R2: 0.87
  - F-statistic: 21.61
  - DW: 2.19
- Italy
  - Δb_{t−1}: -0.05 (-0.6)
  - x_t: 0.45 (4.1)
  - g_t: -0.58 (-2.6)
  - Δb_{t−2}: -0.07 (-1.1)
  - x_{t−1}: -0.58 (-3.4)
  - g_{t−1}: 0.56 (2.8)
  - EC term: -0.44 (-2.2)
  - b_{t−2}: -0.10 (-2.1)
  - Cons.: 13.62 (2.5)
  - AR(1): 0.74
  - R2: 0.74
  - F-statistic: 7.76
  - DW: 2.33
- Spain
  - Δb_{t−1}: -0.49 (-0.9)
  - x_t: 0.27 (3.3)
  - g_t: -0.79 (-3.4)
  - Δb_{t−2}: -0.15 (-2.0)
  - x_{t−1}: -0.49 (-4.0)
  - g_{t−1}: 0.86 (3.4)
  - EC term: -0.36 (-2.5)
  - b_{t−2}: 0.00 (0.0)
  - Cons.: -0.43 (-0.3)
  - AR(1): 0.77
  - R2: 0.77
  - F-statistic: 11.44
  - DW: 1.90
- Switzerland
  - Δb_{t−1}: -0.15 (-2.2)
  - x_t: 0.06 (0.7)
  - g_t: -0.90 (-4.9)
  - Δb_{t−2}: -0.10 (-1.9)
  - x_{t−1}: -0.19 (-2.1)
  - g_{t−1}: 0.63 (3.1)
  - EC term: -0.51 (-2.5)
  - b_{t−2}: 0.01 (0.2)
  - Cons.: 0.13 (0.1)
  - AR(1): 0.86
  - R2: 0.86
  - F-statistic: 17.85
  - DW: 2.01
- United Kingdom
  - Δb_{t−1}: -0.07 (-1.0)
  - x_t: 0.51 (6.9)
  - g_t: -0.74 (-4.0)
  - Δb_{t−2}: 0.05 (0.8)
  - x_{t−1}: -0.59 (-7.5)
  - g_{t−1}: 0.71 (3.1)
  - EC term: -0.16 (-2.9)
  - b_{t−2}: -0.02 (-0.4)
  - Cons.: -1.07 (-0.4)
  - AR(1): 0.93
  - R2: 0.93
  - F-statistic: 44.54
  - DW: 1.44

Notes for Group 2
- Point estimates significant at the 5 percent level are indicated in bold, estimates significant at the 10 percent level are indicated in italics. t-statistics in parentheses. Only positive and statistically significant estimates indicate Bohn-type fiscal sustainability.

### Baseline Results for Group 3 (Both debt and primary balance are I(1) & not cointegrated; regress Δs_t on Δb_{t−1})
- Belgium
  - Δb_{t−1}: 0.21 (3.8)
  - x_t: 0.35 (2.0)
  - g_t: -1.25 (-5.1)
  - Const.: -0.01 (0.0)
  - AR(1): 0.51 (3.6)
  - R2: 0.46
  - F-statistic: 7.54
  - DW: 1.84
- Hungary
  - Δb_{t−1}: 0.16 (1.6)
  - x_t: 0.34 (1.8)
  - g_t: -0.91 (-2.9)
  - Const.: -0.30 (-0.4)
  - AR(1): 0.20 (0.8)
  - R2: 0.22
  - F-statistic: 1.51
  - DW: 1.65
- Norway
  - Δb_{t−1}: 0.076 (0.9)
  - x_t: -0.07 (-0.1)
  - g_t: -1.75 (-6.7)
  - Const.: 0.35 (0.2)
  - AR(1): 0.35 (2.0)
  - R2: 0.54
  - F-statistic: 10.15
  - DW: 1.49

Notes for Group 3
- Point estimates significant at the 5 percent level are indicated in bold, estimates significant at the 10 percent level are indicated in italics. t-statistics in parentheses. Only positive and statistically significant estimates indicate Bohn-type fiscal sustainability.

### Stability Test: Point Estimates of Lagged Debt Before and After the Global Financial Crisis
- Group definitions reiterated
  - Group 1: Debt is I(1) and the primary balance is I(0); regress s_t on Δb_{t−1}.
  - Group 2: Both variables are I(1) and cointegrated; estimate an ECM: Δs_t on γ(s_{t−1} − αΔb_{t−2} − β).
  - Group 3: Both variables are I(1) and not cointegrated; regress Δs_t on Δb_{t−1}.
- Pre-GFC / Post-GFC estimates (selected entries reproduced exactly as provided)
  - United States
    - Pre-GFC: -0.00 (-0.1)
    - Post-GFC: -0.08 (-1.0)
  - Austria
    - Pre-GFC: -0.00 (-0.1)
    - Post-GFC: 0.03 (0.5)
  - Czechia
    - Pre-GFC: -0.08 (-0.4)
    - Post-GFC: -0.00 (-0.0)
  - Estonia
    - Pre-GFC: 0.80 (1.4)
    - Post-GFC: 0.02 (0.2)
  - Finland
    - Pre-GFC: 0.03 (0.4)
    - Post-GFC: 0.01 (0.2)
  - Germany
    - Pre-GFC: -0.02 (-0.2)
    - Post-GFC: 0.00 (0.1)
  - Greece
    - Pre-GFC: -0.16 (-2.3)
    - Post-GFC: -0.01 (-0.0)
  - Iceland
    - Pre-GFC: -0.01 (-0.1)
    - Post-GFC: 0.02 (0.2)
  - Latvia
    - Pre-GFC: 0.04 (0.3)
    - Post-GFC: 0.01 (0.2)
  - Lithuania
    - Pre-GFC: -0.40 (-1.6)
    - Post-GFC: 0.00 (0.1)
  - Luxembourg
    - Pre-GFC: -0.13 (-0.8)
    - Post-GFC: -0.01 (-0.3)
  - Netherlands
    - Pre-GFC: 0.08 (1.0)
    - Post-GFC: -0.17 (-2.9)
  - Poland
    - Pre-GFC: 0.08 (1.2)
    - Post-GFC: -0.02 (-0.3)
  - Portugal
    - Pre-GFC: 0.01 (0.1)
    - Post-GFC: 0.03 (0.4)
  - Slovakia
    - Pre-GFC: -0.11 (-1.0)
    - Post-GFC: 0.08 (0.8)
  - Slovenia
    - Pre-GFC: -0.06 (-0.8)
    - Post-GFC: -0.06 (-1.3)
  - Sweden
    - Pre-GFC: -0.06 (-0.9)
    - Post-GFC: 0.04 (0.7)
  - Denmark
    - Pre-GFC: 0.03 (0.5)
    - Post-GFC: 0.05 (2.4)
  - France
    - Pre-GFC: 0.02 (0.6)
    - Post-GFC: -0.08 (-1.0)
  - Ireland
    - Pre-GFC: 0.01 (0.8)
    - Post-GFC: 0.00 (0.1)
  - Italy
    - Pre-GFC: 0.10 (2.1)
    - Post-GFC: (noted) -0.00 (-0.2)
  - Spain
    - Pre-GFC: -0.10 (-1.3)
    - Post-GFC: 0.02 (0.6)
  - Switzerland
    - Pre-GFC: 0.13 (2.9)
    - Post-GFC: -0.01 (-0.2)
  - United Kingdom
    - Pre-GFC: -0.00 (-0.0)
    - Post-GFC: 0.26 (1.3)
  - Belgium
    - Pre-GFC: 0.16 (1.6)
    - Post-GFC: 0.24 (3.3)
  - Hungary
    - Pre-GFC: 0.54 (3.9)
    - Post-GFC: 0.03 (0.4)
  - Norway
    - Pre-GFC: 0.03 (0.3)
    - Post-GFC: 0.21 (1.3)

Notes for Stability Test
- Point estimates of the elasticity of the cyclically-adjusted primary fiscal balance with respect to lagged debt significant at the 5 percent level are indicated in bold, estimates significant at the 10 percent level are indicated in italics. t-statistics in parentheses. Only positive and statistically significant estimates indicate Bohn-type fiscal sustainability.

*Source: Authors’ calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025070-print-pdf.pdf_