## EXECUTIVE SUMMARY

## Source details

**Canonical URL:** [EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025149-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025149-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025149-source-pdf.pdf.json)

---

### A. Introduction
- Corporate income tax (CIT) receipts in Ireland have grown substantially over the past decade, driven by a small number of large multinational enterprises (MNEs).
- Large and concentrated CIT receipts have supported strong headline fiscal outcomes but increase exposure to unpredictable revenue swings from external shocks or policy shifts.
- Ireland faces significant spending needs in housing, infrastructure, climate mitigation/adaptation, population aging, defense, and digital transformation.
- The existing EU fiscal framework has limitations for Ireland:
  - It relies on GDP as the denominator while Modified Gross National Income (GNI*) is argued to be more appropriate to measure the domestic economy in Ireland.
  - It does not capture the disproportionately large revenue risks specific to Ireland.

### B. Calibrating a Prudent Debt Anchor (risk-based approach)
- Overall proposal:
  - Adopt a medium-term debt anchor as the cornerstone, complemented by an operational fiscal rule in the form of multi-year expenditure ceilings.
  - Use a risk-based framework to calibrate the debt anchor, accounting for potential shocks to revenue from volatile MNE operations.
  - Consider both top-down (cross-country FDI panel) and bottom-up (firm-level Compustat) quantitative approaches.
  - Use GNI* as the denominator when excess CIT revenues are transient; prefer net debt (gross debt − cash balance) when using GNI*.
- Top-down (FDI risk) simulation baseline assumptions:
  - FDI and the domestic economy: 4.25 percent (nominal) over the medium-term.
  - Nominal expenditure growth: 5 percent annually (baseline, in line with the previous government’s rule).
  - CIT assumed to grow in line with FDI; other revenues assumed to grow in line with the domestic economy.
- Top-down simulation findings:
  - If gross debt starts at about 40 percent of GDP as of end 2024, there is a 5 percent probability of exceeding 60 percent of GDP by 2030 under the stated path.
  - Simulations suggest gross debt anchors ranging from 30 to 40 percent of GDP or GNI* would put Ireland on a comfortable path over the medium term.
  - With a risk tolerance of 10 percent, the GNI*-based net debt anchor is about 40 percent.
- Bottom-up (firm risk) simulation assumptions and findings:
  - Top 10 corporate taxpayers account for more than 50 percent of Ireland’s total CIT; top 10 MNEs contributed over 95 percent of total CIT from MNEs in Ireland in the base year for firm-level mapping.
  - Firm exit (Poisson) assumption: 5 percent per year (baseline λ = 0.045 used in robustness checks).
  - A gross debt to GDP ratio of 43 percent or below would ensure the probability of breaching the 60 percent benchmark by 2030 is no more than 10 percent.
  - To ensure net debt does not breach 60 percent of GNI* by 2030 with 90 percent probability, net debt would need to be around 36 percent of GNI*, compared to the current level of about 60 percent.
  - Both approaches point toward a net debt-to-GNI* anchor of around 40 percent as prudent.
- Additional model and parameter assumptions (FDI model):
  - Cross-country panel covering 212 countries and regions, 1990–2023; small and open economy sample of 58 countries used for FDI dynamics.
  - Trend growth assumptions used in simulations: 2¼ percent real growth rate for GDP, GNI* and FDI stock; real government spending trend: 3 percent; government debt interest rate and inflation: both 2 percent per annum.

### C. Calibrating an Operational Rule
- Rationale:
  - A debt anchor alone is insufficient; it needs an operational rule to guide short-term policy and ensure compliance.
  - Expenditure-based multi-year ceilings are recommended because they are easier to monitor/enforce, allow automatic stabilizers to operate, and are less vulnerable to optimistic revenue forecasts.
- Design elements for expenditure rules:
  - Coverage, legal basis, monitoring, consequences for non-compliance, escape clause and triggering conditions, correction mechanism, independent oversight, periodic reviews, and links to annual budgets and the medium-term fiscal and structural plan.
  - Multi-year expenditure ceilings should cover all institutions under the general government classification and be informed by a revenue path consistent with medium-term macroeconomic projections.
  - Expenditure limits are defined net of discretionary revenue measures (consistent with the previous government’s net spending rule).
- Fiscal reaction functions (FRFs):
  - Typical FRF specification: p_{t+1} = β0 p_t + β1 γ_t + β2 b_t, where p_t is the primary balance, γ_t is the output gap (or other cyclical measure), and b_t denotes public debt.
  - Estimating the output gap in Ireland is challenging due to GDP distortions; alternatives include GNI*, modified domestic demand, and the unemployment gap.
  - Unemployment gap estimated with a two-state Markov switching model (Appendix I).
- Model-based optimization and implied expenditure growth:
  - Optimization results (net debt and GNI*, 2024 estimates from Budget 2025 as initial values) suggest an average nominal expenditure growth around 5 percent over the next ten years, with somewhat higher growth initially.
  - Across methods, initial 5-year spending growth averages and 10-year averages:
    - First 5 years: between 6 ¼ and 6 ½ percent.
    - First 10 years: around 5½ percent.
  - Specific FRF-implied paths (baseline control problem):
    - Baseline FRF solution: p_{t+1} = −0.43 u_t^� − 0.57 (p_t − p̄) + 0.15 (b_t − b̄).
    - Implementing the baseline rule stabilizes unemployment gap, primary balance and debt at target levels.
    - Implied expenditure growth (baseline): 6.2 percent (first 5 years average), 5.4 percent (first 10 years average), converging to 4 ¼ percent.
- Robustness and alternative FRF specifications:
  - Debt weight variations, multiplier variations, minimax (θ = 70) and robust two-model approaches produce FRF coefficient sets and expenditure growth paths broadly similar to baseline:
    - Minimax FRF coefficients: −0.73, −1.09, 0.29; expenditure growth: 6.3 percent (first 5 years), 5.3 percent (first 10 years).
    - Robust two-model FRF coefficients: −0.53, −0.89, 0.28; expenditure growth: 6.4 percent (first 5 years), 5.4 percent (first 10 years).
- Revenue shock scenario and use of buffers:
  - Illustrative permanent shock: assumed CIT windfall of about 4.5 percent of GNI* disappears over three years, lowering the revenue ratio to about 38 percent thereafter.
  - Two expenditure response scenarios:
    - Full adjustment: nominal expenditure increases only between 1 and 3 ½ percent in the first three years, later catching up.
    - Smoothing: expenditure grows at about 4 percent per annum; this postpones achievement of the debt anchor toward the end of the simulation horizon.
  - Drawing on financial assets outside net debt (e.g., the two savings funds) would cushion initial effects and still allow the debt anchor to be achieved within a reasonable timeframe.
  - Under the full-adjustment scenario, expenditure remains lower—about 90 percent of that in the baseline in level terms.

### D. Conclusion and recommended framework
- Recommended architecture:
  - A prudent medium-term debt anchor of about 40 percent (net debt to GNI*) combined with a net spending rule as the operational target.
  - Set the operational rule in terms of expenditure (net of discretionary revenue measures) with multi-year ceilings informed by cyclical conditions and debt levels.
  - Consider both gross debt to GDP and net debt to GNI* as metrics; GNI* may better reflect domestic economy size when CIT is volatile.
- Institutional and credibility elements:
  - Escape clause and triggering conditions; predefined correction mechanism for breaches; independent oversight (Irish Fiscal Advisory Council role highlighted); periodic reviews; embed links to annual budgets and the medium-term fiscal framework.
  - Integrate the two 2024 savings funds—the Future Ireland Fund and the Infrastructure, Climate and Nature Fund—into the strengthened fiscal framework.
- Public investment and rule design choices:
  - Do not exclude public investment (no “golden rule”); consider overall ceiling with subceilings for current and capital expenditure and strengthened public investment management.
  - Nominal vs. real ceilings trade-offs: nominal simpler; real preserves purchasing power but requires index choice and is harder to communicate.
  - Growth-rate ceilings vs. level ceilings: growth-rate ceilings easier to communicate and continuous.
  - Treatment of automatic stabilizers: exclude cyclical spending (e.g., unemployment benefits) to preserve countercyclical capacity.
  - Ireland’s financial assets (about 40 percent of GNI*) provide a buffer; part will count toward net debt if net debt anchor is adopted.

### Appendix and Technical Materials (high-level)
- Appendix I: Technical Appendix covering:
  - A. FDI Risks: cross-country panel VAR(1) for cyclical components of GNI and FDI; mapping FDI to capital, output, profits, and tax revenue; simulations use trend assumptions and VAR-derived shocks.
  - B. Concentration Risks: firm-level Compustat analysis for top MNEs, panel VAR(1) for firm revenue and pre-tax income ratios, Poisson exit process with baseline λ = 0.045; alternative unit-root revenue growth specification considered.
  - C. Fiscal Reaction Function: Markov switching unemployment gap estimation; state-space control problem and quadratic loss minimization; calibration and robustness checks (minimax and robust stabilization).
- Key parameter and numeric summaries preserved from the technical appendix:
  - Unemployment gap recent range: -0.3 to -0.6 percentage points.
  - Okun’s law coefficient: -0.3.
  - Implied positive output gaps: between 1 and 2 percent.
  - Autoregressive coefficient ρ: 0.72.
  - Fiscal multiplier α (baseline): 0.2; alternative multiplier: 0.4.
  - Nominal potential growth: 4 ¼ percent.
  - Nominal interest rate: 2 ½ percent.
  - Real potential growth: 2 ½ percent.
  - Inflation: 2 percent.
  - Equilibrium real interest rate: 0.5 percent.
  - Tax to GNI ratio (τ) baseline: 43 percent; alternative scenario drop to about 38 percent.
  - Objective function weights: unemployment gap = 1; fiscal impulse (R) = 10; debt anchor weight baseline = 0.5 (alternatives 1 and 0.1).
  - Baseline FRF coefficients: −0.43, −0.57, 0.15.
  - Minimax FRF coefficients: −0.73, −1.09, 0.29.
  - Robust two-model FRF coefficients: −0.53, −0.89, 0.28.
  - Expenditure growth (baseline): 6.2 percent (first 5 years average), 5.4 percent (first 10 years average), converging to 4 ¼ percent.
  - Expenditure growth (minimax): 6.3 percent (first 5 years average), 5.3 percent (first 10 years average).
  - Expenditure growth (robust stabilizing): 6.4 percent (first 5 years average), 5.4 percent (first 10 years average).
  - Minimax robustness parameter used in simulations: θ = 70.

- Implementation summary:
  - Combining a net debt-to-GNI* anchor around 40 percent, multi-year expenditure ceilings implying average nominal expenditure growth of around 5–6 percent, and institutional safeguards would enhance Ireland’s fiscal resilience to volatile MNE-driven revenues while allowing for stepped-up productive public investment within transparent, predictable limits.

*Source: IMF Working Paper — Options for Strengthening Ireland’s Fiscal Framework (EXECUTIVE SUMMARY).*

### EXECUTIVE SUMMARY ___________________________________________________________________3

### EXECUTIVE SUMMARY

### A. Introduction
- Section title: "A. Introduction" (page 4).
- Associated figure: "1. Developments of CIT Revenues" (Figure 1).

### B. Calibrating a Prudent Debt Anchor
- Section title: "B. Calibrating a Prudent Debt Anchor" (page 5).
- Associated figures:
  - "2. CIT and FDI Variation" (Figure 2).
  - "3. Simulated Macro Variables" (Figure 3).
  - "4. Simulated Debt Fan Charts: FDI Risks" (Figure 4).
  - "5. Simulated Debt Fan Charts: Firm Risks" (Figure 5).
  - "6. Debt Trajectories Under Different Expenditure Paths" (Figure 6).

### C. Calibrating an Operational Rule
- Section title: "C. Calibrating an Operational Rule" (page 11).
- Associated figures:
  - "7. Optimal Trajectories" (Figure 7).
  - "8. Revenue Ratio and Expenditure Growth" (Figure 8).
  - "9. Revenue Shock Scenario" (Figure 9).

### D. Conclusion
- Section title: "D. Conclusion" (page 15).

### Appendix and Technical Materials
- Appendix title: "APPENDIX" with "I. Technical Appendix" (page 18).
- Appendix subcomponents:
  - "A. FDI Risks" (page 18).
  - "B. Concentration Risks" (page 22).
  - "C. Fiscal Reaction Function" (page 25).

### Publication and Document Metadata
- Document series: "IMF WORKING PAPERS".
- Report title as listed: "Options for Strengthening Ireland’s Fiscal Framework".
- PDF filename provided: "wpiea2025149-source-pdf - EXECUTIVE SUMMARY ___________________________________________________________________3".
- Canonical source URL provided: "https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025149-source-pdf.pdf".

*International Monetary Fund — EXECUTIVE SUMMARY (IMF WORKING PAPERS): Options for Strengthening Ireland’s Fiscal Framework*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Background and motivation
- Corporate income tax (CIT) receipts in Ireland have grown substantially over the past decade, driven by a small number of large multinational enterprises (MNEs).
- Large and concentrated CIT receipts have supported strong headline fiscal outcomes but increase exposure to unpredictable revenue swings from external shocks or policy shifts.
- Ireland faces significant spending needs to alleviate supply constraints (notably in housing and infrastructure) and address long-term demographic and other pressures (climate mitigation/adaptation, population aging, defense, digital transformation).
- The existing EU fiscal framework has limitations for Ireland because:
  - It relies on GDP as the denominator while Modified Gross National Income (GNI*) is more appropriate to measure the domestic economy in Ireland.
  - It does not capture the disproportionately large revenue risks specific to Ireland.

### Overall proposal
- Propose strengthening Ireland’s national fiscal framework with:
  - Adoption of a medium-term debt anchor as the cornerstone.
  - An operational fiscal rule in the form of multi-year expenditure ceilings.
- The proposed framework is designed to complement Ireland’s obligations under the EU fiscal rules, while taking into consideration country-specific circumstances, notably the use of an appropriate debt concept and denominator.
- Methodology has broader applicability to other advanced and developing economies, conditional on data availability.

### Calibrating a prudent debt anchor (risk-based approach)
- A risk-based framework is used to calibrate the debt anchor, accounting for potential shocks to revenue stemming from volatile MNE operations.
- Two quantitative approaches are employed to quantify MNE-related risks:
  - Top-down: cross-country FDI panel dataset to gauge risks associated with MNE operations.
  - Bottom-up: firm-level data (Compustat) to estimate firm- or sector-specific risks given high CIT concentration.
- Debt anchors are set to limit the probability that debt will reach levels that could trigger negative market reactions (notably crossing the EU 60 percent debt-to-GDP benchmark).
- Key calibration considerations:
  - Use of GNI* as the denominator is argued to better capture Ireland’s domestic tax base when excess CIT revenues are transient.
  - Use of net debt (gross debt – cash balance) is recommended when GNI* is used as the denominator, to account for government liquid assets.

### Findings from simulations and calibration
- Baseline assumptions for top-down simulations:
  - FDI and the domestic economy assumed to grow by 4.25 percent in nominal terms over the medium-term.
  - Nominal expenditure assumed to grow by 5 percent annually (in line with the previous government’s rule).
  - CIT assumed to grow in line with FDI; other revenues assumed to grow in line with the domestic economy.
- Top-down (FDI risk) simulation results:
  - If gross debt starts at about 40 percent of GDP as of end 2024, there is a 5 percent probability of exceeding 60 percent of GDP by 2030 under the stated path.
  - Simulations suggest debt anchors ranging from 30 to 40 percent of GDP or GNI* would put Ireland on a comfortable path over the medium term.
  - With a risk tolerance of 10 percent, the GNI*-based net debt anchor is about 40 percent.
- Bottom-up (firm risk) simulation results:
  - Top 10 corporate taxpayers account for more than 50 percent of Ireland’s total CIT.
  - Firm-specific shocks calibrated using historical distributions of profits for large multinational companies in the chemical and IT sectors; an exit rate of 5 percent per year is assumed.
  - A gross debt to GDP ratio of 43 percent or below would ensure the probability of breaching the 60 percent benchmark by 2030 is no more than 10 percent.
  - To ensure net debt does not breach 60 percent of GNI* by 2030 with 90 percent probability, net debt would need to be brought down to around 36 percent of GNI*, compared to the current level of about 60 percent.
  - Both approaches suggest that a ratio of net debt to GNI* of around 40 percent would provide a prudent anchor for fiscal policy.

### Expenditure rule and fiscal reaction functions (FRFs)
- A debt anchor alone is insufficient for effective fiscal management; it needs to be complemented with an operational rule that guides short-term policy decisions.
- Recommend multi-year expenditure ceilings (expenditure-based rules) as they are more effective in reducing fiscal policy procyclicality and have been adopted by a number of countries.
- Fiscal reaction functions (FRFs) are useful to guide calibration of expenditure rules.
- Under plausible assumptions about key economic parameters and policy preferences, FRF calibration points to:
  - An average nominal expenditure growth of around 5–6 percent in the medium term.
- Simulation implications for expenditure growth:
  - The 5 percent annual nominal spending growth assumed in the baseline would ensure a downward path for debt over a ten-year period.
  - A 6-percent expenditure growth annually could still anchor debt over the medium term, although the debt trajectory would begin to trend up toward the end of the simulation horizon.
  - Higher constant expenditure growth rates may not be compatible with long-term fiscal sustainability; even 5 percent may not suffice in the very long term under the annual revenue growth assumptions used.

### Design features for an effective framework
- For the framework to be effective it should include:
  - An escape clause and triggering conditions.
  - A correction mechanism.
  - Independent oversight.
  - Periodic reviews.
  - Links to annual budgets and the medium-term fiscal and structural plan.
- The paper abstracts from discretionary revenue measures in its analysis and defines expenditure limits as net of such measures, consistent with the previous government’s net spending rule.
- Broader tax base reforms (not analyzed here) would allow higher investment spending while reducing reliance on CIT and remaining within net expenditure limits under the enhanced framework.

### Key numeric findings and targets (preserved as in source)
- Baseline growth assumptions:
  - FDI and domestic economy: 4.25 percent (nominal) over the medium-term.
  - Nominal expenditure growth: 5 percent annually (baseline); 6 percent considered in scenarios.
- Debt anchors and probabilities:
  - Gross debt starting at about 40 percent of GDP as of end 2024 → 5 percent probability of exceeding 60 percent of GDP by 2030 under baseline path.
  - Debt anchors suggested by simulations:
    - Gross debt to GDP: range from 30 to 40 percent (top-down/bottom-up results).
    - Net debt to GNI*: around 40 percent (prudent anchor with risk tolerance considerations).
    - Gross debt to GDP of 43 percent or below → ≤10 percent probability of breaching 60 percent by 2030 (bottom-up).
    - To ensure net debt does not breach 60 percent of GNI* by 2030 with 90 percent probability → net debt needs to be around 36 percent of GNI* (current level ~60 percent).
- Firm exit assumption in bottom-up simulations: 5 percent per year.
- Previous government’s nominal expenditure growth rule: annual growth not to exceed 5 percent during 2021–25.

### Conclusion
- A well-designed fiscal framework combining a prudent medium-term debt anchor (net debt to GNI* around 40 percent), operational multi-year expenditure ceilings (implying average nominal expenditure growth of around 5–6 percent), and supporting institutional features would enhance Ireland’s fiscal resilience against volatile MNE-driven revenues, help balance competing priorities, and contribute to policy certainty, predictability, and transparency conducive to confidence and investment.

*Source: IMF Working Paper — Options for Strengthening Ireland’s Fiscal Framework (EXECUTIVE SUMMARY).*

### 4.25 percent.

### wpiea2025149-source-pdf - 4.25 percent.

### C. Calibrating an Operational Rule
- A debt anchor alone is insufficient; it must be complemented by an operational rule (e.g., expenditure or fiscal balance rule) to guide short-term policy and ensure compliance.
- Expenditure rules are preferred over budget balance or revenue rules because:
  - They are easier to monitor and enforce (governments have direct control on spending).
  - They allow automatic stabilizers to operate.
  - They are less vulnerable to overly optimistic revenue forecasts.
- Design elements that matter for expenditure rules: coverage, legal basis, monitoring, and consequences for non-compliance.
- For Ireland, multi-year expenditure ceilings covering all institutions under the general government classification are recommended.
- Multi-year ceilings can be informed by a revenue path consistent with medium-term macroeconomic projections and can help prevent procyclical spending in times of positive revenue shocks.
- Fiscal reaction functions (FRFs) provide a conceptual framework to calibrate expenditure ceilings. A typical FRF specification:
  - p_{t+1} = β0 p_t + β1 γ_t + β2 b_t
  - where p_t is the primary balance, γ_t is the output gap (or other cyclical measure), and b_t denotes public debt.

### FRF estimating and calibration challenges and approaches
- Estimating the output gap is particularly difficult for Ireland due to GDP distortions from the large MNE sector.
- Alternatives to the output gap:
  - GNI* (available only at the annual frequency with delay).
  - Modified domestic demand (focuses on domestic activity, excludes external operations).
  - Unemployment gap (requires estimating natural unemployment; offers timely data and fewer revisions). Natural unemployment is estimated here via a two-state Markov switching model (Appendix I).
- FRFs can be:
  - Estimated econometrically (sensitive to data quality and structural breaks—relevant for Ireland’s volatility and 2015 structural break).
  - Calibrated using economic models (useful when historical data are limited; enables forward-looking rule design).
- Optimization and robust control methods can identify FRF parameters that align with policy objectives under uncertainty (Appendix I).

### Model-based optimization results and implied expenditure growth
- Optimization-based results (net debt and GNI*; 2024 estimates from Budget 2025 as initial values) suggest:
  - An average nominal expenditure growth around 5 percent over the next ten years, with somewhat higher growth initially, is consistent with sustainability.
- Model simulations stabilize debt at the anchor (assumed at 40 percent) and the unemployment gap at zero under plausible model parameter assumptions.
- Across methods, initial 5-year spending growth averages and 10-year averages:
  - First 5 years: between 6 ¼ and 6 ½ percent.
  - First 10 years: around 5½ percent.
- After the transition period, FRFs prescribe expenditure to grow at the same rate as nominal revenue.
- Higher spending growth initially is driven by the favorable current fiscal position and a trade-off between:
  - Closing a negative unemployment gap (which implies tighter fiscal policy).
  - Reducing a substantial primary surplus (which requires higher expenditure).
- The model assigns a higher, but not overwhelmingly larger, weight to cyclical position relative to debt, leading to a more persistent unemployment gap and allowing scope for stepped-up productive public investment in the initial period.

### Revenue shock scenario and use of buffers
- Illustrative permanent shock: assumed CIT windfall of about 4.5 percent of GNI* disappears over three years, lowering the revenue ratio to about 38 percent thereafter.
- Two expenditure response scenarios:
  - Full adjustment: nominal expenditure increases only between 1 and 3 ½ percent in the first three years, later catching up.
  - Smoothing: expenditure grows at about 4 percent per annum; this postpones achievement of the debt anchor until toward the end of the simulation horizon.
- Intermediate scenarios are possible.
- The scenario does not account for possible use of financial assets outside net debt (e.g., the two savings funds). Drawing on these assets would cushion initial effects while still allowing the debt anchor to be achieved within a reasonable timeframe.
- Note: In level terms, under the full-adjustment scenario, expenditure remains lower—about 90 percent of that in the baseline.

### D. Conclusion: Recommended fiscal framework and elements for credibility
- Recommended architecture:
  - A prudent debt anchor of about 40 percent combined with a net spending rule as an operational target.
  - Consider both gross debt to GDP and net debt to GNI as debt metrics; GNI may be a more suitable denominator as it reflects domestic economy size more accurately.
  - Operational rule set in terms of expenditure (net of discretionary revenue measures) to limit procyclicality, enhance transparency, and facilitate monitoring/enforcement.
  - Use expenditure ceilings informed by cyclical conditions and debt levels to provide stable and predictable framework.
- Additional elements to ensure credibility, flexibility, and transparency:
  - Escape clause and triggering conditions: legislative status for national fiscal rules; predefined, objective activation conditions for temporary deviations.
  - Correction mechanism: predefined remedial actions if fiscal rules are breached (e.g., automatic spending adjustments or corrective budget plans).
  - Independent fiscal oversight: the Irish Fiscal Advisory Council should monitor compliance, assess forecasts, and uphold credibility.
  - Periodic review: regular reviews to recalibrate rules as macro-fiscal conditions and structures evolve.
  - Link to annual budgets and MTFF: embed the fiscal rule in annual budgeting and align with the medium-term fiscal framework so satisfying the operational rule implies the fiscal anchor is achieved.
  - Integrate the two 2024 savings funds—the Future Ireland Fund and the Infrastructure, Climate and Nature Fund—into the strengthened fiscal framework.

### Implementation design choices for multi-year expenditure ceilings
- Nominal vs. real ceilings:
  - Nominal ceilings: simple and transparent but may fail in unstable inflation.
  - Real ceilings: preserve purchasing power, require choosing an appropriate price index (e.g., HICP, CPI, or GDP deflator) and are harder to communicate.
- Levels vs. growth rates:
  - Growth-rate ceilings are easier to communicate and continuous; if a ceiling is breached in one year, subsequent years’ growth rates must adjust.
- Treatment of cyclical expenditure:
  - Automatic stabilizers (e.g., unemployment benefits) could be excluded to avoid constraining countercyclical policy.
- Public investment:
  - Avoid excluding public investment from the ceiling (golden rule) — not recommended due to risks of creative accounting, low-return project selection, and excessive borrowing.
  - Consider an overall expenditure ceiling with separate subceilings for current and capital expenditure and strengthened public investment management.
  - A slower adjustment toward the debt anchor could allow upfront growth-enhancing investment transparently.
- Link to financial assets:
  - Ireland’s substantial financial assets (about 40 percent of GNI*) provide a buffer; part held in cash instruments will count toward net debt if a net debt anchor is adopted.
  - A significant portion will remain outside the fiscal rule scope and serve as a buffer/financing source for future pressures (e.g., ageing, climate change).
  - Net worth targets pose operational and conceptual challenges (volatility from discount rate revisions, public comprehension, no limits on borrowing for investment, illiquidity of nonfinancial assets).

*Source: IMF staff summary of the IMF Working Paper content.*

### Appendix I.  Technical Appendix

### Appendix I.  Technical Appendix

### A. FDI Risks

- Sample and data
  - Cross-country panel dataset covering 212 countries and regions from 1990 to 2023.
  - Inward FDI stock data in nominal US dollars from countries’ official external sector statistics.
  - Domestic macroeconomic and fiscal variables (GDP, GNI (GNI* for Ireland), GDP deflator, fiscal revenue, government spending, primary balance, gross domestic debt) from the IMF’s WEO database; measured in current prices in local currency.
  - GNI of sample countries deflated by their GDP deflators for estimation in real terms.
  - Sample of small and open economies: countries in the 30th to 90th percentiles for GDP size and the 50th to 99th percentiles for openness (FDI-GDP ratio), resulting in a sample of 58 countries (Appendix I. Table 1).

- Model specification
  - Two-stage approach to public debt projections.
  - Stage 1: panel VAR(1) model for cyclical components of GNI and FDI (trend-stationary):
    - Denote cyclical components as y*_{it} and i_{it}.
    - HP-filter GNI and FDI for each country; compute percentage deviations from filtered trends.
    - Model: [y*_{it}, i_{it}]' = u_i + B [y*_{i,t-1}, i_{i,t-1}]' + ε_{it}.
    - Assumptions: country fixed effects u_i ~ Gaussian; innovations ε_{it} independent, Gaussian, allowed heteroskedasticity; all countries share same autoregression matrix B.
  - Stationarity tests:
    - Im-Pesaran-Shin (IPS) for FDI: rejects null that all countries’ FDI time series have unit roots (alternative: at least one country is trend-stationary).
    - Levin-Lin-Chu (LLC) test for FDI: p-value = 0.00, rejects null that all countries have unit roots.
    - Individual Dickey-Fuller (DF) tests: null of unit root rejected at the 95 percent confidence level for most countries.
    - Ireland: DF test fails to reject unit root (driven by Global Financial Crisis turmoil and one-off intangible asset transfer in 2015).
  - Stage 2: translate FDI to GDP and MNE CIT, and then to fiscal revenue and debt.
    - Assumption: FDI translates to capital in domestic production one-for-one.
    - MNE production function: Y_{it} = A_{it} K_{it}^{α} L_{it}^{1−α}, with α being capital income share (capital share close to one for intellectual property).
    - Labor supply: perfectly elastic at fixed wage; matches increases in capital as FDI settles in Ireland. Changes: Δy_{it} = Δk_{it} = Δl_{it}.
    - Monopolistic competition: firm profits proportional to output with margin determined by demand elasticity η > 1:
      - Δτ_{it}^c = Δπ_{it} = (1/η) Δy_{it} = (1/η) Δk_{it}.
    - Personal income taxes from labor: Δτ_{it}^p = Δl_{it}.
    - Aggregate corporate income tax percentage change:
      - Δτ_t^c = s_{c,dom} Δτ_{t,c,dom} + s_i^c Δτ_{it}^c = s_{c,dom} Δτ_{t,c,dom} + (1/η) s_i^c Δk_{it} = s_{c,dom} Δτ_{t,c,dom} + (1/η) s_{c,FDI} Δk_t.
    - Shares: s_{c,dom} + s_{c,FDI} = 1. Similarly s_{p,dom} and s_{p,FDI} for personal income tax.
    - Domestic firm CIT and PIT assumed to track GNI* one-for-one: Δτ_{t,p,dom} = Δτ_{t,c,dom} = Δy_t^*.
    - Total percentage change of fiscal revenue:
      - Δτ_t = μ Δτ_t^c + (1−μ) Δτ_t^p
      - = (μ s_{c,dom} + (1−μ) s_{p,dom}) Δy_t^* + [μ / η s_{c,FDI} + (1−μ) s_{p,FDI}] Δk_t
      - where μ is the share of tax revenue from corporate income.
    - Difference between GDP and GNI* assumed to be value added generated by MNE operations; its growth rate assumed to track FDI growth:
      - Δln(Y_t − Y_t^*) = Δk_t.
    - Government budget and debt dynamics:
      - Primary balance PB_t = T_t − G_t.
      - Debt evolution: B_t = (1 + R_t) B_{t−1} − PB_t.

- Parameter assumptions for simulations
  - Cyclical components of FDI and GNI simulated from VAR(1) estimates and added to pre-determined trends that assume a 2¼ percent real growth rate for GDP, GNI* and FDI stock.
  - Trend growth for real government spending: 3 percent.
  - Government debt interest rate and inflation: both 2 percent per annum.

### B. Concentration Risks

- Sample and data
  - Firm data from Compustat: North America Fundamental Annual dataset (corporate financials) and Index Monthly Prices dataset (delisting information).
  - Selected top 100 US firms in the tech and pharmaceutical industries (GIND codes 502030, 352020, 451030, 452020, 452010) in 1990 by average 1990 revenue.
  - Delisting treatment: “bankruptcy” counted as bankrupt; “merger and acquisition” treated as survival for the whole sample period.
  - Collected top 10 MNEs’ revenue in Ireland in the base year from local sources.
  - For simulations: assume a real log trend growth of 2¼ percent for all MNEs. Other assumptions same as FDI risk model.

- Model specification
  - Two-stage approach for public debt projections under firm concentration risks.
  - Panel VAR(1) for individual firm cyclical log revenue rrev_{it} and pre-tax income to revenue ratio pri_{it}:
    - [rrev_{it}, pri_{it}]' = u_i + B [rrev_{i,t−1}, pri_{i,t−1}]' + ε_{it}.
  - Exit dynamics: add independent Poisson exiting rule with constant arrival rate λ.
    - Baseline λ = 0.045 matches average duration of 17 years in the sample (constant survival probability each period 1 − λ).
    - Average duration formula with constant λ:
      - E[mim_m(T,30)] = (1/λ) [1 − r^{−30 λ}] (formula as presented).
  - Historical observations:
    - Revenue trajectory of top 52 firms since 1990 shows bankruptcy events relatively evenly distributed across 30 years, except during two major financial crises.
    - Bankruptcy more likely during revenue declines but can occur amid steady revenue growth.
  - Top 10 MNEs in Ireland contributed over 95 percent of total CIT from MNEs in Ireland.
  - Simulated pre-tax income for firm i:
    - PI_{it} = (pp̄_i + pp̃_{it}) re p (rrev̄_{it} + rreṽ_{it}) * re_isst_{it}, where hat variables are simulated based on VAR estimates and re_isst_{it} indicates survival in market (survives with probability 1 − λ each period).
  - Total pre-tax income from MNEs:
    - Π_t = sum_{i=1}^{10} PI_{it} (then divided by top 10 pre-tax income share in all MNEs in 2024).
  - Corporate tax change mapping:
    - With Δτ_{it}^c = Δπ_{it} = (1/η) Δy_{it} = (1/η) Δk_{it}, aggregate corporate income tax percentage change:
      - Δτ_t^c = s_{c,dom} Δτ_{t,c,dom} + s_{c,FDI} Δπ_t.
    - Personal income taxes mapped similarly; remainder of procedure same as FDI risk model.

- Revenue growth model alternative
  - Individual firm revenues may be non-stationary; alternative assumption: firm growth has a unit root and revenue growth rate and income-to-revenue ratio jointly follow a VAR.
    - [rrevv_{it}, pri_{it}]' = u_i + B [rrevv_{i,t−1}, pri_{i,t−1}]' + ε_{it}.
  - Simulations with unit-root firm growth:
    - Feature more extreme upside and downside risks; asymmetry evident because declines in top MNE revenues both increase debt (through lower revenue) and reduce GDP, amplifying debt-to-GDP rises; revenue growth reduces debt-to-GDP.
    - Stationarity tests for top companies’ revenue growth inform model choice.
    - This model tends to generate wider fan charts and lower debt anchors.

### C. Fiscal Reaction Function

- Estimating the unemployment gap
  - Approach: simple Markov switching model of unemployment with two states (“low” and “high” unemployment).
  - Model specification:
    - u_t = û_{s_t} + ε_t, with ε_t ~ N(0, σ_{s_t}^2).
    - û_{s_t} switches between u_0 (low) and u_1 (high). Transition matrix:
      - P = [[p_{00}, p_{10}], [1−p_{00}, 1−p_{10}]].
  - Results:
    - Estimated unemployment in the low-unemployment state ≈ 4 ¾ percent.
    - Estimated unemployment in the high-unemployment state ≈ slightly over 10 ½ percent (with large variance; variances allowed to switch).
    - Transition matrix implies a 96 percent probability of staying in the low-unemployment regime from one period to the next if currently in the low state.
  - Rationale:
    - Historical unemployment dynamics alternate between low (stable) and high (crisis) regimes; two-state Markov model captures regime switching and yields a constant natural unemployment rate by state.

*Source: IMF staff calculations as presented in Appendix I. Technical Appendix*

### Appendix I. Table 3. Ireland: Markov Model Estimation Results

### Appendix I. Table 3. Ireland: Markov Model Estimation Results

### Unemployment regime and unemployment gap
- Based on the probability plot (Figure 2), the economy has been in the low-unemployment regime since the recovery from the financial crisis, except for a short period during the Covid-19 pandemic (Figure 2, left panel).
- The unemployment gap is defined as 푢푢
푡푡
� = 푢푢
푡푡
− 푢푢�
푠푠
푡푡
.
- Recently, the unemployment gap has been negative in the range of -0.3 to -0.6 percentage points which, using Okun’s law with a coefficient of -0.3 (Conefrey et al., 2015), translates into positive output gaps between 1 and 2 percent.

### Simple model of fiscal policy (structure and key equations)
- Unemployment gap dynamics (policy influence):
  - (1) u�
t+1 = ρ 푢푢
푡푡
� + α η
푡푡
  - ηt denotes the fiscal impulse, defined as the change in the cyclically adjusted balance expressed as a ratio to potential output (푌푌
�
푡푡
). ρ is an autoregressive parameter.
- Cyclically adjusted balance decomposition:
  - 푃푃퐵퐵
푡푡
= 푃푃퐵퐵
푡푡
푐푐푐푐
+ 푃푃퐵퐵
푡푡
푐푐
  - With revenue elasticity = 1 and expenditure elasticity = 0, and defining the output gap 훾훾
푡푡
= (푌푌
푡푡
− 푌푌
�
푡푡
)/푌푌
�
푡푡
, the cyclical component becomes 푃푃퐵퐵
푡푡
푐푐 = τ 훾훾
푡푡
푌푌
�
푡푡
, and in ratios: p
t = p
t
c a
 + τ γ
t
.
- Fiscal impulse representation:
  - η
푡푡
:= Δ p
푡푡+1
c a
 = p
푡푡+1
− τ γ
푡푡+1
− (p
푡푡
− τ γ
푡푡
)
  - (2) p
푡푡+1
= p
푡푡
+ τ γ
푡푡+1
− τ γ
푡푡
+ η
푡푡
- Public debt evolution:
  - (3) b
푡푡+1
=
(1 + r)/(1 + g) p
푡푡
b
푡푡
− p
푡푡+1
  - where r and g denote interest rate and potential growth rate, respectively.
- Okun’s law connects output and unemployment gaps: 푢푢
푡푡
� = β γ
푡푡
.
- State-space form and control problem:
  - State equation (4): y
푡푡+1
= A y
푡푡
+ B η
푡푡
+ C v
푡푡
.
  - Quadratic loss (5): minimize E0 Σ∞
t=0 1/2 [ (y
t
− y
�
)⊤ Q (y
t
− y
�
) + η
t
⊤ R η
t
 ] subject to dynamics.
  - Target vector y
� = [0, p̅, b̅]⊤; Q has weights w
u on unemployment gap and w
b on debt; R is scalar weight on fiscal impulse.

### Calibration, parameter values, and baseline assumptions
- Autoregressive coefficient for the unemployment gap: ρ = 0.72.
- Fiscal multiplier for unemployment (α): set at 0.2 (an OLS estimate of (1), half of Monacelli et al. (2010) estimate).
- Okun’s law parameter: β = -0.3 (Conefrey et al., 2015).
- Nominal potential growth and interest rate: 4 ¼ and 2 ½ percent respectively (2 ½ percent real potential growth, 2 percent inflation and 0.5 percent equilibrium real interest rate).
- Tax to GNI ratio (τ): 43 percent (baseline).
- Objective function weights:
  - weight on unemployment gap = 1.
  - weight on fiscal impulse (control variable) R = 10.
  - weight on the debt anchor (w
b) = 0.5 in the baseline; alternative scenarios consider w
b = 1 and w
b = 0.1.

### Baseline fiscal reaction function (FRF) and interpretation
- Baseline FRF (solution to (6)-(7) with normal disturbances):
  - p
푡푡+1
= −0.43 u
푡푡
� − 0.57 (p
푡푡
− p̅) + 0.15 (b
푡푡
− b̅)
- Interpretation:
  - The rule prescribes fiscal tightening (higher primary balance) when the unemployment gap is negative (output gap positive) and when debt is above target.

### Optimal trajectories, expenditure growth, and transitional dynamics
- Implementing the baseline rule stabilizes unemployment gap, primary balance and debt at target levels.
- Primary surpluses need to be maintained for the first 5 years to allow net debt to converge to the desired level; surpluses decline gradually.
- Under the constant revenue-to-GDP ratio assumption, implied expenditure growth:
  - average rate about 6.2 percent in the first 5 years;
  - average rate about 5.4 percent in the first 10 years;
  - eventual convergence of expenditure growth to 4 ¼ percent in line with the potential growth rate.

### Sensitivity analysis (alternative weights and multiplier)
- FRF coefficients under alternative parametrizations (hats denote deviations from targets):
  - Baseline: -0.43  -0.57  0.15
  - Debt weight of 1: -0.46  -0.65  0.21
  - Debt weight of 0.1: -0.36  -0.40  0.07
  - Multiplier of 0.4: -0.56  -0.68  0.15
- Findings:
  - Increasing the weight on debt (to 1) strengthens the FRF response to debt deviations.
  - Increasing the fiscal multiplier for unemployment (to 0.4 as in Monacelli et al. (2010)) increases the response to the unemployment gap; reaction to debt remains similar.
  - Differences in average expenditure growth in the first 5 years across scenarios are less than 0.5 ppts; differences essentially disappear over a 10-year horizon.

### Robustness to uncertainty: minimax and robust stabilization approaches
- Minimax optimal approach (Hansen and Sargent style):
  - Additional constraint: Σ∞
t=0 v
t
⊤ v
t
≤ δ.
  - Equivalent minimax problem includes a penalty − θ v
t
⊤ v
t
; threshold for θ is around 62; simulations use θ = 70.
  - Minimax FRF:
    - p
푡푡+1
= −0.73 u
푡푡
� − 1.09 (p
푡푡
− p̅) + 0.29 (b
푡푡
− b̅)
  - This rule prescribes a stronger response to deviations from targets.
  - Expenditure growth under minimax: average about 6.3 percent in the first 5 years and 5.3 percent in the first 10 years.
- Robust stabilization (invariant set / feedback rule approach):
  - Targets a feedback rule and an invariant region around equilibrium that keeps the system within bounds for bounded shocks (ellipsoidal disturbance sets).
  - Example disturbance bounds implied: persistent shocks or additive uncertainties ~0.75 ppts in unemployment equation and ~0.5 ppts in the debt equation in absolute value (80 percent confidence ellipse illustration).
  - Robust FRF under parametric uncertainty (two-model representation, tax ratio drop from 43 percent to about 38 percent) yields:
    - p
푡푡+1
= −0.53 u
푡푡
� − 0.89 (p
푡푡
− p̅) + 0.28 (b
푡푡
− b̅)
  - Robust stabilizing rule implies somewhat lower expenditure growth in the first year but average paths similar to optimal and minimax:
    - average 6.4 percent in the first 5 years;
    - average 5.4 percent in the first 10 years.

### Key statistics and numeric summaries (preserved exactly as in source)
- Unemployment gap recent range: -0.3 to -0.6 percentage points.
- Okun’s law coefficient used: -0.3.
- Implied positive output gaps: between 1 and 2 percent.
- Autoregressive coefficient ρ: 0.72.
- Fiscal multiplier α (baseline): 0.2.
- Alternative multiplier considered: 0.4.
- Nominal potential growth: 4 ¼ percent.
- Nominal interest rate: 2 ½ percent.
- Real potential growth: 2 ½ percent.
- Inflation: 2 percent.
- Equilibrium real interest rate: 0.5 percent.
- Tax to GNI ratio (τ) baseline: 43 percent; alternative scenario drop to about 38 percent.
- Objective weights: unemployment gap = 1; fiscal impulse weight = 10; debt anchor weight baseline = 0.5 (alternatives 1 and 0.1).
- Baseline FRF coefficients: −0.43, −0.57, 0.15.
- Minimax FRF coefficients: −0.73, −1.09, 0.29.
- Robust two-model FRF coefficients: −0.53, −0.89, 0.28.
- Expenditure growth (baseline): 6.2 percent (first 5 years average), 5.4 percent (first 10 years average), converging to 4 ¼ percent.
- Expenditure growth (minimax): 6.3 percent (first 5 years average), 5.3 percent (first 10 years average).
- Expenditure growth (robust stabilizing): 6.4 percent (first 5 years average), 5.4 percent (first 10 years average).
- Minimax robustness parameter used in simulations: θ = 70 (threshold around 62).

*Source: CSO, IMF staff calculations — Appendix I, Table 3 and accompanying text, Options for Strengthening Ireland’s Fiscal Framework (Working Paper No. WP/2025/149).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025149-source-pdf.pdf_