## 1. Introduction

## Source details

**Canonical URL:** [1. Introduction](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022164-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022164-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022164-print-pdf.pdf.json)

---

### Key historical fiscal developments and policy question
- Since the asset price bubble burst in the early 1990s, the Japanese government have run fiscal deficits throughout business cycles, with repeated episodes of fiscal consolidation being interrupted or delayed by shocks.
- Public debt to GDP ratio:
  - about 60 percent of GDP in 1990
  - about 240 percent of GDP in 2019
  - about 260 percent of GDP after COVID-19 fiscal packages and the recession
- The COVID-19 pandemic and the war in Ukraine pose significant uncertainty for the outlook.
- Central policy question: When and how should Japan unwind stimulus and switch towards fiscal consolidation, balancing output stabilization and debt sustainability?

### Aim and approach of the paper
- The paper assesses the appropriateness of Japan’s past and future fiscal stance using a normative stochastic structural model: the Buffer-Stock Model of the Government (Fournier, 2019a).
- The model provides a utility-maximizing fiscal stance path that balances:
  - cycle stabilization (output gap management and hysteresis avoidance)
  - debt sustainability (market-access risk and risk premia)
- Relationship to literature:
  - Builds on normative fiscal-policy models (Lucas and Stokey, 1983; Chari, Christiano and Kehoe, 1994; Aiyagari et al., 2002)
  - Related applied fiscal-stance work for advanced economies (Kanda, 2011; Carnot, 2014; Bankowsky and Ferdinandusse, 2017)
  - Closest to Yoshino et al. (2015, 2018) but differs by explicitly modeling stochastic output shocks, distinguishing structural primary balance from automatic stabilizers, accounting for hysteresis, and providing backward-looking assessment within the same framework.

### Model conceptual principles and main features
- Government as forward-looking agent holding a debt buffer analogous to precautionary savings in a consumer buffer-stock model.
- Policy implications generated by the model:
  - Governments should smooth the cycle. Counter-cyclical fiscal policy dampens recessions, reduces hysteresis, and improves short-term utility.
  - Governments need to generate surpluses to restore buffers when public debt is high; not needed when debt is low.
  - Highly indebted governments should react less to adverse shocks because the insurance value of the debt buffer rises as the buffer shrinks.
- Model structure — key components:
  - Government chooses change in structural primary balance (st_t pb) to maximize household utility subject to:
    - debt sustainability constraints with a probability of losing market access rising with debt and debt changes
    - output dynamics with exogenous shocks and hysteresis (scarring) effects
    - feedbacks: primary balance affects output via fiscal multiplier; output affects primary balance via automatic stabilizers
    - adjustment costs to fiscal policy changes captured in the resource constraint
  - Per-period utility:
    - u(c_t, L_t) = c_t^(1−σ)/(1−σ) − ξ y_t^*^(1−σ) L_t^(1+η)/(1+η)
    - where σ is risk aversion, η is inverse Frisch elasticity, ξ calibrated so utility peaks at output equal to potential
  - Debt market pressure:
    - Interest rate increases with public debt and with the change in debt (risk premium components).
    - Probability of losing market access:
      - P(lma) = [1 + exp(d1 (1 − d_t / d − d2 (d_t − d_t−1)) )]^(−1)
      - d1 governs debt limit uncertainty; d2 governs effect of debt change; d is debt level where P(lma)=50% absent debt change.
    - Loss of market access forces debt to be constant under an adverse shock of d3 σ.
  - Output and hysteresis:
    - Output produced by linear labor production function; productivity affected by permanent hysteresis of crisis.
    - Fiscal multiplier:
      - ∂gap(pb_t, v_t)/∂pb_t = −m1 (1 − m2 gap(pb_t, v_t))
      - m1 is baseline multiplier; m2 magnifies multiplier in downturns.
  - Primary balance decomposition:
    - pb_t = pb_t^st + a · gap_t
    - a is automatic stabilizer (semi-elasticity of primary balance to the gap)
  - Aggregate resource constraint:
    - c_t = y_t (1 − χ (Δpb_t^st)^2)
    - χ captures fiscal adjustment costs (implementation and tax uncertainty costs).

### Calibration to Japan — baseline parameters
- Welfare function:
  - Discount factor β 0.99
  - Risk aversion σ 2
  - Labor elasticity η 1/0.3
  - Weight of labor ξ 1
- Fiscal parameters:
  - Fiscal multiplier when the gap is null m1 0.65
  - Fiscal multiplier sensitivity to shocks m2 5
  - Automatic stabilizers (primary balance semi-elasticity to the gap) a 0.41
  - Adjustment cost χ 3
- Interest rate and debt parameters:
  - Growth-adjusted interest rate at current debt level 0%
  - Effect of debt level on the risk premium α 1.0%
  - Effect of debt change on the risk premium α2 0.5%
  - Debt level at which the risk to lose market access is 50% d 300%
  - Debt limit accuracy d1 3
  - Effect of debt change on the risk to lose market access is d2 1
  - Adverse scenario coefficient in case of loss of market access d3 -1%
- Economy parameters:
  - Potential GDP per capita growth 0.9%
  - Shock persistence ρ 0.60
  - Shock size σ 3.8%
  - Hysteresis 10%
  - Hysteresis threshold -1%

### Calibration details and empirical support
- Effect of debt level on the risk premium (α):
  - Estimated via a regression following Tokuoka (2010) with additional control for U.S. long-term interest rate.
  - Result: an increase in gross-debt-to-GDP ratio by one percentage point is associated with a rise in JGB yields by one basis point (0.01), consistent with other estimates.
- Fiscal multiplier (m1) rationale:
  - Japan classified into “Medium-multiplier” group (multiplier of 0.4~0.6) per Batini et al. (2014); both bounds increased by 30 percent due to the Bank of Japan’s policy rate being close to the effective lower bound, yielding a range of 0.5~0.8.
  - Baseline uses the mid-point value 0.65.
- Growth-adjusted interest rate (r−g):
  - Historical r−g differentials between 1990 and 2020: negative in many years in the 2010s, positive for most of the 1990s, affecting debt dynamics.

### Regression evidence (selected results)
- Regression summary:
  - R square: 0.93
  - Observations: 80 (quarterly data from Q4 2000 to Q3 2020)
  - Dependent variable: nominal 10 year bond yield
  - Specification: nominal 10 year bond yield = 훽0 + 훽1 gross debt + control variables
- Key coefficients (Table 2):
  - Gross debt 0.01 **
  - BOJ's holdings of JGBs -0.01 ***
  - Private sector financial assets -0.01
  - Share of foreign JGB holdings -0.05 **
  - (Notes) 1. *** p < 0.01, ** p < 0.05, * p < 0.1.
- Data sources: MOF and Flow of Funds series for debt, BoJ holdings, foreign holdings, and private net financial assets; additional controls include short-term interest rate, inflation, real GDP growth, and the US long-term interest rate.

### Historical fiscal context and quantified contributions to debt accumulation (1990–2020)
- Increase in debt to GDP ratio: 196.0 (In percent of GDP)
- Cumulative contributions to changes in gross public debt between 1990 and 2020 (In percent of GDP):
  - Primary deficits: 123.7
  - Pension: 95.6
  - Healthcare and long-term care: 82.2
  - Other primary expenditures: 20.3
  - Residual: 9.9
  - r-g (interest rate-growth differential): 62.4
  - Primary revenues: -46.2
  - Other social benefits: -7.2
  - Public investments: -21.0

### Fiscal developments during the COVID-19 pandemic
- Supplementary budgets in 2020 and 2021:
  - Four large supplementary budgets amounting to about 18 ½ percent of GDP
- Primary deficits:
  - 2019: 2½ percent of GDP
  - 2020: 8 percent of GDP
  - 2021: 7 percent of GDP

### Model-based policy findings and recommendations
- Near term:
  - The model recommends a wider near-term structural primary deficit than current policy plans to narrow the output gap and minimize hysteresis.
  - This supports disbursement of contingent measures included in the November 2021 and April 2022 economic packages in response to shocks (Omicron wave and the war in Ukraine).
- Medium term:
  - After 2024 the model calls for gradual fiscal adjustment to anchor public debt on a downward path and rebuild fiscal buffers.
  - Pace of consolidation is sensitive to the interest rate-growth differential:
    - With r-g = -0.3 the model suggests a more gradual consolidation; public debt can be reduced even with a negative structural primary balance.
    - With r-g = +0.3 the model requires a higher structural primary balance and faster debt reduction.
- Policy design implications:
  - Fiscal support should be sized flexibly according to epidemiological, economic, and geopolitical developments.
  - Automatic stabilizers are emphasized because they react quickly and accurately to economic developments.
  - Targeted support to households and firms is important given higher multipliers relative to untargeted support.

### Sensitivity and scenario analysis (selected quantitative details)
- Sensitivity checks for recommended structural primary balance in 2022:
  - Lower (higher) multiplier scenario: multiplier (m1) is 0.5 (0.8)
  - Lower adjustment cost scenario: adjustment cost set at 1.5
  - Larger hysteresis scenario: hysteresis is 100% larger than in the baseline
  - Higher elasticity of interest to debt scenario: alpha is set at 1.5
- Key sensitivity observations:
  - Without scarring (no hysteresis) the model recommends less near-term stimulus.
  - If the fiscal multiplier is lower, the model calls for a wider fiscal deficit.
  - Sensitivities to the elasticity of interest rate to debt and to the adjustment cost are small.
- Scenario narratives:
  - Upside scenario: quicker and stronger recovery driven by pent-up demand and large household savings; temporary support smaller and phased out sooner; frontloaded fiscal consolidation.
  - Downside scenario: recovery hindered by pandemic or escalation of war in Ukraine; more stimulus advisable even if it increases public debt, followed by larger fiscal adjustment later.

### Conclusions — historical assessment and policy trade-offs
- Over the past three decades, chronic primary deficits, rising age-related spending, and large shocks led to unprecedented public debt accumulation; pension and healthcare and long-term care made the largest contributions.
- Retrospective model-based analysis suggests the fiscal stance in the 1990s and the early 2000s was overall looser than the model recommendations.
- Policy trade-off:
  - Premature fiscal support withdrawal would risk derailing the recovery.
  - Large fiscal imbalances point to a need for fiscal consolidation.
- Model recommendations:
  - Near term: fiscal policy should be supportive to narrow the output gap and minimize hysteresis.
  - Medium term: gradual fiscal consolidation to stabilize debt and rebuild fiscal buffers against demographic pressures.
- Sensitivity:
  - Recommended pace of adjustment is sensitive to r-minus-g assumptions; model scenarios include r-g = -0.3% and r-g = 0.3%.

### Methodological note on debt decomposition and further research avenues
- Debt decomposition follows the IMF Debt Sustainability Framework for Market Access Countries and attributes changes in debt since 1990 to primary balance, interest rate-growth differential, and a residual including asset-side changes and interest revenues.
- Allocation of the primary balance contribution across spending and revenue categories uses 1990 category shares adjusted so the 1990 primary balance becomes zero; no exchange-rate contribution as Japan’s public debt is all denominated in yen.
- Suggested further research:
  - Distinguish between fiscal instruments: revenue versus expenditure.
  - Study explicit interaction between fiscal and monetary policy.
  - Explore fiscal space in a low interest rate environment, the link between price stability and fiscal credibility, and fiscal consequences under interest-rate uncertainty.
  - Examine optimal design of automatic stabilizers in addition to discretionary fiscal responses.

*Source: IMF Working Paper — “The Fiscal Stance in Japan” (extracted from wpiea2022164-print-pdf).*

### 1. Introduction ........................................................................................................

### 1. Introduction

### Key historical fiscal developments and policy question
- Since the asset price bubble burst in the early 1990s, the Japanese government have run fiscal deficits throughout business cycles, with repeated episodes of fiscal consolidation being interrupted or delayed by shocks.
- Public debt to GDP ratio:
  - about 60 percent of GDP in 1990
  - about 240 percent of GDP in 2019
  - about 260 percent of GDP after COVID-19 fiscal packages and the recession
- The COVID-19 pandemic and the war in Ukraine pose significant uncertainty for the outlook.
- Central policy question: When and how should Japan unwind stimulus and switch towards fiscal consolidation, balancing output stabilization and debt sustainability?

### Aim and approach of the paper
- The paper assesses the appropriateness of Japan’s past and future fiscal stance using a normative stochastic structural model: the Buffer-Stock Model of the Government (Fournier, 2019a).
- The model provides a utility-maximizing fiscal stance path that balances:
  - cycle stabilization (output gap management and hysteresis avoidance)
  - debt sustainability (market-access risk and risk premia)
- Comparison with related literature:
  - Builds on normative fiscal-policy models (Lucas and Stokey, 1983; Chari, Christiano and Kehoe, 1994; Aiyagari et al., 2002)
  - Related applied fiscal-stance work for advanced economies (Kanda, 2011; Carnot, 2014; Bankowsky and Ferdinandusse, 2017)
  - Closest to Yoshino et al. (2015, 2018) but differs by explicitly modeling stochastic output shocks, distinguishing structural primary balance from automatic stabilizers, accounting for hysteresis, and providing backward-looking assessment within the same framework.

### Paper structure
- Section 2: Model economy and calibration to Japan.
- Section 3: Japan’s fiscal policy since 1990 and retrospective evaluation via the model.
- Section 4: Model-recommended optimal future fiscal path and scenario analyses.
- Section 5: Conclusions.

### The Buffer-Stock Model of the Government — Conceptual principles
- Analogy: government as a forward-looking agent holding a debt buffer similar to a consumer’s precautionary savings in the buffer-stock model (Deaton, 1991).
- Key policy principles generated by the model:
  - Governments should smooth the cycle. Counter-cyclical fiscal policy dampens recessions, reduces hysteresis, and improves short-term utility.
  - Governments need to react to rising debt by generating surpluses to restore buffers when public debt is high; not needed when debt is low.
  - Highly indebted governments should react less to adverse shocks because the insurance value of the debt buffer rises as the buffer shrinks, incentivizing preservation of buffers.

### Model structure — main features
- Government chooses change in structural primary balance (st_t pb) to maximize household utility subject to:
  - debt sustainability constraints with a probability of losing market access rising with debt and debt changes
  - output dynamics with exogenous shocks and hysteresis (scarring) effects
  - feedbacks: primary balance affects output via fiscal multiplier; output affects primary balance via automatic stabilizers
  - adjustment costs to fiscal policy changes captured in the resource constraint
- Per-period utility:
  - u(c_t, L_t) = c_t^(1−σ)/(1−σ) − ξ y_t^*^(1−σ) L_t^(1+η)/(1+η)
  - where σ is risk aversion, η is inverse Frisch elasticity, ξ calibrated so utility peaks at output equal to potential
- Debt market pressure:
  - Interest rate increases with public debt and with the change in debt (risk premium components).
  - Probability of losing market access:
    - P(lma) = [1 + exp(d1 (1 − d_t / d − d2 (d_t − d_t−1)) )]^(−1)
    - d1 governs debt limit uncertainty; d2 governs effect of debt change; d is debt level where P(lma)=50% absent debt change.
  - Loss of market access forces debt to be constant under an adverse shock of d3 σ.
- Output and hysteresis:
  - Output produced by linear labor production function; productivity affected by permanent hysteresis of crisis.
  - Fiscal multiplier:
    - ∂gap(pb_t, v_t)/∂pb_t = −m1 (1 − m2 gap(pb_t, v_t))
    - m1 is baseline multiplier; m2 magnifies multiplier in downturns.
- Primary balance decomposition:
  - pb_t = pb_t^st + a · gap_t
  - a is automatic stabilizer (semi-elasticity of primary balance to the gap)
- Aggregate resource constraint:
  - c_t = y_t (1 − χ (Δpb_t^st)^2)
  - χ captures fiscal adjustment costs (implementation and tax uncertainty costs).

### Calibration to Japan — baseline parameters (Table 1)
- Welfare function:
  - Discount factor β 0.99
  - Risk aversion σ 2
  - Labor elasticity η 1/0.3
  - Weight of labor ξ 1
- Fiscal parameters:
  - Fiscal multiplier when the gap is null m1 0.65
  - Fiscal multiplier sensitivity to shocks m2 5
  - Automatic stabilizers (primary balance semi-elasticity to the gap) a 0.41
  - Adjustment cost χ 3
- Interest rate and debt parameters:
  - Growth-adjusted interest rate at current debt level 0%
  - Effect of debt level on the risk premium α 1.0%
  - Effect of debt change on the risk premium α2 0.5%
  - Debt level at which the risk to lose market access is 50% d 300%
  - Debt limit accuracy d1 3
  - Effect of debt change on the risk to lose market access is d2 1
  - Adverse scenario coefficient in case of loss of market access d3 -1%
- Economy parameters:
  - Potential GDP per capita growth 0.9%
  - Shock persistence ρ 0.60
  - Shock size σ 3.8%
  - Hysteresis 10%
  - Hysteresis threshold -1%

### Calibration details and empirical support
- Effect of debt level on the risk premium (α):
  - Estimated via a regression following Tokuoka (2010) with additional control for U.S. long-term interest rate.
  - Result: an increase in gross-debt-to-GDP ratio by one percentage point is associated with a rise in JGB yields by one basis point (0.01), consistent with other estimates (Mochida 0.7 bps; comparisons to France and US calibrations).
- Fiscal multiplier (m1) rationale:
  - Classified Japan into “Medium-multiplier” group (multiplier of 0.4~0.6) per Batini et al. (2014) bucket approach.
  - Both bounds increased by 30 percent due to the Bank of Japan’s policy rate being close to the effective lower bound for a long time, yielding a range of 0.5~0.8.
  - Baseline uses the mid-point value 0.65.
- Growth-adjusted interest rate (r−g):
  - Figure 1 (referenced) shows r−g differentials between 1990 and 2020: negative in many years in the 2010s, positive for most of the 1990s, impacting debt dynamics.

### Regression results (Table 2 — explanatory variables and coefficients)
- Gross debt 0.01 **
- BOJ's holdings of JGBs -0.01 ***
- Private sector financial assets -0.01
- Share of foreign JGB holdings -0.05 **
- (Notes) 1. *** p < 0.01, ** p < 0.05, * p < 0.1.

*IMF Working Paper — “The Fiscal Stance in Japan”: 1. Introduction (extracted from wpiea2022164-print-pdf)*

### 2. R square: 0.93, Observations: 80 (quarterly data from Q4 2000 to Q3 2020)

### 2. R square: 0.93, Observations: 80 (quarterly data from Q4 2000 to Q3 2020)

### Data, regression, and model setup
- Sample and fit:
  - R square: 0.93
  - Observations: 80 (quarterly data from Q4 2000 to Q3 2020)
- Dependent variable:
  - nominal 10 year bond yield
- Regression specification (as presented):
  - nominal 10 year bond yield = 훽0 + 훽1 gross debt + control variables
- Data sources:
  - 10-Year Benchmark Government Bond Yield (EOP, % p.a.): MOF
  - General Government Gross Debt in percent of GDP (excluding the bonds issued to finance the Fiscal Investment and Loan Program): Flow of Funds (FoF)
  - Central government debt held by the BoJ in percent of GDP: FoF
  - Share of foreign holdings of JGBs: FoF
  - Private net financial assets held by the household and corporate sectors in percent of GDP: FoF
- Other control variables:
  - short-term interest rate, inflation, real GDP growth rate, and the US long-term interest rate

### Key model assumptions and long-run calibrations
- Long-run baseline assumption for the interest rate minus growth differential:
  - r-minus-g long-run value assumed as zero in the baseline
  - Sensitivity scenarios: +0.3 and -0.3 for r-minus-g
- Automatic stabilizer (semi-elasticity of the primary balance to the output gap):
  - a = 0.4 (per Price et al., 2015 estimate for Japan)
- WEO April 2022 database used for projections; beyond 2027 key variables converge gradually over 2028~2030 to long-run values:
  - potential growth rate: 0.3 percent
  - inflation: 1 percent
  - 10-year JGB yield: 1.3 percent
  - Note: 10-year JGB yield assumption deviates from the DSA assumption (2.0 percent) to match the baseline r-minus-g assumption

### Historical fiscal context and quantified contributions to debt accumulation (1990–2020)
- Increase in debt to GDP ratio: 196.0 (In percent of GDP)
- Cumulative contributions to changes in gross public debt between 1990 and 2020 (In percent of GDP):
  - Primary deficits: 123.7
  - Pension: 95.6
  - Healthcare and long-term care: 82.2
  - Other primary expenditures: 20.3
  - Residual: 9.9
  - r-g (interest rate-growth differential): 62.4
  - Primary revenues: -46.2
  - Other social benefits: -7.2
  - Public investments: -21.0

### Fiscal developments during the COVID-19 pandemic
- Supplementary budgets in 2020 and 2021:
  - Four large supplementary budgets amounting to about 18 ½ percent of GDP
- Primary deficits:
  - 2019: 2½ percent of GDP
  - 2020: 8 percent of GDP
  - 2021: 7 percent of GDP

### Model-based policy findings and recommendations
- Near term:
  - The model recommends a wider near-term structural primary deficit than current policy plans to narrow the output gap and minimize hysteresis.
  - This can be interpreted as supporting disbursement of contingent measures included in the November 2021 and April 2022 economic packages in response to shocks (Omicron wave and the war in Ukraine).
- Medium term:
  - After 2024 the model calls for gradual fiscal adjustment to anchor public debt on a downward path and rebuild fiscal buffers.
  - The pace of consolidation is sensitive to the interest rate-growth differential:
    - With r-g = -0.3 the model suggests a more gradual consolidation; public debt can be reduced even with a negative structural primary balance.
    - With r-g = +0.3 the model requires a higher structural primary balance and faster debt reduction.
- Policy design implications:
  - Fiscal support should be sized flexibly according to epidemiological, economic, and geopolitical developments.
  - Automatic stabilizers are emphasized as useful because they react quickly and accurately to economic developments.
  - Targeted support to households and firms is important given higher multipliers relative to untargeted support.

### Sensitivity and scenario analysis (selected quantitative details)
- Sensitivity checks reported for recommended structural primary balance in 2022:
  - Baseline vs alternatives note:
    - Lower (higher) multiplier scenario: multiplier (m1) is 0.5 (0.8)
    - Lower adjustment cost scenario: adjustment cost set at 1.5
    - Larger hysteresis scenario: hysteresis is 100% larger than in the baseline
    - Higher elasticity of interest to debt scenario: alpha is set at 1.5
  - Key sensitivity observations:
    - Without scarring (no hysteresis) the model recommends less near-term stimulus.
    - If the fiscal multiplier is lower, the model calls for a wider fiscal deficit (highlighting importance of targeting).
    - Sensitivities to the elasticity of interest rate to debt and to the adjustment cost are small.
- Scenario analyses for near-term fiscal policy (qualitative):
  - Upside scenario: quicker and stronger recovery driven by pent-up demand and large household savings; temporary support smaller and phased out sooner; frontloaded fiscal consolidation.
  - Downside scenario: recovery hindered further by pandemic or escalation of war in Ukraine; more stimulus advisable even if it increases public debt, followed by larger fiscal adjustment later.
- Recovery-path implication:
  - Scale and timing of additional fiscal support should be flexibly attuned to epidemiological, economic, and geopolitical developments.

*Italic: Source: IMF Working Paper — “The Fiscal Stance in Japan” (figures, tables, and text extracted from the chapter).*

### 5. Conclusions

### 5. Conclusions

### Historical assessment and drivers of debt accumulation
- Over the past three decades, the Japanese government ran chronic primary deficits, driven by repeated stimuli after the bubble burst, a rise in age-related spending, and large shocks that halted the momentum for fiscal consolidation.
- Public debt piled up to an unprecedented level, to which the pension and healthcare and long-term care expenditures made the largest contribution.
- Retrospective model-based analysis suggests the fiscal stance in the 1990s and the early 2000s was overall looser than the model recommendations.

### Near-term versus medium-term policy advice from the model
- Policy trade-off:
  - Premature fiscal support withdrawal would risk derailing the recovery.
  - Large fiscal imbalances point to a need for fiscal consolidation.
- Model recommendations:
  - Near term: fiscal policy should be supportive with a view to narrowing the output gap and minimizing hysteresis.
  - Medium term: fiscal consolidation at a gradual pace to stabilize debt and rebuild fiscal buffers in the face of intensifying demographic pressures.
- Sensitivity:
  - The recommended pace of adjustment is sensitive to the r-minus-g assumptions, reaffirming the critical role the interest rate-growth differential plays in debt dynamics for highly indebted countries like Japan.
  - Model scenarios shown include r-g=-0.3% and r-g=0.3%.

### Methodological note on debt decomposition (appendix summary)
- The decomposition follows the IMF Debt Sustainability Framework for Market Access Countries and attributes changes in debt since 1990 to:
  - Primary balance
  - Interest rate-growth differential (automatic debt dynamics15)
  - Residual, which includes asset-side changes and interest revenues
- Allocation of the primary balance contribution across spending and revenue categories:
  - Uses each category’s 1990 level (in percent of GDP) adjusted for the primary surplus in 1990 (+3.2 percent of GDP) as a benchmark.
  - Adjustment is done by revising up each expenditure category and revising down the revenue category on a pro-rata basis so the primary balance for 1990 after adjustment becomes zero.
  - Example provided: non-interest revenue actual in 1990 was 29.1 percent; adjusted benchmark becomes 27.4 (=29.1-3.2/2). The difference between actual non-interest revenue in a given year (e.g., 28.6 in 1991) and 27.4 is counted as the contribution of primary revenue to changes in debt (i.e., -1.2 ppt contribution in 1991).
- There is no exchange-rate contribution as Japan’s public debt is all denominated in yen.

### Gaps and avenues for further research and policy analysis
- Extend analysis to distinguish between fiscal instruments: revenue versus expenditure.
- Study explicit interaction between fiscal and monetary policy (monetary policy was implicit in this paper, including a higher multiplier reflecting the zero lower bound context).
- Specific topics of interest:
  - Fiscal space in a low interest rate environment.
  - The link between price stability and fiscal credibility.
  - Fiscal consequences when there is uncertainty in interest rate movement (such as when policymakers face a situation with inflation and employment moving in different directions).
- Examine optimal design of automatic stabilizers in addition to discretionary fiscal responses.

*Source: IMF staff calculations and text from "5. Conclusions" (wpiea2022164-print-pdf).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022164-print-pdf.pdf_