## 1. Actual and Derived Debt-to-GDP (U.S.)

## Source details

**Canonical URL:** [1. Actual and Derived Debt-to-GDP (U.S.)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15137.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15137.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15137.pdf.json)

---

### Introduction and overview
- Public debt in advanced economies rose from around 70 percent of GDP in 2007 to over 105 percent in 2013 (IMF, 2014).
- The paper studies how primary balance, interest rate, growth, and inflation interact to determine sovereign debt dynamics using a structural vector auto-regression (SVAR) framework.
- Core debt accumulation relationship (Equation 1):
  - d_t = (1 + i_t^eff)/(1 + g_t)(1 + π_t) * d_{t-1} - pb_t
  - where d_t is sovereign debt-to-GDP ratio; i_t^eff is the effective interest rate; g_t is real growth; π_t is the GDP deflator; pb_t is the primary balance to GDP ratio.

### Methodology and identification
- Model components:
  - SVAR estimated for endogenous variables corresponding to components of the debt equation.
  - The marginal interest rate on new borrowing (not the monetary policy rate) is used in the VAR; the effective interest rate used in the debt equation is derived from the marginal rate and average maturity.
- Identification and auxiliary structures:
  - Structural identification transforms reduced-form errors into structural shocks via an A matrix with imposed restrictions.
  - Elasticities (θ terms) for primary balance responses to growth and inflation are taken from Girouard and Andre (2005).
  - The contemporaneous elasticity of the primary balance with respect to the marginal interest rate is set to zero.
  - Relationship between marginal and effective interest rates (Equation 4): i_t^eff = (1/n_t) * Σ_{j=1}^{n_t} i_{t-j}^marg where n_t is average maturity.
- Impulse response functions (IRFs):
  - Debt is treated endogenously in IRF construction by combining the SVAR with the debt accumulation equation.
  - IRFs computed following Cherif and Hasanov (2012) using a bootstrapping method (2000 repetitions) to obtain median IRFs and 16% and 68% quantiles.
- Data:
  - Quarterly data for 15 OECD countries (Australia, Canada, Belgium, Denmark, France, Finland, Germany, Greece, Ireland, Italy, Norway, Spain, Sweden, the United Kingdom and the United States), mostly covering 1999-2014; some countries have longer series.

### Key empirical findings (median responses across all countries)
- The paper reports 375 IRFs (5 responses to each of 5 shocks across 15 countries); Table 2 summarizes median peak responses (peak response to each shock).
- Table 2 (median peak responses):
  - Shock: Primary Balance (positive shock)
    - Primary Balance: 1.00
    - Marginal Interest Rate: 0.04
    - Debt to GDP ratio: -1.25
  - Shock: Real Growth (negative shock)
    - Primary Balance: -0.60
    - Marginal Interest Rate: 0.01
    - Debt to GDP ratio: 1.60
  - Shock: Marginal Interest Rate (positive shock)
    - Primary Balance: 0.7
    - Marginal Interest Rate: 1.00
    - Debt to GDP ratio: 0.45
  - Shock: Debt to GDP ratio (positive shock)
    - Primary Balance: 0.21
    - Marginal Interest Rate: -0.04
    - Debt to GDP ratio: 1.00
- Interpretation of key magnitudes:
  - A positive 1 unit (peak) shock to the primary balance has a median associated decline in debt-to-GDP of -1.25 (peak).
  - A 1 percent fall in growth has a peak association with a 0.6 percent loosening in the structural primary balance (primary balance response = -0.60).
  - A 1 percent increase in the marginal interest rate leads to an average tightening of the primary balance of around 0.7 (peak).
  - A positive shock to debt elicits a government reaction in the primary balance (0.21 peak) and a small negative median marginal interest rate response (-0.04 peak) in the pseudo-IRF construction.

### Qualitative dynamics and implications
- Primary balance behavior:
  - The primary balance responds actively to growth shocks and is highly sensitive to changes in the marginal interest rate.
  - The response to growth shocks is counter-cyclical (despite automatic stabilizers being controlled for), indicating discretionary fiscal policy: a 1 percent fall in growth → peak primary balance response of -0.60.
- Interest rate and debt interactions:
  - Marginal interest rate shocks raise debt (median peak debt response = 0.45) and prompt an average tightening in the primary balance (0.7 peak), consistent with a fiscal reaction to higher borrowing costs.
  - Positive shocks to debt generate government fiscal responses (primary balance peak = 0.21) in the pseudo-IRF framework.
- Nonlinearities and feedbacks:
  - Combining the SVAR with the debt accumulation equation introduces nonlinear feedbacks between debt and its components; these are accounted for in IRF construction and imply IRFs that may not be smooth.
- Model performance:
  - Figure 1 and Figure 2 (U.S. example) show that the debt accumulation equation and the marginal-to-effective interest rate auxiliary equation provide close approximations to actual series for the United States (actual vs fitted comparisons).

### Key empirical findings on fiscal responses and debt dynamics
- Primary balance responds positively to debt: "a 1pt increase in debt, on average generates a 0.2pt tightening in the primary balance."
- Governments in the sample actively react to shocks that impact debt dynamics; responses are designed to stabilize debt and mitigate second round effects.
- Estimated market reaction to shocks in the primary balance, growth and debt are close to zero on the sample average, but notable cross-country differences exist.

### Sample split and rationale (constrained vs unconstrained monetary policy)
- Countries split into two groups:
  - "Constrained" monetary policy: sovereigns without full control over monetary policy (currency union or fixed exchange rate) — Belgium, Denmark, France, Finland, Germany, Greece, Ireland, Italy and Spain.
  - "Unconstrained" monetary policy: sovereigns with unconstrained monetary policy (typically ‘inflation targeters’) — the Australia, Canada, Norway, Sweden, the United Kingdom and the United States.
- Reasons why independent monetary policy may affect debt dynamics:
  - Better control of nominal GDP to stabilize debt-to-GDP without fiscal consolidation.
  - Use of seigniorage revenues to help repay debt (potentially higher inflation).
  - Large purchases of government debt can coordinate creditors and reduce likelihood of runs/multiple equilibria.
- The estimation approach cannot precisely determine causality between monetary policy and interest rate responses.

### Responses by monetary regime (summary)
- Response of interest rates:
  - Marginal interest rate response to a +1pt primary balance shock: small increase, peaking at around 0.08pts, similar for both groups.
  - Marginal interest rate response to a -1pt growth shock: essentially insignificant for both groups.
  - Response to a +10pt debt shock (pseudo IRFs; no confidence bands calculated):
    - Constrained group: marginal interest rate increases sharply and persists for around 1.5 years.
    - Unconstrained group: marginal interest rate persistently falls.
- Response of the primary balance:
  - Shock to marginal interest rate (+1pt):
    - Constrained group: large and persistent increase in primary balance.
    - Unconstrained group: virtually no reaction.
  - Shock to growth (-1pt):
    - Constrained group: persistently tighter primary balance to stabilize debt.
    - Unconstrained group: structural primary balance does not react; monetary policy can offset demand shocks.
  - Shock to debt (+10pt):
    - Both groups tighten fiscal policy, but the constrained group’s response is much larger.
    - Confirms Bohn-type behavior: primary balance reacts proportionally to the debt level, with stronger tightening where monetary policy is constrained.

### Evolution of debt following shocks (group differences)
- Growth shock (-1pt):
  - Median debt levels increase by around 2–2.5 percent of GDP at the peak for both groups.
  - Constrained group: slightly larger peak impact due to larger increases in marginal interest rates, but faster return to pre-shock level via stronger primary balance response.
  - Unconstrained group: slower, more measured reduction in debt.
- Interest rate shock (+1pt):
  - Unconstrained group: impact on debt quickly dies out.
  - Constrained group: impact on debt is highly persistent.
- Primary balance shock (negative 1pts):
  - Unconstrained group: peak impact on debt is 2 pts and is highly persistent; debt is maintained at higher level for a long period.
  - Constrained group: reduce debt at a much quicker rate.
- Debt shock (+10pt):
  - Both groups show debt returning to pre-shock level after around 5 years; underlying mechanisms differ (unconstrained rely more on stimulating growth, constrained on fiscal consolidation).

### Diagnostics and robustness
- Stationarity:
  - ADF and KPSS tests used. Four countries—Japan, Portugal, Austria and the Netherlands—had at least one non-stationary series and were dropped from the sample.
- Residuals:
  - Jarque-Bera test for normality produced mixed results; reliance on assumption that residuals are asymptotically normally distributed.
- Robustness checks:
  - Lag length varied (e.g., 2 vs 4 quarters): IRFs maintained same sign, similar magnitude and statistical significance.
  - Sample restricted to exclude crisis years (2008-2014): U.S. and U.K. show no significant change; other countries with shorter series show wider confidence bands and some shocks become insignificant.
  - SVAR identification varied (Cholesky decomposition vs imposing automatic stabilizer elasticities): results largely unchanged.
  - Removing outliers (e.g., Norway) does not materially change results.
- Sample period: 1999-2014 (characterized by adoption of the euro and subsequent euro area crisis episodes).

### Conclusion and policy implications
- The SVAR with endogenous debt accumulation captures interactions among primary balance, interest rate, growth and inflation, and shows feedback effects that can mitigate or amplify debt shocks.
- Sovereign credit markets:
  - Do not systematically respond to shocks to growth or the primary balance on the sample average.
  - Are sensitive to the debt level, especially for constrained monetary policy countries where market interest rates react positively to higher debt.
  - For unconstrained policy countries, higher debt may be associated with downward pressure on long-term rates.
- Policy implications:
  - Monetary policy matters for sovereign debt sustainability; unconstrained monetary regimes have more tools to manage debt dynamics (monetary offset, seigniorage, debt purchases).
  - Constrained monetary policy countries are more reliant on fiscal adjustment to stabilize debt.

### Suggested extensions
- Use a panel VAR to derive more systematic median IRFs.
- Extend analysis to emerging market economies, requiring addition of exchange rate term to control for foreign currency denominated debt.
- Consider Bayesian SVAR estimation or local projection models (Jordà, 2005) as alternative specifications.

*Source: _wp15137 - 1. Actual and Derived Debt-to-GDP (U.S.), IMF working paper (extracted content).*

### 1. Actual and Derived Debt-to-GDP (U.S.) ...............................................................................

### 1. Actual and Derived Debt-to-GDP (U.S.)

### Introduction and overview
- Recent years saw a rapid increase in public debt in advanced economies: from around 70 percent of GDP in 2007 to over 105 percent in 2013 (IMF, 2014).
- The paper examines how primary balance, interest rate, growth, and inflation interact to determine sovereign debt dynamics using a structural vector auto-regression (SVAR) framework.
- Core accounting relationship (debt accumulation equation, Equation 1):
  - d_t = (1 + i_t^eff)/(1 + g_t)(1 + π_t) * d_{t-1} - pb_t
  - where d_t is sovereign debt-to-GDP ratio; i_t^eff is the effective interest rate; g_t is real growth; π_t is the GDP deflator; pb_t is the primary balance to GDP ratio.

### Methodology and identification
- Model components:
  - SVAR estimated for endogenous variables corresponding to components of the debt equation.
  - The marginal interest rate on new borrowing (not the monetary policy rate) is used in the VAR; the effective interest rate used in the debt equation is derived from the marginal rate and average maturity.
- Identification and auxiliary structures:
  - Structural identification transforms reduced-form errors into structural shocks via an A matrix with imposed restrictions; elasticities (θ terms) for primary balance responses to growth and inflation are taken from Girouard and Andre (2005); the contemporaneous elasticity of the primary balance with respect to the marginal interest rate is set to zero.
  - Relationship between marginal and effective interest rates (Equation 4): i_t^eff = (1/n_t) * Σ_{j=1}^{n_t} i_{t-j}^marg where n_t is average maturity.
- Impulse response functions (IRFs):
  - Debt is treated endogenously in IRF construction by combining the SVAR with the debt accumulation equation; IRFs computed following Cherif and Hasanov (2012) using a bootstrapping method (2000 repetitions) to obtain median IRFs and 16% and 68% quantiles.
- Data:
  - Quarterly data for 15 OECD countries (Australia, Canada, Belgium, Denmark, France, Finland, Germany, Greece, Ireland, Italy, Norway, Spain, Sweden, the United Kingdom and the United States), mostly covering 1999-2014; some countries have longer series.

### Key empirical findings (median responses across all countries)
- The paper reports 375 IRFs (5 responses to each of 5 shocks across 15 countries); Table 2 summarizes median peak responses. Responses reflect the peak response to each shock.
- Table 2 (median peak responses):
  - Shock: Primary Balance (positive shock)
    - Primary Balance: 1.00
    - Marginal Interest Rate: 0.04
    - Debt to GDP ratio: -1.25
  - Shock: Real Growth (negative shock)
    - Primary Balance: -0.60
    - Marginal Interest Rate: 0.01
    - Debt to GDP ratio: 1.60
  - Shock: Marginal Interest Rate (positive shock)
    - Primary Balance: 0.7
    - Marginal Interest Rate: 1.00
    - Debt to GDP ratio: 0.45
  - Shock: Debt to GDP ratio (positive shock)
    - Primary Balance: 0.21
    - Marginal Interest Rate: -0.04
    - Debt to GDP ratio: 1.00
- Interpretation of key magnitudes:
  - A positive 1 unit (peak) shock to the primary balance has a median associated decline in debt-to-GDP of -1.25 (peak).
  - A 1 percent fall in growth has a peak association with a 0.6 percent loosening in the structural primary balance (i.e., primary balance response = -0.60).
  - A 1 percent increase in the marginal interest rate leads to an average tightening of the primary balance of around 0.7 (peak).
  - A positive shock to debt elicits a government reaction in the primary balance (0.21 peak) and a small negative median marginal interest rate response (-0.04 peak) in the pseudo-IRF construction.

### Qualitative dynamics and implications
- Primary balance behavior:
  - The primary balance responds actively to growth shocks and is highly sensitive to changes in the marginal interest rate.
  - The response to growth shocks is counter-cyclical (despite automatic stabilizers being controlled for), indicating discretionary fiscal policy: a 1 percent fall in growth → peak primary balance response of -0.60.
- Interest rate and debt interactions:
  - Marginal interest rate shocks raise debt (median peak debt response = 0.45) and prompt an average tightening in the primary balance (0.7 peak), consistent with a fiscal reaction to higher borrowing costs.
  - Positive shocks to debt generate government fiscal responses (primary balance peak = 0.21) in the pseudo-IRF framework.
- Nonlinearities and feedbacks:
  - Combining the SVAR with the debt accumulation equation introduces nonlinear feedbacks between debt and its components; these are accounted for in IRF construction and imply IRFs that may not be smooth.
- Model performance:
  - Figure 1 and Figure 2 (U.S. example) show that the debt accumulation equation and the marginal-to-effective interest rate auxiliary equation provide close approximations to actual series for the United States (actual vs fitted comparisons).

*Italic: Source: _wp15137 - 1. Actual and Derived Debt-to-GDP (U.S.), IMF working paper (extracted content).*

### 0.7 percent. This reaction may be explained by the governments’ desire to stabilize debt in

### _wp15137 - 0.7 percent. This reaction may be explained by the governments’ desire to stabilize debt in

### Key empirical findings on fiscal responses and debt dynamics
- Primary balance responds positively to debt: "a 1pt increase in debt, on average generates a 0.2pt tightening in the primary balance."
- Governments in the sample actively react to shocks that impact debt dynamics; responses are designed to stabilize debt and mitigate second round effects.
- Estimated market reaction to shocks in the primary balance, growth and debt are close to zero on the sample average, but notable cross-country differences exist.

### Sample split and rationale
- Countries are split into two groups:
  - "Constrained" monetary policy: sovereigns without full control over monetary policy (currency union or fixed exchange rate).
  - "Unconstrained" monetary policy: sovereigns with unconstrained monetary policy (typically ‘inflation targeters’).
- Reasons why independent monetary policy may affect debt dynamics:
  - Better control of nominal GDP to stabilize debt-to-GDP without fiscal consolidation.
  - Use of seigniorage revenues to help repay debt (potentially higher inflation).
  - Large purchases of government debt can coordinate creditors and reduce likelihood of runs/multiple equilibria.
- Countries defined as having constrained monetary policy in the sample include – Belgium, Denmark, France, Finland, Germany, Greece, Ireland, Italy and Spain.
- Countries defined as having unconstrained monetary policy in the sample include – the Australia, Canada, Norway, Sweden, the United Kingdom and the United States.

### Response of interest rates
- Marginal interest rate response to a +1pt primary balance shock: small increase, peaking at around 0.08pts, similar for both groups.
- Marginal interest rate response to a -1pt growth shock: essentially insignificant for both groups.
- Response to a +10pt debt shock (pseudo IRFs; no confidence bands calculated):
  - Constrained group: marginal interest rate increases sharply and persists for around 1.5 years.
  - Unconstrained group: marginal interest rate persistently falls.
- Interpretation: constrained countries face perceived higher credit risk after debt shocks; unconstrained countries may use monetary policy tools (asset purchases, forward guidance) to manipulate long-term rates and stabilize debt dynamics. The estimation approach cannot precisely determine causality.

### Response of the primary balance
- Shock to marginal interest rate (+1pt):
  - Constrained group: large and persistent increase in primary balance.
  - Unconstrained group: virtually no reaction.
- Shock to growth (-1pt):
  - Constrained group: persistently tighter primary balance to stabilize debt.
  - Unconstrained group: structural primary balance does not react; monetary policy can offset demand shocks.
- Shock to debt (+10pt):
  - Both groups tighten fiscal policy, but the constrained group’s response is much larger.
  - Confirms that the primary balance reacts proportionally to the debt level (Bohn-type behavior), with stronger tightening where monetary policy is constrained.
- Caveat: many constrained-group countries were subject to euro area Stability and Growth Pact (SGP) rules; authors argue SGP is unlikely to be driving results because rules were broken and unconstrained-group countries also had fiscal rules but behaved differently.

### Evolution of debt following shocks
- Growth shock (-1pt):
  - Median debt levels increase by around 2–2.5 percent of GDP at the peak for both groups.
  - Constrained group: slightly larger peak impact due to larger increases in marginal interest rates, but faster return to pre-shock level via stronger primary balance response.
  - Unconstrained group: slower, more measured reduction in debt.
- Interest rate shock (+1pt):
  - Unconstrained group: impact on debt quickly dies out.
  - Constrained group: impact on debt is highly persistent.
- Primary balance shock (negative 1pts):
  - Unconstrained group: peak impact on debt is 2 pts and is highly persistent; debt is maintained at higher level for a long period.
  - Constrained group: reduce debt at a much quicker rate.
- Debt shock (+10pt):
  - Both groups show debt returning to pre-shock level after around 5 years; underlying mechanisms differ (unconstrained rely more on stimulating growth, constrained on fiscal consolidation).

### Diagnostics and robustness
- Stationarity:
  - ADF and KPSS tests used. Four countries—Japan, Portugal, Austria and the Netherlands—had at least one non-stationary series and were dropped from the sample.
- Residuals:
  - Jarque-Bera test for normality produced mixed results; reliance on assumption that residuals are asymptotically normally distributed.
- Robustness checks:
  - Lag length varied (e.g., 2 vs 4 quarters): IRFs maintained same sign, similar magnitude and statistical significance.
  - Sample restricted to exclude crisis years (2008-2014): U.S. and U.K. show no significant change; other countries with shorter series show wider confidence bands and some shocks become insignificant.
  - SVAR identification varied (Cholesky decomposition vs imposing automatic stabilizer elasticities): results largely unchanged.
  - Removing outliers (e.g., Norway) does not materially change results.
- Sample period: 1999-2014 (characterized by adoption of the euro and subsequent euro area crisis episodes).

### Conclusion and implications
- The SVAR with endogenous debt accumulation captures interactions among primary balance, interest rate, growth and inflation, and shows feedback effects that can mitigate or amplify debt shocks.
- Sovereign credit markets:
  - Do not systematically respond to shocks to growth or the primary balance on the sample average.
  - Are sensitive to the debt level, especially for constrained monetary policy countries where market interest rates react positively to higher debt.
  - For unconstrained policy countries, higher debt may be associated with downward pressure on long-term rates.
- Policy implications:
  - Monetary policy matters for sovereign debt sustainability; unconstrained monetary regimes have more tools to manage debt dynamics (monetary offset, seigniorage, debt purchases).
  - Constrained monetary policy countries are more reliant on fiscal adjustment to stabilize debt.
- Suggested extensions:
  - Use a panel VAR to derive more systematic median IRFs.
  - Extend analysis to emerging market economies, requiring addition of exchange rate term to control for foreign currency denominated debt.
  - Consider Bayesian SVAR estimation or local projection models (Jordà, 2005) as alternative specifications.

*Source: IMF working paper content unit _wp15137 (excerpts provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15137.pdf_