## wpiea2020231-print-pdf

## Source details

**Canonical URL:** [wpiea2020231-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020231-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020231-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020231-print-pdf.pdf.json)

---

### I. INTRODUCTION — scope and main research question
- Climate change context:
  - Global average surface temperature has risen by 1.1 degrees Celsius since 1880 and global annual mean temperatures could increase by as much as 4 degrees Celsius over the next century.
- Research gap and objective:
  - No prior empirical research on how climate change affects sovereign default risk; this paper estimates the impact of climate change on the probability of sovereign default for a panel of countries.
- Sample and data:
  - 116 countries over the period 1995–2017.
  - Sovereign default data from a database developed by the Bank of Canada and the Bank of England.
  - Climate vulnerability and resilience measured by ND-GAIN indices.
  - Macroeconomic and financial controls from IMF IFS, WEO, and World Bank WDI databases.

### II. Key findings (empirical results)
- Core empirical conclusions:
  - Climate change vulnerability and resilience have significant effects on the probability of sovereign debt default, especially among low-income countries.
  - Countries with greater vulnerability face a higher likelihood of debt default; higher resilience is associated with a lower risk of sovereign default.
- Quantified effects (benchmark/logit specifications and summaries):
  - Climate change vulnerability:
    - Coefficient range reported across specifications: 0.122 to 0.407.
    - Benchmark statement: "a one percentage point increase in climate change vulnerability is associated with an increase of 0.41 percent in the probability of debt default."
  - Climate change resilience:
    - Coefficient range reported across specifications: -0.324 to -0.402.
    - Benchmark statement: "a one percentage point increase in climate change resilience is associated with a decrease of 0.33 percent in the likelihood of debt default."
- Robustness and sensitivity:
  - Results remain robust to inclusion of additional controls (foreign reserves, external debt, budget balance, political and institutional factors).
  - Alternative estimators used: probit model confirms the impact (with caveats about fixed effects and the incidental parameter problem).
  - Models estimated for full sample and sub-samples (EMEs and LICs).

### III. Data and measurement details
- ND-GAIN indices:
  - ND-GAIN database covers 184 countries over 1995–2017.
  - Composite indices based on 45 indicators: 36 variables for vulnerability and 9 variables for resilience.
  - Vulnerability measures exposure, sensitivity, and capacity to adapt across six sectors: food, water, health, ecosystem services, human habitat, infrastructure.
  - Resilience assesses capacity to apply economic investments and convert them to adaptation actions across three areas: economic, governance, social resilience.
  - Summary statistics (as reported): Mean value of climate change resilience is 33.7; minimum 0.2; maximum 71.3.
- Sovereign default definition:
  - Default defined when the database reports a non-missing and non-zero nominal value of government debt in default in a given year.
  - Default considered when debt service is not made on the due date or within a specified grace period, payments are not made within a guarantee time frame, or creditors incur material economic losses on sovereign debt held due to government action.

### IV. Empirical strategy and model specification
- Baseline framework:
  - Binary choice: logistic regression (logit) modeling Pr(D = 1 | X) where D = 1 denotes sovereign default in a year.
  - Structural specification includes country fixed effects (η_i) and time effects (μ_t); robust standard errors clustered at country level.
  - Main explanatory variables: lagged measures of climate change vulnerability and resilience (i, t-1).
- Control variables (lagged by one year):
  - Government debt (share of GDP)
  - Real GDP per capita
  - Real GDP growth
  - Consumer price inflation
  - Financial development (domestic credit to private sector as share of GDP)
  - Terms-of-trade index
  - Real effective exchange rate (REER) index
- Estimation samples:
  - Full sample (116 countries, 1995–2017).
  - Subsamples: EMEs and LICs; results reported separately.

### V. Representative reported results and key coefficients
- Logit summary highlights (reported examples from Tables 1–2):
  - Climate vulnerability reported coefficient examples: 0.4071**, 0.0279*, 0.1223***.
  - Climate resilience reported coefficient examples: -0.3327***, -0.3240***, -0.4017*.
  - Pseudo R-squared examples (Table 1): 0.4260, 0.1390, 0.3220.
  - Pseudo R-squared examples (Table 2): 0.4398, 0.4165, 0.5095.
  - Number of observations examples (Table 1): 952, 1,026, 875.
  - Number of observations examples (Table 2): 9526, 942, 4,258.

- Probit Model (Table 3) selected coefficients and exact values:
  - Climate vulnerability: 0.0568***; 0.0216**; 0.0637*** (standard errors (0.005)(0.009)(0.010))
  - Climate resilience: -0.0281***; -0.0169***; -0.0270*** (standard errors (0.004)(0.005)(0.005))
  - Real GDP growth: -0.0052; -0.0029; -0.0009; -0.0012; -0.0147; -0.0202 (standard errors (0.005)(0.005)(0.005)(0.005)(0.012)(0.014))
  - Real GDP per capita (t-1): -0.0594***; -0.0557***; -0.0843***; -0.0862***; 0.0537*; 0.0291 (standard errors (0.011)(0.012)(0.014)(0.015)(0.028)(0.028))
  - Inflation (t-1): 0.0005; -0.0009; 0.0002; -0.0002; 0.0259***; 0.0260*** (standard errors (0.002)(0.002)(0.002)(0.002)(0.009)(0.009))
  - Government debt (t-1): 0.0075***; 0.0084***; 0.0118***; 0.0125***; 0.0012; 0.0016 (standard errors (0.001)(0.001)(0.001)(0.001)(0.001)(0.001))
  - Financial development (t-1): -0.0107***; -0.0122***; -0.0117***; -0.0102***; -0.0128***; -0.0088*** (standard errors (0.001)(0.002)(0.002)(0.002)(0.003)(0.003))
  - Terms-of-trade (t-1): 0.1049; 0.2935**; 0.3388*; 0.4068**; -0.0709; -0.2497 (standard errors (0.123)(0.122)(0.175)(0.182)(0.176)(0.195))
  - REER (t-1): 0.1194; 0.2876; 0.1815; 0.1741; 0.0146; -0.1239 (standard errors (0.184)(0.197)(0.226)(0.229)(0.373)(0.344))
  - Number of countries: 116; 116; 61; 61; 55; 55
  - Number of observations: 1,901; 1,901; 1,026; 1,026; 875; 875
  - Pseudo R-squared: 0.2143; 0.1818; 0.1298; 0.1313; 0.2859; 0.3131

- Appendix Table A2 (Climate Vulnerability by Type of Debt) selected coefficients:
  - Climate vulnerability: 0.1488**; 0.0105; 0.0409; 0.0637; 0.0208; 0.2475*** (standard errors (0.066)(0.115)(0.070)(0.110)(0.062)(0.077))
  - Real GDP growth: -0.0096; -0.0844*; -0.0332; -0.1587**; -0.0809*; -0.0126 (standard errors (0.024)(0.048)(0.037)(0.074)(0.050)(0.030))
  - Real GDP per capita (t-1): -0.4800***; -0.1406; 0.0054; -0.0185; 0.0671; -0.2661 (standard errors (0.175)(0.289)(0.115)(0.263)(0.142)(0.214))
  - Government debt (t-1): 0.0164***; 0.0853***; 0.0260***; 0.0184; 0.0205***; 0.0799*** (standard errors (0.004)(0.015)(0.007)(0.015)(0.005)(0.011))
  - Financial development (t-1): -0.0289***; -0.1162***; 0.0003; -0.1524***; -0.0504*; -0.0813*** (standard errors (0.010)(0.034)(0.009)(0.057)(0.027)(0.022))
  - REER (t-1): 0.0914; -1.8403; -3.4492***; -3.6714**; 1.3509; -3.8011*** (standard errors (0.568)(1.521)(1.170)(1.826)(1.011)(1.170))
  - Number of countries: 9; 16; 23; 0; 23; 277
  - Number of observations: 1,418; 956; 520; 384; 348; 852

- Appendix Table A3 (Climate Resilience by Type of Debt) selected coefficients:
  - Climate resilience: -0.1818***; 0.1463*; -0.0953**; -0.0428; -0.1641**; -0.1081** (standard errors (0.047)(0.084)(0.043)(0.091)(0.068)(0.053))
  - Real GDP per capita (t-1): -0.4309**; -0.1192; 0.0446; -0.0959; 0.0209; -0.2323 (standard errors (0.175)(0.233)(0.110)(0.272)(0.136)(0.207))
  - Financial development (t-1): -0.0198**; -0.1275***; 0.0079; -0.1579***; -0.0199; -0.0874*** (standard errors (0.010)(0.033)(0.010)(0.057)(0.026)(0.023))
  - REER (t-1): 0.2976; -1.7767; -3.2978***; -3.9067**; 2.7642**; -3.3287*** (standard errors (0.570)(1.489)(1.158)(1.903)(1.110)(1.140))
  - Number of observations: 1,418; 956; 520; 384; 348; 852

- Appendix Table 4 (Rare-Events Logistic Estimator) selected coefficients:
  - resilience: -0.0454***; -0.0274***; -0.0437*** (standard errors (0.006)(0.009)(0.009))
  - vulnerability: 0.1052***; 0.0376**; 0.1145*** (standard errors (0.010)(0.016)(0.019))
  - real GDP per capita (t-1): -0.0957***; -0.1077***; -0.1374***; -0.1360***; 0.1069**; 0.0430 (standard errors (0.022)(0.020)(0.026)(0.025)(0.054)(0.056))
  - debt (% GDP) (t-1): 0.0143***; 0.0116***; 0.0205***; 0.0191***; 0.0011; 0.0014 (standard errors (0.002)(0.002)(0.003)(0.002)(0.002)(0.003))
  - private credit (% GDP) (t-1): -0.0210***; -0.0171***; -0.0166***; -0.0190***; -0.0229***; -0.0143*** (standard errors (0.003)(0.002)(0.003)(0.003)(0.005)(0.004))
  - ln Terms of Trade (t-1): 0.4878**; 0.2003; 0.6743**; 0.5562*; -0.1398; -0.5651 (standard errors (0.205)(0.223)(0.303)(0.294)(0.318)(0.367))
  - ln REER (t-1): 0.5407; 0.3043; 0.2377; 0.2657; -0.1082; -0.4402 (standard errors (0.360)(0.319)(0.384)(0.376)(0.779)(0.700))
  - Number of countries: 116; 116; 61; 61; 55; 55
  - Number of observations: 1,901; 1,812; 1,026; 1,026; 875; 875

### VI. Robustness checks and rare-events estimation
- Rationale:
  - Rare-events bias can arise when there are many panel units with no default events; King and Zeng (2001) relogit correction applied.
- Method and outcome:
  - Use the relogit estimator (bias-correction) to reduce mean square error in the presence of rare events.
  - Relogit results (Appendix Table A4) validate baseline findings and remain robust to inclusion of additional controls such as government revenues and exports.

### VII. Interpretation and policy implications
- Interpretation:
  - Physical risks from climate change materially affect sovereign default risk through macroeconomic channels: vulnerability raises default probability, resilience lowers it.
  - Effects are economically meaningful and particularly pronounced in low-income and developing countries.
- Policy recommendations:
  - Enhancing structural resilience through cost-effective mitigation and adaptation investments to reduce sovereign default risk.
  - Strengthening financial resilience via fiscal buffers and insurance schemes to cope with climate-driven fiscal shocks.
  - Improving economic diversification and policy management to reduce the likelihood that climate shocks translate into sovereign default.
  - For low-income countries with limited fiscal capacity: consider mechanisms such as debt-for-nature swaps to mobilize resources for environmental conservation while reducing the debt burden.

*Italic: Content summarized from wpiea2020231-print-pdf - References (source PDF content provided).*

### References .............................................................................................................

### wpiea2020231-print-pdf - References .....................................................................................................

### I. INTRODUCTION — scope and main research question
- Climate change poses an existential threat; global average surface temperature has risen by 1.1 degrees Celsius since 1880 and global annual mean temperatures could increase by as much as 4 degrees Celsius over the next century (IPCC, 2007; Stern, 2007; IPCC, 2014).
- Research gap addressed: no prior empirical research on how climate change affects sovereign default risk; this paper estimates the impact of climate change on the probability of sovereign default for a panel of countries.
- Sample and data sources summarized in the paper:
  - 116 countries over the period 1995–2017.
  - Sovereign default data from a database developed by the Bank of Canada and the Bank of England.
  - Climate vulnerability and resilience measured by ND-GAIN indices.
  - Macroeconomic and financial controls from IMF IFS, WEO, and World Bank WDI databases.

### II. Key findings (empirical results)
- Main empirical result:
  - Climate change vulnerability and resilience have significant effects on the probability of sovereign debt default, especially among low-income countries.
  - Countries with greater vulnerability to climate change face a higher likelihood of debt default compared to more climate-resilient countries.
  - Higher climate change resilience is associated with a lower risk of sovereign default.
- Quantified effects (from benchmark/logit specifications and summary statements):
  - Climate change vulnerability:
    - Coefficient range reported across specifications: 0.122 to 0.407.
    - Benchmark: "a one percentage point increase in climate change vulnerability is associated with an increase of 0.41 percent in the probability of debt default."
  - Climate change resilience:
    - Coefficient range reported across specifications: -0.324 to -0.402.
    - Benchmark: "a one percentage point increase in climate change resilience is associated with a decrease of 0.33 percent in the likelihood of debt default."
- Robustness and sensitivity:
  - Results remain robust to inclusion of additional controls (foreign reserves, external debt, budget balance, political and institutional factors).
  - Alternative estimators used: probit model confirms the impact (with caveats about fixed effects and the incidental parameter problem).
  - Models estimated for full sample and sub-samples (emerging market economies and low-income countries).

### III. Data and measurement details
- ND-GAIN indices:
  - ND-GAIN database covers 184 countries over 1995–2017 (dataset used for climate indices).
  - Composite indices based on 45 indicators: 36 variables for vulnerability and 9 variables for resilience.
  - Vulnerability: measures exposure, sensitivity, and capacity to adapt across six life-supporting sectors—food, water, health, ecosystem services, human habitat, infrastructure.
  - Resilience: assesses capacity to apply economic investments and convert them to adaptation actions across three areas—economic, governance, social resilience.
  - Summary statistics reported in prose:
    - Mean value of climate change resilience is 33.7; minimum 0.2; maximum 71.3.
  - Spatial/time heterogeneity highlighted (maps and figures comparing 1995 vs 2017 for vulnerability and resilience).
- Sovereign default definition and data:
  - Default defined when the database reports a non-missing and non-zero nominal value of government debt in default in a given year.
  - Database considers a default when debt service is not made on the due date or within a specified grace period, payments are not made within a guarantee time frame, or creditors incur material economic losses on sovereign debt held due to government action.

### IV. Empirical strategy and model specification
- Binary choice framework:
  - Baseline: logistic regression (logit) modeling Pr(D = 1 | X) where D = 1 denotes sovereign default in a year.
  - Structural specification described (logit probability and latent-variable formulation).
  - Main explanatory variables: lagged measures of climate change vulnerability and resilience (country i, time t-1).
  - Control variables (lagged by one year) include:
    - Government debt (share of GDP)
    - Real GDP per capita
    - Real GDP growth
    - Consumer price inflation
    - Financial development (domestic credit to private sector as share of GDP)
    - Terms-of-trade index
    - Real effective exchange rate (REER) index
  - Fixed effects and time effects included (η_i and μ_t); robust standard errors clustered at country level to account for heteroskedasticity.
- Estimation samples:
  - Full sample (116 countries, 1995–2017).
  - Subsamples: EMEs and LICs (emerging market economies and low-income countries); results reported separately.
- Representative reported logit results (textual summary of reported table outcomes):
  - Climate vulnerability positive and statistically significant across specifications; reported coefficient examples: 0.4071**, 0.0279*, 0.1223*** (as presented in Table 1).
  - Climate resilience negative and statistically significant across specifications; reported coefficient examples: -0.3327***, -0.3240***, -0.4017* (as presented in Table 2).
  - Benchmark pseudo R-squared and sample sizes mentioned in tables (as presented): Pseudo R-squared values reported include 0.4260, 0.1390, 0.3220 (Table 1) and 0.4398, 0.4165, 0.5095 (Table 2). Number of observations reported include 952, 1,026, 875 (Table 1) and 9526, 942, 4258 (Table 2) as they appear in the source tables.

### V. Interpretation and policy implications
- Interpretation:
  - Physical risks from climate change materially affect sovereign default risk through macroeconomic channels; vulnerability raises default probability, resilience lowers it.
  - Effects are economically meaningful and particularly pronounced in low-income and developing countries.
- Policy implications emphasized in the paper:
  - Enhancing structural resilience through cost-effective mitigation and adaptation can reduce sovereign default risk.
  - Strengthening financial resilience via fiscal buffers and insurance schemes can help cope with climate-driven fiscal shocks.
  - Improving economic diversification and policy management can reduce the likelihood that climate shocks translate into sovereign default.
  - These policy levers are especially relevant for developing countries that are relatively more vulnerable to climate risks.

*Italic: Content summarized from wpiea2020231-print-pdf - References (source PDF content provided).*

### Appendix Tables A2-A3, are broadly consistent with the baseline findings, although a small

### Appendix Tables A2-A3, are broadly consistent with the baseline findings, although a small

### Robustness: rare-events logistic (relogit) estimator
- Rationale:
  - Maximum Likelihood estimates in logistic regression are consistent but only asymptotically unbiased; problem arises when many panel units (default episodes) have no events and country-specific indicators perfectly predict zeroes.
  - King and Zeng (2001) describe rare events as “dozens to thousands of times fewer ones [...] than zeroes.”
- Method:
  - Use the relogit estimator (bias-correction method of King and Zeng (2001)) for dichotomous dependent variables to reduce mean square error in presence of rare events.
  - Model specified as Pr(D=1|X) = Φ(Z′θ) equivalent to Pr(D=1|X) = Φ(γ + β1/COVID??? + β2 X) (equation (3) in source), with Φ(⋅) the logistic function.
  - Variance of estimated coefficients expressed as VCV(θ̂) = (Z′ V Z)−1 where V is diagonal with entries Φ(⋅)⋅[1−Φ(⋅)].
  - Follow King and Zeng (1999a, 1999b) and correct for small sample and rare events biases; sampling design random or conditional on Z′.
- Outcome:
  - Results presented in Appendix Table A4 (relogit) validate the baseline findings and remain robust to inclusion of additional control variables such as government revenues and exports.

### Main empirical findings on climate and sovereign default
- Data and scope:
  - Panel of 116 countries over the period 1995–2017.
- Core results:
  - Climate change vulnerability and resilience significantly affect the probability of sovereign debt default, especially among low-income countries.
  - Greater climate vulnerability associated with higher likelihood of debt default.
  - Greater climate resilience associated with lower likelihood of debt default.
- Robustness:
  - Findings remain robust to alternative measures of debt default, model specifications, and estimation methodologies, including relogit rare-events logistic estimator.

### Key coefficient estimates (selected tables and values preserved exactly)
- From Table 3 (Probit Model) summary (Specification columns 1–6):
  - Climate vulnerability: 0.0568***; 0.0216**; 0.0637*** (standard errors (0.005)(0.009)(0.010))
  - Climate resilience: -0.0281***; -0.0169***; -0.0270*** (standard errors (0.004)(0.005)(0.005))
  - Real GDP growth: -0.0052; -0.0029; -0.0009; -0.0012; -0.0147; -0.0202 (standard errors (0.005)(0.005)(0.005)(0.005)(0.012)(0.014))
  - Real GDP per capita (t-1): -0.0594***; -0.0557***; -0.0843***; -0.0862***; 0.0537*; 0.0291 (standard errors (0.011)(0.012)(0.014)(0.015)(0.028)(0.028))
  - Inflation (t-1): 0.0005; -0.0009; 0.0002; -0.0002; 0.0259***; 0.0260*** (standard errors (0.002)(0.002)(0.002)(0.002)(0.009)(0.009))
  - Government debt (t-1): 0.0075***; 0.0084***; 0.0118***; 0.0125***; 0.0012; 0.0016 (standard errors (0.001)(0.001)(0.001)(0.001)(0.001)(0.001))
  - Financial development (t-1): -0.0107***; -0.0122***; -0.0117***; -0.0102***; -0.0128***; -0.0088*** (standard errors (0.001)(0.002)(0.002)(0.002)(0.003)(0.003))
  - Terms-of-trade (t-1): 0.1049; 0.2935**; 0.3388*; 0.4068**; -0.0709; -0.2497 (standard errors (0.123)(0.122)(0.175)(0.182)(0.176)(0.195))
  - REER (t-1): 0.1194; 0.2876; 0.1815; 0.1741; 0.0146; -0.1239 (standard errors (0.184)(0.197)(0.226)(0.229)(0.373)(0.344))
  - Number of countries: 116; 116; 61; 61; 55; 55
  - Number of observations: 1,901; 1,901; 1,026; 1,026; 875; 875
  - Pseudo R-squared: 0.2143; 0.1818; 0.1298; 0.1313; 0.2859; 0.3131
- From Appendix Table A2 (Climate Vulnerability by Type of Debt) selected coefficients:
  - Climate vulnerability: 0.1488**; 0.0105; 0.0409; 0.0637; 0.0208; 0.2475*** (standard errors (0.066)(0.115)(0.070)(0.110)(0.062)(0.077))
  - Real GDP growth: -0.0096; -0.0844*; -0.0332; -0.1587**; -0.0809*; -0.0126 (standard errors (0.024)(0.048)(0.037)(0.074)(0.050)(0.030))
  - Real GDP per capita (t-1): -0.4800***; -0.1406; 0.0054; -0.0185; 0.0671; -0.2661 (standard errors (0.175)(0.289)(0.115)(0.263)(0.142)(0.214))
  - Government debt (t-1): 0.0164***; 0.0853***; 0.0260***; 0.0184; 0.0205***; 0.0799*** (standard errors (0.004)(0.015)(0.007)(0.015)(0.005)(0.011))
  - Financial development (t-1): -0.0289***; -0.1162***; 0.0003; -0.1524***; -0.0504*; -0.0813*** (standard errors (0.010)(0.034)(0.009)(0.057)(0.027)(0.022))
  - REER (t-1): 0.0914; -1.8403; -3.4492***; -3.6714**; 1.3509; -3.8011*** (standard errors (0.568)(1.521)(1.170)(1.826)(1.011)(1.170))
  - Number of countries: 9; 16; 23; 0; 23; 277 (table formatting indicates samples across columns)
  - Number of observations: 1,418; 956; 520; 384; 348; 852
- From Appendix Table A3 (Climate Resilience by Type of Debt) selected coefficients:
  - Climate resilience: -0.1818***; 0.1463*; -0.0953**; -0.0428; -0.1641**; -0.1081** (standard errors (0.047)(0.084)(0.043)(0.091)(0.068)(0.053))
  - Real GDP per capita (t-1): -0.4309**; -0.1192; 0.0446; -0.0959; 0.0209; -0.2323 (standard errors (0.175)(0.233)(0.110)(0.272)(0.136)(0.207))
  - Financial development (t-1): -0.0198**; -0.1275***; 0.0079; -0.1579***; -0.0199; -0.0874*** (standard errors (0.010)(0.033)(0.010)(0.057)(0.026)(0.023))
  - REER (t-1): 0.2976; -1.7767; -3.2978***; -3.9067**; 2.7642**; -3.3287*** (standard errors (0.570)(1.489)(1.158)(1.903)(1.110)(1.140))
  - Number of observations: 1,418; 956; 520; 384; 348; 852
- From Appendix Table 4 (Rare-Events Logistic Estimator) selected coefficients:
  - resilience: -0.0454***; -0.0274***; -0.0437*** (standard errors (0.006)(0.009)(0.009))
  - vulnerability: 0.1052***; 0.0376**; 0.1145*** (standard errors (0.010)(0.016)(0.019))
  - real GDP per capita (t-1): -0.0957***; -0.1077***; -0.1374***; -0.1360***; 0.1069**; 0.0430 (standard errors (0.022)(0.020)(0.026)(0.025)(0.054)(0.056))
  - debt (% GDP) (t-1): 0.0143***; 0.0116***; 0.0205***; 0.0191***; 0.0011; 0.0014 (standard errors (0.002)(0.002)(0.003)(0.002)(0.002)(0.003))
  - private credit (% GDP) (t-1): -0.0210***; -0.0171***; -0.0166***; -0.0190***; -0.0229***; -0.0143*** (standard errors (0.003)(0.002)(0.003)(0.003)(0.005)(0.004))
  - ln Terms of Trade (t-1): 0.4878**; 0.2003; 0.6743**; 0.5562*; -0.1398; -0.5651 (standard errors (0.205)(0.223)(0.303)(0.294)(0.318)(0.367))
  - ln REER (t-1): 0.5407; 0.3043; 0.2377; 0.2657; -0.1082; -0.4402 (standard errors (0.360)(0.319)(0.384)(0.376)(0.779)(0.700))
  - Number of countries: 116; 116; 61; 61; 55; 55
  - Number of observations: 1,901; 1,812; 1,026; 1,026; 875; 875

### Policy implications
- Enhancing structural resilience reduces sovereign default risk:
  - Invest in cost-effective mitigation and adaptation.
  - Strengthen financial resilience through fiscal buffers and insurance schemes.
  - Improve economic diversification and policy management to cope with climate impacts on public finances.
- For low-income countries with limited fiscal capacity:
  - Debt-for-nature swaps could mobilize resources for environmental conservation while reducing the debt burden (Hansen, 1989).

*Source: Appendix Tables and surrounding text in the supplied IMF working paper content.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020231-print-pdf.pdf_