## 12. Impact of Fiscal Shocks on the Current Account Balance—Public Debt

## Source details

**Canonical URL:** [12. Impact of Fiscal Shocks on the Current Account Balance—Public Debt](https://www.imf.org/-/media/files/publications/wp/2018/wp18170.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18170.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18170.pdf.json)

---

### Introduction and research question
- Objective: quantify the relationship between fiscal and current account deficits in developing countries (EMDEs) and identify channels and heterogeneity.
- Sample: unbalanced panel of 114 EMDEs over the period 1990-2015.
- Identification of fiscal shocks: forecast errors in government spending using IMF WEO vintages (forecasts made in October of the same year), purged of predictable components by projecting on lags of macro variables.
- Estimation method: local projections (Jordà, 1995/2005) with country and time fixed effects; instrument the ratio of government budget balance to GDP with identified fiscal shocks (Ramey and Zubairy, 2018).

### Data and empirical methodology — key elements and statistics
- Fiscal shock measure:
  - 퐹퐸푖,푡|푡−1 = ∆lnG푖,푡 − ∆lnG푖,푡|−1 (forecast error in government spending growth).
- Forecast vintage used to minimize endogeneity:
  - Forecasts made in October of the same year (robust to using April forecasts).
- Purging predictable components:
  - Projection on lags of output, government spending, government revenue, real exchange rate, current account and inflation; residual used as shock.
- Shock distribution (Figure A1):
  - average = -0.1 percent
  - median = 0.1 percent
  - bulk of shocks (between the 1st and the 99th percentile) lie between -99 and 100 percent
- First-stage result (instrument relevance):
  - a 100 percent unanticipated increase in government spending reduces, on average, the budget balance by about 2 percent of GDP
  - F-statistic = 14.35
- Alternative first-stage when using April forecasts:
  - a 100 percent unanticipated increase reduces the budget balance by about 2.5 percent of GDP
  - F-statistic = 19.15
- Forecast quality:
  - mean square error of government spending growth forecasts larger in EMDEs than in AEs (Table A2)
  - robustness checks drop observations with mean square errors above the 5th percentile

### Baseline results and robustness checks
- Main estimate:
  - a 1 percentage point of GDP unanticipated improvement in the budget balance improves the current account balance by 0.8 percentage point of GDP a year after the shock
  - effect statistically significant up to two years after the shock
- Emerging market subsample:
  - a fiscal shock increases the current account balance by about 1.1 percentage points after a year (not statistically different from baseline)
- Robustness checks that yield similar results:
  - controlling for current and lagged output growth innovations (growth forecast errors)
  - including changes in terms of trade and financial crises as controls
  - using April forecasts
  - excluding large mean-square-error observations

### Comparison with other fiscal shock measures
- Using changes in the cyclically-adjusted budget balance (CAB):
  - a 1 percentage point increase in the CAB improves the current account by 0.25 percentage point of GDP a year after the shock
  - effects become statistically insignificant afterward
  - CAB-based magnitude is substantially smaller than the baseline estimate using forecast-error shocks — likely due to endogeneity/omitted variable bias in CAB measures
- Magnitude for advanced economies (narrative approach comparators):
  - peak twin deficit magnitude about 0.8 (Figure 5 Panel A), close to narrative-based estimates

### Sign, size, and asymmetry of shocks
- No statistically significant difference in the current account response between positive shocks (expansions) and negative shocks (consolidations) using the interaction specification in equation (3).

### Channels: saving, investment, and exchange rate
- Investment channel:
  - an unanticipated improvement in the fiscal balance of 1 percentage point of GDP lowers domestic investment by 0.8 percentage point of GDP a year after the shock
  - investment falls by 1.6 percentage point of GDP at the peak
  - the current account adjustment mainly occurs through a persistent reduction in investment rather than a large increase in saving
- Saving channel:
  - impact on saving is positive but significant only two years after the shock
- Exchange rate channel:
  - nominal exchange rate response: a 1 percent of GDP fiscal consolidation is associated with a depreciation of about 0.7 percent in the year of the shock
  - depreciation of about 1 percent in the medium term (Figure A2)
  - heterogeneity by exchange rate regime: impact on the current account is significantly larger and more persistent in countries with less flexible exchange rate arrangements; effect not statistically different from zero in more flexible regimes

### Heterogeneity by cyclical, structural, and policy factors
- Business cycle (smooth transition specification):
  - fiscal shocks have a larger and more persistent effect on the current account when they occur in recessions
  - effects are not statistically significant in periods of high growth after one year
  - interpretation: larger investment response in weak growth periods, consistent with larger fiscal multipliers in recessions
- Trade openness:
  - threshold: trade openness above the sample average of about 70 percent of GDP defines the “more open” group
  - point estimates suggest larger current account responses in more open economies, but differences are not statistically significant given large standard errors
- Public debt:
  - threshold: public debt-to-GDP above the sample average of about 45 percent of GDP defines the “high-debt” group
  - impact of fiscal shocks is smaller in periods of higher public debt; differences driven by larger investment responses when debt is lower
  - differences in point estimates are generally small relative to standard errors and often not statistically significant

### Quantifying the twin deficits and policy implications
- Twin deficits magnitude:
  - baseline: 1 percent of GDP unanticipated improvement in the budget balance -> 0.8 percentage point of GDP improvement in the current account (average effect)
  - CAB-based measure: 1 percentage point increase in CAB -> 0.25 percentage point of GDP improvement in the current account
- Policy implication:
  - for a given external adjustment target, less fiscal consolidation may be required than commonly assumed if fiscal consolidation is unanticipated and exogenous
  - heterogeneity across states and country characteristics implies that the required fiscal adjustment differs by country and cyclical conditions

### Empirical result: effect of fiscal shocks on Government Budget Balance (Table 1)
- Estimates based on Equation (2); country- and year-fixed effects included; T-statistics in parentheses; ***,** denote significance at 1 and 5 percent, respectively.
  - Shock_t: -1.917***  (-3.79)
  - Shock_t-1: 0.395  (1.09)
  - Shock_t-2: -0.040  (-0.14)
  - Current account_t-1: -0.053**  (-2.49)
  - Current account_t-2: -0.022  (-0.78)
  - N: 1,942
  - R2: 0.12

### Appendix data — government expenditure shocks in EMDEs and forecast accuracy
- Table A1: Government Expenditure Shocks in EMDEs — for each country, Table A1 reports mean shock, SD. Shock, mean square forecast error. Example country entries (exact values reported):
  - Afghanistan: 0.072  0.536  0.266
  - Albania: 0.001  0.261  0.064
  - Algeria: -0.062  0.190  0.038
  - Angola: 0.164  0.599  0.367
  - Argentina: 0.039  0.288  0.079
- Table A2: Mean square error of government spending growth rate forecasts — by country group (exact values):
  - AEs: 2.25
  - EMDEs: 11.9
  - EMs: 9.9
  - LICs: 14.9

*Source: IMF Working Paper — "Impact of Fiscal Shocks on the Current Account Balance—Public Debt", chapter 12, based on an unbalanced panel of 114 EMDEs (1990–2015) and analyses reported in the source PDF.*

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

### References

### Tables
- 1. The effect of fiscal shocks on the Government Budget Balance ..............................24

### Figures
- 1. Impact of Fiscal Shocks on the Current Account ................................................................... 25
- 2. Impact of Fiscal Shocks on the Current Account: EMs vs. LICs .......................................... 26
- 3. Impact of Fiscal Shocks on the Current Account—controlling for other variables ............. 27
- 4. Impact of Fiscal Shocks on the Current Account—robustness to shock measures ............. 28
- 5. Impact of Fiscal Shocks on the Current Account—comparison to the literature ................. 29
- 6. Impact of Fiscal Shocks on the Current Account: Sign of the Shock ................................... 30
- 7. Impact of Fiscal Shocks on Investment and Saving  .............................................................. 31
- 8. Impact of CAB changes on Investment and Saving ............................................................... 32
- 9. Impact of Fiscal Shocks on the Current Account Balance: recessions vs. expansions  ....... 33
- 10. Impact of Fiscal Shocks on the Current Account Balance—Trade Openness ................... 34
- 11. Impact of Fiscal Shocks on the Current Account Balance—Exchange Rate Regimes  .... 35

*Source: wp18170 - References (wp18170 - References.pdf).*

### 12. Impact of Fiscal Shocks on the Current Account Balance—Public Debt....................36

### 12. Impact of Fiscal Shocks on the Current Account Balance—Public Debt

### Introduction and research question
- Objective: quantify the relationship between fiscal and current account deficits in developing countries (EMDEs) and identify channels and heterogeneity.
- Sample: unbalanced panel of 114 EMDEs over the period 1990-2015.
- Identification of fiscal shocks: forecast errors in government spending using IMF WEO vintages (forecasts made in October of the same year), purged of predictable components by projecting on lags of macro variables.
- Estimation method: local projections (Jordà, 1995/2005) with country and time fixed effects; instrument the ratio of government budget balance to GDP with identified fiscal shocks (Ramey and Zubairy, 2018).

### Data and empirical methodology — key elements and statistics
- Fiscal shock measure: 퐹퐸푖,푡|푡−1 = ∆lnG푖,푡 − ∆lnG푖,푡|−1 (forecast error in government spending growth).
- Forecast vintage used to minimize endogeneity: forecasts made in October of the same year (robust to using April forecasts).
- Purging predictable components: projection on lags of output, government spending, government revenue, real exchange rate, current account and inflation; residual used as shock.
- Shock distribution (Figure A1): average = -0.1 percent, median = 0.1 percent; bulk of shocks (between the 1st and the 99th percentile) lie between -99 and 100 percent.
- First-stage result (instrument relevance): a 100 percent unanticipated increase in government spending reduces, on average, the budget balance by about 2 percent of GDP; F-statistic = 14.35.
- Alternative first-stage when using April forecasts: a 100 percent unanticipated increase reduces the budget balance by about 2.5 percent of GDP; F-statistic = 19.15.
- Forecast quality: mean square error of government spending growth forecasts larger in EMDEs than in AEs (Table A2); robustness checks drop observations with mean square errors above the 5th percentile.

### Baseline results and robustness checks
- Main estimate: a 1 percentage point of GDP unanticipated improvement in the budget balance improves the current account balance by 0.8 percentage point of GDP a year after the shock; statistically significant up to two years after the shock.
- Emerging market subsample: a fiscal shock increases the current account balance by about 1.1 percentage points after a year (not statistically different from baseline).
- Robustness:
  - Controlling for current and lagged output growth innovations (growth forecast errors) yields similar results.
  - Including changes in terms of trade and financial crises as controls yields similar results.
  - Using April forecasts and excluding large mean-square-error observations produce results similar to baseline.

### Comparison with other fiscal shock measures
- Using changes in the cyclically-adjusted budget balance (CAB):
  - A 1 percentage point increase in the CAB improves the current account by 0.25 percentage point of GDP a year after the shock; effects become statistically insignificant afterward.
  - The CAB-based magnitude is substantially smaller than the baseline estimate using forecast-error shocks — likely due to endogeneity/omitted variable bias in CAB measures.
- Magnitude for advanced economies (narrative approach comparators): peak twin deficit magnitude about 0.8 (Figure 5 Panel A), close to narrative-based estimates.

### Sign, size, and asymmetry of shocks
- Results indicate no statistically significant difference in the current account response between positive shocks (expansions) and negative shocks (consolidations) using the interaction specification in equation (3).

### Channels: saving, investment, and exchange rate
- Investment channel:
  - An unanticipated improvement in the fiscal balance of 1 percentage point of GDP lowers domestic investment by 0.8 percentage point of GDP a year after the shock and by 1.6 percentage point of GDP at the peak.
  - The current account adjustment mainly occurs through a persistent reduction in investment rather than a large increase in saving.
- Saving channel:
  - Impact on saving is positive but significant only two years after the shock.
- Exchange rate channel:
  - Nominal exchange rate response: a 1 percent of GDP fiscal consolidation is associated with a depreciation of about 0.7 percent in the year of the shock, and of about 1 percent in the medium term (Figure A2).
  - Heterogeneity by exchange rate regime: impact on the current account is significantly larger and more persistent in countries with less flexible exchange rate arrangements; effect not statistically different from zero in more flexible regimes.

### Heterogeneity by cyclical, structural, and policy factors
- Business cycle (smooth transition specification):
  - Fiscal shocks have a larger and more persistent effect on the current account when they occur in recessions; effects are not statistically significant in periods of high growth after one year.
  - Interpretation: larger investment response in weak growth periods, consistent with larger fiscal multipliers in recessions.
- Trade openness:
  - Threshold: trade openness above the sample average of about 70 percent of GDP defines the “more open” group.
  - Point estimates suggest larger current account responses in more open economies, but differences are not statistically significant given large standard errors.
- Public debt:
  - Threshold: public debt-to-GDP above the sample average of about 45 percent of GDP defines the “high-debt” group.
  - Impact of fiscal shocks is smaller in periods of higher public debt; differences driven by larger investment responses when debt is lower.
  - Differences in point estimates are generally small relative to standard errors and often not statistically significant.

### Quantifying the twin deficits and policy implications
- Twin deficits magnitude:
  - Baseline: 1 percent of GDP unanticipated improvement in the budget balance -> 0.8 percentage point of GDP improvement in the current account (average effect).
  - CAB-based measure: 1 percentage point increase in CAB -> 0.25 percentage point of GDP improvement in the current account.
- Policy implication: for a given external adjustment target, less fiscal consolidation may be required than commonly assumed if fiscal consolidation is unanticipated and exogenous; heterogeneity across states and country characteristics implies that the required fiscal adjustment differs by country and cyclical conditions.

*Source: IMF Working Paper — "Impact of Fiscal Shocks on the Current Account Balance—Public Debt", chapter 12, based on an unbalanced panel of 114 EMDEs (1990–2015) and analyses reported in the source PDF.*

### REFERENCES

### wp18170 - REFERENCES

### Empirical result: effect of fiscal shocks on Government Budget Balance
- Table 1 estimates (dependent variable: government budget balance; estimates based on Equation (2); country- and year-fixed effects included):
  - Shock_t: -1.917***  (-3.79)
  - Shock_t-1: 0.395  (1.09)
  - Shock_t-2: -0.040  (-0.14)
  - Current account_t-1: -0.053**  (-2.49)
  - Current account_t-2: -0.022  (-0.78)
  - N: 1,942
  - R2: 0.12
  - Note: T-statistics based on robust clustered standard errors in parentheses. ***,** denote significance at 1 and 5 percent, respectively.

### Main empirical design and interpretation (shared notes across figures)
- Baseline shock interpretation:
  - Responses show the effect of 1 percentage point of GDP exogenous increase in the fiscal balance.
  - X-axis in figures indicates years after the shock at t=0.
  - Blue lines denote the response; dotted lines denote 90 percent confidence bands.
  - Estimates are based on Equation (2) unless otherwise noted; red lines in many panels indicate the baseline response in Figure 1.
- Definition of expansionary vs. contractionary shocks (Figure 6):
  - Expansionary (contractionary) shocks are fiscal shocks with an unanticipated change in government expenditure in percent of GDP below (above) zero.
  - Results in Figure 6 are based on Equation (3).
- Regime- and characteristic-based estimates (Figures 9–12):
  - Figure 9: Estimates based on Equation (4) — low- (Panel A) versus high-growth (Panel B) regime.
  - Figures 10–12: Estimates based on Equation (5) — subgroup comparisons by trade openness, exchange rate regime flexibility, and initial public debt-to-GDP ratio.

### Figures — topics and focal comparisons
- Figure 1: Impact of Fiscal Shocks on the Current Account (percentage points of GDP). Baseline response to 1 percentage point of GDP exogenous increase in the fiscal balance; 90 percent confidence bands.
- Figure 2: Impact by country group — EMs vs. LICs (percentage points of GDP).
  - Panel A: Emerging Markets.
  - Panel B: Low-income Countries.
  - Red lines indicate baseline response in Figure 1.
- Figure 3: Controlling for additional variables (percentage points of GDP).
  - Panel A: Controlling for growth news.
  - Panel B: Controlling for crisis and changes in terms of trade.
  - Red lines indicate baseline response in Figure 1.
- Figure 4: Robustness to shock measures (percentage point of GDP).
  - Panel A: Shocks based on forecasts made in April of the same year.
  - Panel B: Excluding large forecast errors.
  - Note: Large forecast errors are defined as those with mean square errors above the 5th percentile of the distribution.
  - Red lines indicate baseline response in Figure 1.
- Figure 5: Comparison to the literature (percentage point of GDP).
  - Panel A: Advanced Economies — response to 1 percentage point of GDP exogenous increase in the fiscal balance in AEs.
  - Panel B: Developing Economies — CAB versus Unanticipated Fiscal Shocks; blue lines denote response to 1 percentage point of GDP increase in the cyclically-adjusted fiscal balance.
  - Red lines indicate baseline response in Figure 1.
- Figure 6: Impact by sign of the shock (percentage point of GDP).
  - Panel A: Expansionary Shocks.
  - Panel B: Contractionary Shocks.
  - Results based on Equation (3). Red lines indicate baseline response in Figure 1.
- Figure 7: Impact on Investment and Saving (percentage point of GDP).
  - Panel A: Investment.
  - Panel B: Saving.
  - Responses to 1 percentage point of GDP exogenous increase in the fiscal balance.
- Figure 8: Impact of CAB changes on Investment and Saving (percentage point of GDP).
  - Panel A: Investment.
  - Panel B: Saving.
  - Blue lines: response to 1 percentage point of GDP increase in the CAB. Red lines: response to exogenous shocks reported in Figure 7.
- Figure 9: Recessions vs. expansions (percentage point of GDP).
  - Panel A: Recessions (low-growth regime).
  - Panel B: Expansions (high-growth regime).
  - Estimates based on Equation (4). Red lines indicate baseline response in Figure 1.
- Figure 10: Trade openness split (percentage point of GDP).
  - Panel A: More Open Economies.
  - Panel B: Less Open Economies.
  - Estimates based on Equation (5). Red lines indicate baseline response in Figure 1.
- Figure 11: Exchange rate regimes (percentage point of GDP).
  - Panel A: More Flexible Exchange Rate Regimes.
  - Panel B: Less Flexible Exchange Rate Regimes.
  - Estimates based on Equation (5). Red lines indicate baseline response in Figure 1.
- Figure 12: Public Debt split (percentage point of GDP).
  - Panel A: Higher Initial Public Debt-to-GDP ratio.
  - Panel B: Lower Initial Public Debt-to-GDP ratio.
  - Estimates based on Equation (5). Red lines indicate baseline response in Figure 1.
- Appendix Figure A1: Distribution of Government Expenditure Shocks in EMDEs (density plot of fiscal shock).
- Appendix Figure A2: Impact of Fiscal Shocks on the Nominal Exchange Rate (percentage). Response of the nominal exchange rate to 1 percentage point of GDP exogenous increase in the fiscal balance; 90 percent confidence bands.

### Appendix data — government expenditure shocks in EMDEs (Table A1) and forecast accuracy (Table A2)
- Table A1: Government Expenditure Shocks in EMDEs — for each country, Table A1 reports:
  - Mean shock
  - SD. Shock
  - Mean square forecast error
  - Example country entries (exact values as reported):
    - Afghanistan: 0.072  0.536  0.266
    - Albania: 0.001  0.261  0.064
    - Algeria: -0.062  0.190  0.038
    - Angola: 0.164  0.599  0.367
    - Argentina: 0.039  0.288  0.079
    - (Table continues across EMDE sample with country-specific mean shock, SD, and mean square forecast error as reported.)
  - Source: IMF World Economic Outlook and authors calculations.
- Table A2: Mean square error of government spending growth rate forecasts — by country group (exact values):
  - AEs: 2.25
  - EMDEs: 11.9
  - EMs: 9.9
  - LICs: 14.9
  - Source: IMF World Economic Outlook and authors calculations.
  - Note: The group classification is based on the IMF WEO.

*Source: wp18170 - REFERENCES (IMF World Economic Outlook and authors calculations, figures and tables as presented in the source PDF).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18170.pdf_