1. World Economic Outlook World Economic Outlook Fall 2010

2. Will It Hurt? Macroeconomic Effects of Fiscal Consolidation Daniel Leigh, Pete Devries, Charles Freedman, Jaime Guajardo, Douglas Laxton, and Andrea Pescatori with support from Murad Omoev, Min Song and Jessie Yang.

3. Setting the Scene Advanced economies face challenge of fiscal consolidation. Advanced economies face challenge of fiscal consolidation. What are the macro effects of tax hikes and spending cuts? What are the macro effects of tax hikes and spending cuts? Role of monetary policy, international trade, tax-spending composition, perceived sovereign risk. Role of monetary policy, international trade, tax-spending composition, perceived sovereign risk. Government Debt/GDPFiscal Balance/GDP Source: IMF World Economic Outlook database. Note: Advanced economy weighted average. General government.

4. Main Results Fiscal consolidations are contractionary : GDP falls, unemployment rate rises in the short term. Fiscal consolidations are contractionary : GDP falls, unemployment rate rises in the short term. Monetary easing (ER↓, R↓) + NX boom = key cushioning role. Monetary easing (ER↓, R↓) + NX boom = key cushioning role. Contractionary for both spending cuts and tax hikes but more for tax hikes. Key: R↓ more for spending cuts. Contractionary for both spending cuts and tax hikes but more for tax hikes. Key: R↓ more for spending cuts. Contractionary for both high sovereign default risk and low risk but more for low risk. Contractionary for both high sovereign default risk and low risk but more for low risk. Long-term gains. Long-term gains.

5. Roadmap Looking at History (Empirical Analysis) Looking at History (Empirical Analysis) New approach for identifying episodes. New approach for identifying episodes. Fresh evidence: short-term effects. Fresh evidence: short-term effects. Model Simulations: GIMF Model Simulations: GIMF Zero lower bound, synchronized adjustment. Zero lower bound, synchronized adjustment. Long-term effects. Long-term effects.

6. Identifying Fiscal Consolidation Conventional approach: outcome-based (CAPB). Conventional approach: outcome-based (CAPB). Sample selection bias  expansionary effects. Sample selection bias  expansionary effects. Alesina and Ardagna (2010), many others. Alesina and Ardagna (2010), many others. Action-based definition: historical accounts and records (OECD Economic Surveys, IMF documents, budgets). Action-based definition: historical accounts and records (OECD Economic Surveys, IMF documents, budgets). 15 OECD countries 1980-2009: 173 cases of fiscal consol 15 OECD countries 1980-2009: 173 cases of fiscal consol G7, AUS, BEL, DNK, FIN, IRL, PRT, ESP, SWE. G7, AUS, BEL, DNK, FIN, IRL, PRT, ESP, SWE. Mean size of 173 cases: 1% of GDP. Mean size of 173 cases: 1% of GDP.

7. Example: Fiscal Consolidation in Japan in 1997 Fiscal consolidation totals 1.8 % of GDP. Fiscal consolidation totals 1.8 % of GDP. Tax hikes: of 1.2 % of GDP, spending cuts of 0.6 % of GDP. Tax hikes: of 1.2 % of GDP, spending cuts of 0.6 % of GDP. Part of medium-term strategy of fiscal consolidation (Fiscal Structural Reform Act) in response to large structural fiscal deficit and future demographic pressures. Part of medium-term strategy of fiscal consolidation (Fiscal Structural Reform Act) in response to large structural fiscal deficit and future demographic pressures. Sources: 1997 IMF Staff Report for Japan. Sources: 1997 IMF Staff Report for Japan.

8. Macroeconomic Effects Estimation approach: Romer-Romer-style. Estimation approach: Romer-Romer-style. g: growth rate of real GDP. g: growth rate of real GDP. FC: action-based consolidation in % of GDP. FC: action-based consolidation in % of GDP. Cumulate responses to estimate GDP level. Cumulate responses to estimate GDP level. Robustness: different lag lengths (up to 4), no lags of growth. Similar results.

9. Fiscal Consolidation is Contractionary Impact of 1% of GDP fiscal consolidation. Impact of 1% of GDP fiscal consolidation. GDP down ½ percent. Unemployment rate up ⅓ point. GDP down ½ percent. Unemployment rate up ⅓ point. Note: Consolidation in year t=1. Point estimates and one standard error bands.

10. Usually: Monetary Mitigation Interest Rate (Basis points) Exchange Rate (Percent) Monetary conditions ease in response to fiscal consol. Monetary conditions ease in response to fiscal consol. Interest rates fall. Interest rates fall. Currency looses value (both real and nominal). Currency looses value (both real and nominal). Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands.

11. Transmission Channel: Net Exports NX increase plays key offsetting role. Contribution ↑ 0.5%. NX increase plays key offsetting role. Contribution ↑ 0.5%. Domestic demand ↓ 1%. Domestic demand ↓ 1%. Exports rise 1%, imports fall 1%. Exports rise 1%, imports fall 1%. Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands.

12. Does Composition Matter? GDP Unemployment Rate Tax-based Spending-based Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands. Tax-based vs. spending-based consolidation. Tax-based vs. spending-based consolidation. Both are contractionary, but spending-based less so. Both are contractionary, but spending-based less so.

13. Tax-based Spending-based Domestic Demand Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands. Does Composition Matter? (Cont.)

14. Composition and Monetary Policy Spending-based: interest rate cuts, currency loses value. Spending-based: interest rate cuts, currency loses value. Tax-based: interest rate hikes. Tax-based: interest rate hikes. Policy Rate (Basis points) Real Effective Exchange Rate (Percent) Tax-based Spending-based Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands.

15. Role of Perceived Sovereign Default Risk Low perceived risk  “Keynesian” contraction. Low perceived risk  “Keynesian” contraction. High risk  milder contraction. High risk  milder contraction. Denmark/Ireland = outliers. Denmark/Ireland = outliers. Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands. Low risk High risk Denmark (1983) and Ireland (1987)

16. Contrast with the Literature Our sample using AA (2010) episodes  expansionary effects. Our sample using AA (2010) episodes  expansionary effects. Interpretation: sample selection bias. Interpretation: sample selection bias. GDPUnemployment Rate Action-based approach (large > 1.5%) Standard approach (∆CAPB/GDP > 1.5%) Note: Impact of additional 1% of GDP consolidation in year t=1. Point estimates and one standard error bands.

17. Canada and Rest of the World. Canada and Rest of the World. With zero interest rate floor  contractionary effect doubles. With zero interest rate floor  contractionary effect doubles. ZIRF + synchronized adjustment  effect doubles again. ZIRF + synchronized adjustment  effect doubles again. GIMF: Zero Interest Rate Floor GIMF: Zero Interest Rate Floor With zero interest rate floor Without zero interest rate floor Canada-only fiscal consolidation Global fiscal consolidation Impact of Fiscal Consolidation of 1% of GDP on GDP

18. GIMF: Long-term Effects G3 debt/GDP ↓ 10pp  interest rate ↓ 30 bps. G3 debt/GDP ↓ 10pp  interest rate ↓ 30 bps. Interest payments ↓  taxes ↓ (or spending ↑) Interest payments ↓  taxes ↓ (or spending ↑) GDP ↑ range of outcomes: GDP ↑ range of outcomes: Minimum: 0.5% (savings used to raise general transfers) Minimum: 0.5% (savings used to raise general transfers) Maximum: 1.5% (savings used to cut capital income taxes) Maximum: 1.5% (savings used to cut capital income taxes) Rest of world benefits (lower interest rate). Rest of world benefits (lower interest rate).

19. Lessons for Today Fiscal consolidations are contractionary in short-term. Fiscal consolidations are contractionary in short-term. Monetary easing (ER↓, R↓) + NX boom = key cushioning role. But less today (zero R, synchronized). Monetary easing (ER↓, R↓) + NX boom = key cushioning role. But less today (zero R, synchronized). Less contractionary for high risk than for low risk. Less contractionary for high risk than for low risk. Reforms needed: retirement age, entitlement programs. Reforms needed: retirement age, entitlement programs. Long-term gains. Lower interest rates, lower taxes. Long-term gains. Lower interest rates, lower taxes.

20. Appendix

21. Episodes of Fiscal Consolidation: Action-based vs. Standard Approach DEU 1996 JPN 1999 IRL 2009 ITA 1993 IRL 1982 FIN 1993 FIN 1992 BEL 1984 JPN 2006 FIN 2000 Action-based approach Standard approach (change in cyclically adjusted primary balance)

22. Tax vs. Spending: Additional Results GDP Unemployment Rate Net Exports Contribution Domestic Demand Contribution Tax-based Spending-based Note: Impact of 1% of GDP consolidation in year t=1. Point estimates and one standard error bands.

23. Composition and Monetary Policy (cont.) Why do monetary conditions ease less for tax-based? Why do monetary conditions ease less for tax-based? Central banks view spending cuts favorably (signal discipline)? Central banks view spending cuts favorably (signal discipline)? Inflation impact of (indirect) tax hikes. Inflation impact of (indirect) tax hikes. Policy Rate (Basis points) GDP (Percent) Tax-based (indirect) Tax-based (direct) Spending-based Note: Impact of 1% of GDP consolidation in year t=1.

24. Spending Composition Transfers cuts more benign than cuts to government consumption or investment. Transfers cuts more benign than cuts to government consumption or investment. Confidence effects? Results only suggestive. Confidence effects? Results only suggestive. Note: Impact of 1% of GDP consolidation in year t=1. Cuts to public investment Cuts to public consumption Cuts to government transfers

25. Long-Term Effects (GIMF)