IMF-周期性财政乘数_政策组合与金融摩擦之谜（英）-2025.5_52页_9mb

Summary of IMF Working Paper: Cyclical Fiscal Multipliers: Policy Mix and Financial Friction Puzzle

This paper analyzes the dynamic relationship between U.S. fiscal multipliers and the business cycle from 1949 to 2018 using a Time-Varying Parameter Vector Autoregression (TVP-VAR) model. The key findings are:

1. Time-Varying Fiscal Multipliers

• Fiscal multipliers were counter-cyclical before the late 1980s, rising during recessions, but became significantly smaller after the 1980s, peaking during certain recessions and declining during others (e.g., the Great Recession).
• This evolution is driven by shifts in the fiscal-monetary policy mix. Multipliers are larger during periods with a fiscally-led policy mix (active fiscal policy, passive monetary policy) and smaller under a monetary-led policy mix (passive fiscal policy, active monetary policy). This aligns with findings from Leeper et al. (2017) and Bianchi and Ilut (2017).

2. Role of Financial Frictions

• The study challenges the literature’s assumption that financial frictions enhance fiscal multipliers by promoting growth. Instead, evidence shows that financial frictions became increasingly constraining in the 2000s, dampening fiscal stimulus effectiveness. Notably, during the Global Financial Crisis, multipliers were significantly lower due to heightened credit spreads, rising household debt, and constrained borrowing (following Gilchrist and Zakrajšek, 2012).

3. Policy Implications

• The findings highlight the importance of accounting for state-dependent multipliers in economic models. The structural break in the 1980s, attributed to monetary policy shifts under Paul Volcker and Reagan’s fiscal stance, marks the transition from efficacy of fiscal stimulus to reduced impact in later recessions.
• The paper underscores the need to integrate professional forecasts into TVP-VAR models to address shock anticipation issues without altering policy anticipation conclusions.

4. Methodological Contributions

• The paper introduces a novel hybrid identification strategy combining sign and short-term zero restrictions, avoiding pitfalls of Cholesky decomposition in fiscal VARs. Using Hamilton (2018)'s linear projection method produces multipliers rather than elasticities, enhancing interpretability.
• Bayesian estimation with Tobit and Lasso priors regulates parameter time variation, enabling analysis of large TVP-VAR models efficiently.

5. Limitations and Future Research

• The model omits key variables (e.g., public debt) limiting the capture of shock transmission mechanisms. Further research could incorporate additional variables or test the multipliers framework in broader economies. Replication using less flexible methods like the Leeper et al. (2017) framework is recommended.

Conclusion

The study demonstrates that fiscal multipliers are not stable over the business cycle, reinforcing the need for dynamic policy adjustments. It introduces a financial friction puzzle, revealing a reversal in the historical belief that greater integration, near-zero rates, and high debt exacerbate crises rather than fiscal dynamics.