2025-06-08-IMF-周期性财政乘数_政策组合与金融摩擦之谜（英）_53页_9mb

Cyclical Fiscal Multipliers: Policy Mix and Financial Friction Puzzle

Zamid Aligishiev and Hamed Ghiaie
IMF Working Paper WP/25/108

Introduction

This paper investigates dynamic relationships between U.S. government expenditure multipliers and the economy's cyclical position from 1949 to 2018 using a Time-Varying Parameter Vector Autoregression (TVP-VAR) model. It challenges literature that assumes stable fiscal-multiplier relationships with business cycles. The paper was published by the IMF in 2025.

Key Methodology

1. Advanced TVP-VAR Modeling:
   • Extends standard TVP-VAR model with time-varying parameters to analyze 4-quarter lags
   • Incorporates Tobit prior for parameter stability and Lasso prior for variance regulation
2. Identification Strategy:
   • Combines sign restrictions with short-term zero restrictions to identify structural shocks
   • Uses Hamilton's linear projection method for detrending data
   • Employs NBER-defined recessions and incorporates professional forecasts for policy anticipation
3. Unique Implementation:
   • Model allows for four lagged coefficients capturing delayed fiscal effects
   • Free from H-P filter biases by using linear projection for detrending

Main Findings

Multiplier Dynamics

• Period Differences:
  • 1949–late 1980s: Counter-cyclical multipliers higher during recessions
  • Post–1980s: Pro-cyclical multipliers with significantly reduced effectiveness during recessions
• Global Financial Crisis Exception: Especially low multipliers (negative values) during 2008 crisis
• Cyclical Pattern: Multipliers higher during early recessions but lower (occasionally negative) during later 2000s recessions

Determinants

• Policy Mix Effects:
  • Fiscal multipliers significantly larger during periods of fiscal-led policy mixes
  • Reduced effectiveness under monetary-led policy mixes
• Financial Frictions: Reverse relationship observed since 2000s—higher financial frictions correlate with lower fiscal multipliers
• Policy Shift Hypothesis: Changes in fiscally vs. monetarily led policy mixes explain fluctuations

Statistical Verification

• Regime Change: Identified shift around late 1980s due to monetary policy framework changes
• Predictability: Found that interest rate spreads predicted multiplier changes well until mid-1990s
• Robustness Checks: Results unchanged when controlling for policy anticipation using professional forecasts

Policy Implications

1. Challenges common view of counter-cyclical multipliers, demonstrating their time-variation with economic slack
2. Highlights need for policy assessment context-specific to business cycles
3. Reveals that multipliers may decline during certain recent recessions rather than just any recession
4. Shows importance of financial conditions in determining fiscal transmission mechanisms

Innovation & Contribution

• First to use TVP-VAR framework for state-dependent fiscal multipliers
• Provides novel empirical evidence of reversed financial friction-multiplier relationship since 2000s
• Challenges prevailing literature on fiscal multipliers and policy effectiveness
• Offers methodology expanding TVP-VAR applications with professional forecasts and improved identification

Methodological Advantages

• Uses full-period identification with fewer restrictions than alternative approaches
• Avoids the rescaling bias problem of tax shock calculations in fiscal VARs
• Implements Bayesian estimation with proper accounting for parameter uncertainty
• Adaptable for future applications across different economies