Heteroscedasticity-Robust Quantile Regression Decomposition for Cross-Country Fiscal Multiplier Estimation Under Structural Break Regimes

Abstract

Accurate estimation of fiscal multipliers remains a fundamental challenge in macroeconomic policy calibration, particularly across heterogeneous sovereign economies subject to asymmetric structural breaks. This study develops a heteroscedasticity-robust quantile regression decomposition framework integrating Markov-switching error correction mechanisms to isolate regime-contingent fiscal transmission channels across 47 OECD and emerging-market economies over the period 1995–2023. Employing panel-corrected standard errors alongside penalized B-spline smoothing for nonlinear trend extraction, our methodology yields multiplier estimates that are both distributional-tail-sensitive and structurally invariant. Empirical findings confirm significant multiplier heterogeneity conditioned on debt-to-GDP thresholds, monetary accommodation regimes, and output gap polarity. The framework substantially outperforms conventional OLS-based impulse response approaches in out-of-sample fiscal forecast accuracy, offering a replicable toolkit for evidence-based consolidation strategy design among policymakers navigating post-pandemic fiscal normalization.