<p>Government spending plays a vital role in driving economic growth, particularly in developing economies like Ethiopia, where public investment serves as a key policy instrument. This study addresses a significant gap in the Ethiopian fiscal-growth literature, where most previous research assumes a linear relationship between government expenditure and economic growth, overlooking possible nonlinear and asymmetric dynamics. Using annual data from 1980 to 2024, the study applies the Nonlinear Autoregressive Distributed Lag (NARDL) model to capture both the magnitude and direction of fiscal shocks. By decomposing government expenditure into positive (expansionary) and negative (contractionary) changes, the study provides new empirical evidence on whether increases and decreases in spending influence growth differently. The results reveal that increases in government expenditure have no significant long-run effect on economic growth, while reductions in spending exert positive effects in both the short and long run. This implies that excessive or inefficient government spending may constrain growth, whereas disciplined and well-targeted fiscal consolidation enhances it. The findings also show that Wagner’s Law does not hold in the Ethiopian context, as economic growth does not Granger-cause government expenditure. Overall, the study contributes to the existing literature by introducing a nonlinear perspective to Ethiopia’s fiscal policy analysis and offering valuable insights for designing efficient, growth-oriented, and context-specific fiscal strategies.</p>