Gone with the cycle: The asymmetric impact of business cycle on growth

This paper investigates the asymmetric effects of business cycle volatility on economic growth in a panel of 99 countries over 1951–2019. Within a β-convergence framework, we decompose cyclical dynamics into three components: position relative to trend, amplitude of deviations, and phase. Three key findings emerge. First, we find robust evidence of β-convergence, which is stronger in per-capita terms than in aggregate output, showing that demographic dynamics partly mask the pace of productivity-driven catch-up. Second, large deviations from trend carry a significant growth penalty, pointing to stable slope asymmetry across specifications. Third, phase effects are not statistically significant, indicating that expansions and recessions do not produce structural jumps in average growth. A Gaussian-mixture clustering of volatility regimes further reveals that pooled regressions are dominated by low-volatility observations, while in high-volatility environments convergence is sharper and volatility costs are more severe. Overall, the evidence supports models in which volatility magnifies effective risk and depresses investment, and highlights the need for stabilization strategies that are both counter-cyclical and volatility-sensitive.