Revisiting the environmental Phillips curve hypothesis: a nonlinear panel data approach

This study re-examines the EPC hypothesis between unemployment and carbon emissions across 1991-2021, using a novel of pooled mean group nonlinear autoregressive distributed lag model and panel quantile regression. Diverging from prior linear assumptions, we extend the examination on the nonlinear nexus by incorporating the asymmetries of unemployment in the estimation. Results are compared across emission and unemployment extremes. Results reveal no short-term EPC validity but uncover a complex long-term nexus: rising unemployment reduces emissions in high-unemployment economies, yet increases them in low-emission countries. Conversely, falling unemployment drives emission growth in carbon-intensive nations, reflecting rebound effects from industrial recovery.