Has the Paris Agreement Shaped Emission Trends? A Panel VECM Analysis of Energy, Growth, and CO$_2$ in 106 Middle-Income Countries

Abstract

Rising CO$_2$ emissions remain a critical global challenge, particularly in middle-income countries where economic growth drives environmental degradation. This study examines the long-run and short-run relationships between CO$_2$ emissions, energy use, GDP per capita, and population across 106 middle-income countries from 1980 to 2023. Using a Panel Vector Error Correction Model (VECM), we assess the impact of the Paris Agreement (2015) on emissions while conducting cointegration tests to confirm long-run equilibrium relationships. The findings reveal a strong long-run relationship among the variables, with energy use as the dominant driver of emissions, while GDP per capita has a moderate impact. However, the Paris Agreement has not significantly altered emissions trends in middle-income economies. Granger causality tests indicate that energy use strongly causes emissions, but GDP per capita and population do not exhibit significant short-run causal effects. Variance decomposition confirms that energy shocks have the most persistent effects, and impulse response functions (IRFs) show emissions trajectories are primarily shaped by economic activity rather than climate agreements. Robustness checks, including autocorrelation tests, polynomial root stability, and Yamagata-Pesaran slope homogeneity tests, validate model consistency. These results suggest that while global agreements set emissions reduction goals, their effectiveness remains limited without stronger national climate policies, sectoral energy reforms, and financial incentives for clean energy adoption to ensure sustainable economic growth.