Abstract
This paper attempts to find out the relationship of renewable and non-renewable energy utilization with CO2 emission in Bangladesh. The analysis is done using the Autoregressive Distributed Lag (ARDL), Dynamic Ordinary Least Squares (DOLS), Fully Modified OLS (FMOLS) and Canonical Cointegrating Regression (CCR) techniques to observe the robustness of the findings. Results postulate that non-renewable energy and GDP has statistically significant positive association with CO2 emission whereas, renewable energy is negatively associated. From all estimation techniques, it is found that 1% augment in non-renewable energy utilization will boost CO2 emission by on an average more than 0.75%. Moreover, Toda -Yamamoto (TY) and vector autoregression impulse response procedure have been employed and the findings confirm that non-renewable energy is causing CO2 emission but renewable energy is not causing CO2 emission. The paper adds value to the present energy-emission study in a way that in Bangladesh context there are paucity of studies that concentrate on the effect of both renewable and non-renewable energy utilization on CO2 emission and there exists no study with multiple estimation techniques with a view to get robust findings. Another contribution is the dynamic effects of impulse response function also complement the ARDL, DOLS, FMOLS and CCR findings. As findings are responsive to the methodology used, therefore, for robust results, multiple techniques are employed. This study recommends increasing the share of renewable energy will play a critical role to reduce CO2 emission consequently global warming.