Abstract
The purpose of this article is to empirically investigate the impact of economic growth, oil consumption, financial development, industrialization and trade openness on carbon dioxide (CO2) emissions, particularly in relation to major oil-consuming developing economies. This study utilizes annual data from 1980 to 2012 on a panel of 18 developing countries. Our empirical analysis employs robust panel cointegration tests and a vector error correction model (VECM) framework. The empirical results of three panel cointegration models suggest that there is a significant long-run equilibrium relationship among economic growth, oil consumption, financial development, industrialization, trade openness and CO2 emissions. Similarly, results from VECMs show that economic growth, oil consumption and industrialization have a short-run dynamic bidirectional feedback relationship with CO2 emissions. Long-run (error-correction term) bidirectional causalities are found among CO2 emissions, economic growth, oil consumption, financial development and trade openness. Our results confirm that economic growth and oil consumption have a significant impact on the CO2 emissions in developing economies. Hence, the findings of this study have important policy implications for mitigating CO2 emissions and offering sustainable economic development.
Notes
1 According to the EIA, the terms ‘oil’ and ‘petroleum’ are sometimes used interchangeably.
2 Due to space limitation, detailed equations and a detailed discussion of the panel cointegration models are excluded.
3 All developed countries (as defined by the World Bank) are excluded from the analysis, because developing countries often contribute higher CO2 emissions, which eventually degrade the environment. Hence, we aim to understand the major contributors of CO2 emissions in developing countries and provide potential policy implications for mitigating CO2 emissions and sustainable economic growth.
4 Growth rates are calculated using original data (before converting into natural logarithms).
5 Unconditional correlations are calculated using data in natural logarithms.