Investigating the Impact of Energy Use on Carbon Emissions: Evidence from a Non-parametric Panel Data Approach

Existing evidence for the role of energy consumption as a significant determinant of carbon emissions is very well studied. However, there are only a handful of studies in the South Asian context which are specific to individual country data and the use of parametric models. Thus, the impact of energy use on carbon emissions is investigated in this study by using a non-parametric panel data approach for five South Asian countries. The non-parametric approach discovers the true relationship from the data itself rather than having any pre-defined functional form. The data in this study was compiled for five South Asian countries—India, Pakistan, Bangladesh, Nepal, and Sri Lanka using the World Development Indicators database. The study period in this study is 1978 to 2011. The estimated impact of energy consumption on carbon emissions is found to be linear and positive in nature. More specifically, there is a monotonically increasing impact of energy use on carbon emissions. However, the estimated impact of income is found to be nonlinear in nature and it shows an N-shaped type relationship. The study also investigates the impact of some other contributing factors such as trade and population density. The results indicate that in the initial stages of globalization, more openness in trade has no significant impact but in the later stages there is a positive impact. However, there is no empirical evidence that population density is a determining factor for carbon emissions.