Impact of Time Trend and Heterogeneity on EKC in Asian Countries

Introduction

The relationship between pollution and income has been studied for decades after the seminal work of Grossman & Krueger (1991) and Shafik & Bandyopadhyay (1992). Several empirical efforts investigate a specific conjecture, called the Environmental Kuznets Curve (hereafter EKC), which proposes a non-linear relationship thereby depicting an inverted ‘U’ shaped link between economic development and pollution. The analogy of EKC echoes the exact relationship between economic growth and income inequality which was provided by Kuznets (1955) in his 1954 presidential address to the American Economic Association. However, several studies observed that the relationship might be N-shaped, thereby implying that the environmental degradation will begin to increase further above a specific level of income¹. The conclusion on the shape of EKC is traditionally inferred from reduced-form models based on the signs of coefficient of GDP² and GDP³. These models are based on a prior assumption of the underlying distribution and functional form. Such an imposed assumption would lead to weak identification and biased turning points. As described by Panayotou (1997) estimation by these models is just like a black box. Besides, the cubic form of the model where the first-order condition is of quadratic nature comments on the N-shaped pattern without testing the real nature of the turnaround points. The study by Sinha et al. (2018) derives validity conditions to give a proper comment on the shape and turning points arising out of the model. Another aspect of the EKC model is the use of a wide range of explanatory variables other than income that is used in the investigation of EKC hypothesis and may create problems of multicollinearity. Moreover, some variables that change over time and cannot be observed directly also influence the development-environment relationship.

This study has been designed to address these limitations and investigate the relationship between economic development and environmental pollution for the Asian countries. Over the past few decades, the economies of Asia have experienced a spectacular growth. The presence of attributes that change over time, such as exogenous technical progress, individual awareness of the environment, and alterations in the production structure and input composition, all contribute to the growth process. Although the momentum of growth of the Asian countries are promising, the path ahead requires resolution of difficult issues and challenges to continue the growth momentum. One such risk is to tackle the issue of global warming which is mainly caused by greenhouse gas (hereafter, GHG) emissions. The economies of Asia are already the largest GHG emitter in the world (Dixon and Kwakwa,2016). Therefore, pollution is the by-product of the growth and development of Asia. Thus, it is meaningful to investigate the shape of EKC by emphasising the time related factors.

To the best of our knowledge, there is no attempt to test this validity condition in the case of Asian countries. Besides, a linear time trend variable is used in the framework of EKC and the use of a linear time trend variable in testing the validity condition is also not carried out in any other work. Assuming the development-environment relationship to be heterogenous, this study uses different income classifications of the overall group of Asian countries by controlling for linear time trends existing in each income group. The reason behind the inclusion of time trend as a variable in the EKC regression is to emphasize the need to find the non-income-related variables besides income, that are affecting pollution levels of the developing countries such as Asia. In this study, that it is likely the impact of linear time trend for different income groups is different is not addressed properly in the existing studies. This paper attempts to bridge up this gap. Accordingly, the expected GHG emissions are calculated for each income group and compare the development-environment path of each income group.

On this note, the specific objectives of this study are as follows: