Environmental Kuznets curve and energy consumption in Malaysia: A cointegration approach

ABSTRACT

This study examines the long- and short-run relationship between carbon dioxide (CO₂) emissions, economic growth, energy consumption, foreign trade, and urbanization based on the Environmental Kuznets Curve (EKC) hypothesis in Malaysia for the period 1980–2008. This study employs the Autoregressive Distributed Lag (ARDL) methodology for empirical analysis. The results obtained suggest the existence of three forms of long-run relationships among variables when CO₂ emissions, foreign trade, and urbanization are the dependent variables. The long-run results support the EKC hypothesis, which assumes an inverted U-shaped relationship between income and environmental degradation. Moreover, both energy consumption and urbanization are found to have positive effects on CO₂ emissions in the long run, while trade has a negative effect on CO₂ emissions in both short and long runs. The stability of the variables in the estimated model is also examined. Results suggest that the estimated model is stable over the study period.

KEYWORDS:

• CO₂ emissions
• environmental Kuznets curve
• energy consumption

Notes

¹ For more details about bounds testing approach advantages see Halicioglu (2009) and Jalil and Mahmud (2009)

² Since the Pesaran et al. (2001) bounds testing approach is applicable for the variables that are I(o) or I(1), the analysis begins by investigating the unit root test of variables using the augmented Dickey–Fuller (1979) ADF and Phillips-Perron (1988) PP tests. The results indicate that all variables are not integrated of order two and above, hence validating the use of bounds testing for cointegration.

³ See Bahmani-Oskooee and Brooks (1999) and Narayan et al. (2oo8).

⁴ Two is the maximum number of lags that can be imposed in this study given the number of variables and the sample size.

⁵ Bahmani-Oskooee and Kantipong (2001) argue that the evidence of no−cointegration in this stage may be attributed tothe fact that the same numbers of lags are imposed on each of the first-differenced variables arbitrarily.

⁶ Since the primary objective of this study is to investigate the presence of EKC, we estimate the equation in which carbon dioxide emissions is the dependent variable.