رشد اقتصادی، حفاظت از انرژی و کاهش انتشار گازهای گلخانه ای: تجزیه و تحلیل مقایسه ای بر اساس داده های پانل برای 8 کشور آسیایی و اقیانوس آرام

This study was conducted to evaluate the causality between energy consumption, GDP growth and carbon emissions for eight Asia-Pacific countries from 1971 to 2005 using the panel data. The results indicate that there are long-run equilibrium relationships between these variables. Additionally, causality from energy consumption to CO2 emissions was observed generally, but there were some opposite relationships also. Parameter estimations of the panel data model indicate that there are great differences in the carbon emissions, the efficiencies of energy use, carbon emissions of unit GDP and unit energy consumption between developed and developing countries. The base carbon emissions, per capita energy consumption and efficiency of energy use in developing countries are far lower than in developed countries; however, the CO2 emissions per unit of energy use is higher. Although developing countries may reduce their CO2 emission per unit energy use, total energy consumption will rise rapidly with economic development. Thus, developing countries must determine how to undergo economic growth while conserving energy and reducing emissions. To respond to global climate change, it is necessary to develop innovative technology for energy use, transform the energy structure and conduct the clean development mechanism.

Global climate warming has become a serious threat to human survival and health. In addition to natural factors, global warming is closely related to CO2 emissions produced by human activities (Soytas and Sari, 2009). Developing appropriate responses to climate change has become the most important environmental problem faced by the international community. The purpose of the Kyoto Protocol signed in 1997 was to restrict greenhouse gas (GHG) emissions in developed countries, and developing countries were asked to undertake the obligations of emission reduction at the Copenhagen Climate Conference in 2009. As a result, the causal relationships between energy consumption, economic growth and carbon emissions have become an international study topic in recent years. The question of whether developing countries should or should not be burdened with emissions reduction in international climate negotiations makes this issue an important aspect impacting international relationships. Global climate change is largely caused by massive GHG emissions that have been produced by developed countries since the industrial revolution, and these countries have the capacity to prevent global warming as well. Based on these facts, it is reasonable for developing countries to require developed nations to reduce GHG emissions gradually based on their current emissions while allowing more carbon emissions by the developing nations. However, developed countries also want to be included in a new round of climate negotiations with developing nations. For China, the largest developing country, although its CO2 emission per capita is only No. 73 worldwide, its total CO2 emissions are the second highest in the world (Haakon et al., 2009). Thus, China faces the dilemma of development and emission reduction. Specifically, China requires a larger carbon emissions space to enhance its industrialization and urbanization, but it must also reduce its GHG emissions. There are different causal links among the energy consumption, economic growth and carbon emissions at different stages of economic development in different countries (Dinda and Coondoo, 2006, Soytas and Sari, 2009 and Huang et al., 2008). Comparing the trends in the efficiency of energy use and carbon emissions per unit GDP change between developed countries and developing countries, we can demonstrate the importance of technological progress in energy saving and carbon emissions, thereby enabling useful policy decisions. This will be of considerable importance for the coordination of national positions on climate change. A review of previously conducted studies revealed that there were complex causal relationships between economic growth and energy consumption. Specifically, short and long term unidirectional links from economic growth to energy consumption or the reverse and bi-directional relationships between variables were observed. Indeed, some studies have shown very different results when data from different groups of countries for the same period were analyzed using the panel Granger test. For example, Narayan and Smyth (2008) revealed long run causality between energy use and income in G-7 countries and found that causality appears to run both ways in four countries (Canada, Italy, Japan and UK), from energy use to income in two countries (US and France), and from income to energy consumption in Germany. Lee (2006) analyzed the causality between energy consumption and income in 11 major industrialized countries and discovered bi-directional causality in the United States, but unidirectional causality running from energy consumption to GDP in Canada, Belgium, the Netherlands and Switzerland. However, this relationship was the opposite in France, Italy and Japan. Moreover, a strong causality from economic growth to energy consumption was observed in 11 oil exporting countries (Mehrara, 2007), but energy consumption and income showed bi-directional causal linkages in 22 OECD countries (Lee et al., 2008). Finally, several studies of Asian countries revealed that energy consumption has a positive impact on economic growth (Lee et al., 2008; Fatai and Oxley, 2004; Wang and Liu, 2007), but there were opposite and bi-directional conclusions (Fatai and Oxley, 2004, Zhao, 2007 and Oh and Lee, 2004). In addition, some empirical studies failed to achieve unanimous conclusions, even if they were conducted in the same country within roughly the same period. For instance, the pioneering work conducted by Kraft and Kraft (1978) demonstrated the existence of Granger causality running from income to energy use for the United States using data covering the period of 1947–1974. However, Stern (1993) observed reverse causality using data for the period of 1947–1990. Moreover, Ang (2007) examined the dynamic causal relationships between two variables for France using panel a vector error correction model (VECM) and found that economic growth exerts a causal influence on energy use in the long run, and energy use points to output growth in the short run. Furthermore, the results obtained using panel data from many countries over various time periods also varied (Mahadevan and Asafu-Adjaye, 2007 and Huang et al., 2008). However, a bi-directional relationship between economic growth and CO2 emissions (Huang et al., 2008 and Soytas et al., 2007), and a unidirectional causality running from energy consumption to carbon emissions (Dinda and Coondoo, 2006, Soytas et al., 2007 and Halicioglu, 2009), has commonly been observed among studies (Table 1). There is a complex nexus between GDP-energy consumption and CO2 emissions. The results obtained by Dinda and Coondoo (2006) suggest that there is more or less a bi-directional causal relationship between per capita GDP and per capita CO2 emission for most countries. Accordingly, the movement of one variable directly affects the other variable through a feedback system. Furthermore, Pao and Tsai(2010) investigated CO2 emission, energy consumption and economic growth in BRIC countries(Brazil, Russian, India and China); their results suggest that emissions strongly Granger-cause both energy consumption and output.Some scholars have evaluated the links between type of energy and economic growth. Yuan et al. (2008) investigated the China case employing a VEC specification and found that Granger causality from electricity and oil consumption to GDP exists, but not from coal. Erol and Yu (1987) also found a unidirectional link from electricity consumption to income in Japan. Ciarreta and Zarraga (2010) investigated the relationship between electricity consumption and actual GDP for a set of 12 European countries and found evidence of a long-run equilibrium relationship between the three series and a negative short-run and strong causality from electricity consumption to GDP. Wolde-Rufael and Menyah (2010) evaluated the causal relationship between the nuclear energy consumption and the actual GDP for nine developed countries. They found a unidirectional causality running from nuclear energy consumption to economic growth in Japan, the Netherlands and Switzerland, the opposite unidirectional causality in Canada and Sweden and a bi-directional causality between economic growth and nuclear energy consumption in France, Spain, the United Kingdom and the United States. Huang et al. (2008) reviewed many previously conducted studies and concluded that the differences in causal relationships obtained using data from the same country could be different due, in part, to differences in research periods or in research methodologies. The most probable reason for these discrepancies is the insufficient number of observed data points. This results from data evaluated in previously conducted studies spanning 30–40 years. Costantini and Martini (2009) felt that the empirical findings describing the causal relationship between energy consumption and economic growth were mixed depending on the functional form adopted, the econometric approach used, the time periods and the sample of countries analyzed. In addition to the above reasons, we believe that the difference in economic structures and energy structure between objective countries also contributes to these mixed results. Based on the methodology used, Costantini and Martini (2009) divided the available literature regarding the relationship between energy use and economic growth into four generations. First-generation studies included Granger causality testing (1969) and cointegration analysis (Sims, 1972 and Granger, 1969), while second generation studies were based on the Granger two-stage procedure (Granger, 1988). These studies evaluated pairs of variables for cointegrating relationships and used estimated error correction models (ECM) to test for Granger causality. Third-generation literature used multivariate estimators (Johansen, 1991) and fourth-generation studies employed recently developed panel methods to test for unit roots, cointegration and Granger causality (Al-Iriani, 2006, Lee and Chang, 2008 and Mahadevan and Asafu-Adjaye, 2007). In the present study, we investigated the causal relationship and interaction between energy consumption, economic growth and carbon emissions for eight Asia-Pacific countries that were divided into two groups (developing countries and developed countries). Furthermore, we compared differences among eight countries in primary energy structure and briefly analyzed the influence of this structure on CO2 emissions. We employed relatively new panel methods to test for unit root, cointegration and Granger causality. Based on the results, we established panel regression models to further estimate the characteristics and changing trends of energy consumption and carbon emissions during the economic growth process, and to compare the differences between developed countries and developing countries. The purpose of this study was to deepen the understanding of the issues that we are facing and to provide a reference for developing countries to develop a low carbon economy. This paper is structured as follows. Section 2 provides a brief introduction to the countries referred to in the study. Section 3 briefly describes the methodology used. Section 4 discusses the data and empirical results. Section 5 provides the study conclusions and policy implications.

The four developed countries have the advantage of technology, therefore, their energy efficiency is greater than that of the four developing countries (Table 2). This means that developing countries could improve their energy efficiency through technological progress, which would enable them to reduce their carbon emissions while they achieve economic growth. Conversely, the per capita energy consumption and per capita CO2 emissions in developed countries are far greater than in developing countries, indicating that carbon emissions in the developing countries will inevitably increase if they wish to achieve the same economic level as developed countries. Overall, developing countries can obtain great benefits with respect to energy use and carbon emissions by learning from developed countries. The panel data model test revealed that there are long-term equilibrium relationships between energy consumption, GDP growth and CO2 emissions for the eight Asia-Pacific countries. Causality from energy consumption to CO2 emissions exists generally, and these findings are particularly evident for developed countries, but there were some opposite relationships also. GDP is responsible for the increase in energy consumption, and there is reverse correlation in developed countries, but there is no reverse correlation in developing nations. There is strong causality between GDP and CO2 emissions over the long run in developed countries, but no such relationship is present in developing countries. However, there is a complex nexus between GDP-energy consumption and CO2 emissions; these causal relationships need to be researched further. The parameter estimations of the panel data model indicate that there are great differences between individual fixed effects and individual variable coefficients for developed and developing countries. These differences represent the bases of carbon emissions, efficiencies of energy use, carbon emissions of unit GDP and unit energy consumption of the two types of countries. In developed countries, the base of carbon emissions is on the higher level, as are the per capita energy consumption and carbon emissions. However, the unit energy use may lead to a greater GDP and produce less carbon emissions because of the high efficiency of energy use. The status of developing countries is quite different. Among them, the highest contribution of unit energy consumption to GDP growth was 1.77 for China. Additionally, the CO2 emission per unit of GDP is 1.87 in China, which was higher than the other developed countries and lower than the developing countries evaluated in the present study. Although developing countries may greatly reduce their CO2 emission per unit of energy use, their total energy consumption will still rise rapidly with economic development. Thus, developing countries must determine how to maintain economic growth while conserving energy and reducing CO2 emissions. This difference is also affected by the energy structure. In developing countries such as China and India, coal accounts for a large proportion of energy consumption, and CO2 emissions per unit of energy consumption are high. Among developed countries, Australia consumes an enormous amount of coal and therefore has high CO2 emissions per unit of energy consumption (Fig. 1). The trend in the time effect is very clear for the eight countries evaluated in this study. The time varying coefficient βtβt increases annually, indicating that technological progress improves the GDP output of unit energy consumption. However, the time fixed effect λtλt also increases annually, which reflects an increase in per capita CO2 emissions. In general, the excessive growth of total energy consumption results in it being difficult to change the increase in per capita CO2 emission by technological progress. Therefore, changes such as transforming the energy structure, improving energy utilization efficiency and controlling the total energy must be implemented to decrease global warming.