مصرف انرژی هسته ای و رشد اقتصادی در کشورهای OECD: تحلیل علیتی ناهمگن پانل مقطعی وابسته

The purpose of this study is to determine the direction causality between nuclear energy consumption and economic growth in OECD countries. The empirical model that includes capital and labor force as the control variables is estimated for the panel of fourteen OECD countries during the period 1980–2007. Apart from the previous studies in the nuclear energy consumption and economic growth relationship, this study utilizes the novel panel causality approach, which allows both cross-sectional dependency and heterogeneity across countries. The findings show that there is no causality between nuclear energy consumption and economic growth in eleven out of fourteen cases, supporting the neutrality hypothesis. As a sensitivity analysis, we also conduct Toda–Yamamoto time series causality method and find out that the results from the panel causality analysis are slightly different than those from the time-series causality analysis. Thereby, we can conclude that the choice of statistical tools in analyzing the nature of causality between nuclear energy consumption and economic growth may play a key role for policy implications.

Determining the direction of causality between energy consumption and economic growth provides important inferences in establishing sound energy policies. A vast literature thereby has been documented on casual relationships between economic growth and consumption of energy sources (electricity, coal, natural gas, and oil) in last two decades.1 The empirical literature has now been focusing on examining the nature of causality between nuclear energy consumption and economic growth due to the fact that nuclear energy is an important source for increasing diversity of energy supplies, for improving energy security, and for providing a low-carbon alternative to fossil fuels (Wolde-Rufael and Menyah, 2010). The importance of nuclear energy has been increasing as a result of its advantages by producing heat and electricity without emitting carbon-dioxide into the atmosphere at the power plant level (OECD, 2005). Increasing importance of nuclear energy leads researches to question to what extent nuclear energy consumption affects economic growth. Investigating the relationship between nuclear energy consumption and economic growth is insightful for understanding the logical reason of investing in nuclear energy for economical concern or for environmental and social concerns. This study aims at determining the direction causality between nuclear energy consumption and economic growth by means of panel causality analysis, which allows us to account for both cross-sectional dependency and cross country heterogeneity. In the empirical analysis, we concentrate on OECD countries that seventeen of thirty three OECD countries have nuclear energy capacity, which produce 85% of world's total nuclear energy production. The empirical model is augmented with the control variables – capital and labor force – in order to avoid omitted variable bias due to the point of view that nuclear energy consumption might not alone stimulate economic growth (Wolde-Rufael and Menyah, 2010). The causality analysis that is applied to the panel of fourteen OECD countries for the period 1980–2007 finds out that neither nuclear energy nor economic growth is the cause of each other in eleven out of fourteen cases, supporting the neutrality hypothesis. As a sensitivity analysis, we compare the results from the panel causality analysis with those from Toda–Yamamoto causality approach, which is not able to take into account for cross-sectional dependency and obtain that the results from the panel causality test are different than those from the time-series causality test. In detecting the existence of causality, we rely upon a panel causality approach instead of time series methods since panel analysis produces more reliable and statistically powerful results than time series analysis by combining information from both cross-section and time dimensions. Unlike the previous panel data studies on the nuclear energy–economic growth nexus, we first test for cross-sectional dependency and heterogeneity across countries due to the fact that ignoring cross-sectional dependency and country specific heterogeneity in a panel causality analysis are potential sources of misleading inferences regarding the direction of causality. Even though time series approaches to causality are able to modeling cross-county heterogeneity, they are not able to take into account cross-sectional dependency across countries. It is the well-known fact that OECD countries are highly integrated and thereby a shock in one country is easily transmitted to other countries through international economic interrelationships. By accounting for cross-sectional dependency across countries, this study differs from the previous studies in the literature and thereby is novel to the nuclear energy–economic growth nexus. We organize the rest of paper as follows. The next section summarizes the hypotheses related to energy consumption and economic growth as well as the empirical literature on the nuclear energy consumption and economic growth nexus. Section 3 is devoted to describe the data, to outline the econometric methods, and to interpret the empirical results. Finally, Section 4 provides a brief summary of the study and the concluding remarks.

This study determines the nature of causal relations between nuclear energy consumption and economic growth in OECD countries. A novel approach to panel Granger causality, which controls for cross-sectional dependency and heterogeneity across countries is applied to the panel of fourteen OECD countries during the period 1980–2007. Furthermore, we also carry out the Toda–Yamamoto causality approach, which is widely utilized by the recent studies to compare the results from the panel causality analysis with those from the time series approach. The panel Granger causality analysis supports the growth hypothesis for Hungary, the conservation hypothesis for UK and Spain, and the neutrality hypothesis for eleven countries. The results from the Toda–Yamamoto approach are slightly different than those from the panel causality method. The Toda–Yamamoto approach provides evidence on the growth hypothesis for Germany and Finland; the conservation hypothesis for Hungary, Japan, Korea, and Sweden; the feedback hypothesis in the UK and the USA; the neutrality hypothesis for other countries. The results from the panel causality analysis imply that the nature of causality between nuclear energy consumption and economic growth seems to be in favor of the neutrality hypothesis in OECD countries. The neutrality between energy consumption and economic growth can be attributed to the fact that energy consumption may be a relatively small component of overall production. Furthermore, the share and hence the contribution of nuclear energy may be even smaller. Thereby, the findings from this study imply that the energy conversation policies would have little or no impact on economic growth. By comparing the result from the panel causality approach with those from the times series method, this paper clearly shows that the choice of statistical techniques in analyzing the direction of causality between nuclear energy consumption and economic growth is crucial for policy implications. We can therefore conclude that the policy implications driven from the causality approach that takes into account cross-sectional dependency seem to be more appropriate to design sound energy policies. In the literature on the energy consumption-economic growth nexus, one of the recent tendencies is to concentrate on investigating causality between nuclear energy, renewable energy and economic growth. In that respect, the future studies may revisit the direction of causality within the context of panel data analysis by taking into account cross-sectional dependency. Second direction for the future studies is to question whether the direction causality between nuclear energy consumption and economic growth is characterized by asymmetric transmission, which provides a room to concentrate on nonlinear causality methods. Another question is that whether there is a threshold impact of nuclear energy consumption and economic growth that can be captured by both nonlinear time series and panel data methods.