مصرف انرژی و رشد اقتصادی برای کشورهای منتخب OECD: شواهد بیشتر از آزمون علیت گرنجر در حوزه فرکانس

This paper aims to reexamine the causal relationship between energy consumption and economic growth for 20 OECD countries. To that end, we employ a Granger causality test in the frequency domain which allows us to distinguish short (temporary) and long-run (permanent) causality. The empirical results could be summarized as following. First, in terms of causality running from GDP to energy consumption, there is a temporary relationship for Australia, Austria, Canada, Italy, Japan, Mexico, the Netherlands, Portugal, the UK, the USA, and a permanent relationship for Austria, Belgium, Denmark, Germany, Italy, Japan, the Netherlands, Norway, and the USA. Second, in terms of causality running from energy consumption to GDP, there is a temporary relationship for Austria, Denmark, Italy, the Netherlands, Norway and Portugal, and a permanent relationship for Belgium, Finland, Greece, Italy, Japan, and Portugal. The main implication of our finding is that the energy policies should take into consideration not only the causality direction between economic growth and energy consumption but also whether it is temporal or permanent and furthermore authorities must design policy actions accordingly.

The two energy crises of the 1970s, which hampered economic growth due to the high energy prices, required governments to implement conservative energy policies. However, reduction in energy consumption may have impacts on economic growth, if energy consumption causes economic growth, then reducing energy consumption could lead to a fall in output and employment. On the other hand, if the causality runs in the opposite direction, then energy conservation policies may be implemented without detrimental effects on output and employment. The current debate about global warming and climate change requires policy makers to take some precautions against the high level of greenhouse gas emissions, such as reducing fossil energy consumption and increasing usage of renewable energy. Some industrialized countries committed themselves to reducing greenhouse gas emission by restricting fossil fuel consumption due to the Kyoto Protocol. Nevertheless, it is argued that decreasing energy consumption may reduce economic growth and increase unemployment since energy is considered an essential factor of production (Stern, 2000). Hence, an examination of the relation between energy consumption and economic growth not only helps to understand the role of energy consumption in sustainable economic growth but also sets a framework for discussion of energy and environmental policies. Consequently, it is important to reveal the direction of causality between these variables. The literature suggests four possible connections between energy consumption and economic growth (Squalli, 2007). The growth hypothesis implies that energy consumption contributes to economic growth both directly and indirectly as a compliment factor to labor and capital in the production process. In addition, it claims that energy conservation policies could make the real GDP to reduce. On the other hand, the conservation hypothesis implies that energy conservation policies have no adverse effects on real GDP. While uni-directional causality running from real GDP to energy consumption supports the conservation hypothesis, the reverse direction of causality supports the growth hypothesis. The feedback hypothesis suggests that there is an interdependency causal link between energy consumption and real GDP and is valid if there is a bi-directional causality. If this is the case, an energy policy aiming at improvements in energy consumption efficiency will not adversely affect real GDP. Finally, the neutrality hypothesis suggests that energy consumption has a little or no impact on real GDP and therefore conservative energy policies will not reduce real GDP. The neutrality hypothesis is valid if there is no causality between energy consumption and economic growth. If energy consumption has no significant effect on the real output, then it would be reasonable for governments to adopt conservative and environment friendly policies. Energy conservation policies aimed at reducing energy consumption must create some methods to reduce consumer demand, such as an appropriate combination of energy taxes and subsidies, to eliminate or at least to mitigate unfavorable effects on economic growth. Policy makers should also encourage industries to adopt technology that reduces pollution. This paper contributes to the energy-growth literature by employing the Granger causality in the frequency domain to analyze both short- and long-run causality and aims to indicate whether there is a change in causality direction over time. To the best of our knowledge, this is the first study investigating causality between these two variables in high and low frequencies. Our analysis of the relationship between energy consumption and economic activity is based on a sample of 20 OECD countries. The remainder of the paper is organized as follows: Section 2 introduces the selected literature, Section 3 presents econometric methodology with empirical results in Section 4, and Section 5 presents our conclusions.

In this paper, we have examined the causal relationship between energy consumption and real GDP for OECD countries. Because energy is one of the basic factors for economic development, it has an important role in economic activities. Energy policies for many countries have more or less been affected by the Kyoto Protocol which aims at reducing emissions to mitigate climate change. However, the target of higher economic development could require energy consumption to increase. Hence, the effects of changes in energy policies on economic growth require more careful scrutiny. The results of traditional Granger causality tests are presented on the third and fourth columns of Table 1. The growth hypothesis is supported for Belgium and Japan, the conservation hypothesis is supported for Australia, Austria, Germany, Italy, the Netherlands, and Portugal, and the neutrality hypothesis is valid for Canada, Denmark, Finland, France, Greece, Mexico, Norway, Spain, Sweden, Turkey, the UK, the USA. Since traditional causality test does not consider time varying nature of the relationship, we perform the frequency causality test and our main findings are as follows. First, Belgium, Italy, and Japan have long term bi-directional causality between energy consumption and economic growth as suggested by the feedback hypothesis. Policy makers in these countries should take into account the feedback effect of real GDP on energy consumption by implementing regulations to reduce energy usage. In addition, energy policies aimed at improvements in energy consumption efficiency will not adversely affect the real GDP in the long run. Second, the neutrality hypothesis is valid for Australia, Canada, France, Mexico, Spain, Sweden, Turkey, and the UK in the long run. Since the neutrality hypothesis indicates that reducing energy consumption does not affect economic growth, conservative energy policies will not negatively affect real GDPs of these countries. Third, the growth hypothesis is suggested for Finland, Greece in the long run, for Austria, Denmark, the Netherlands, Norway in the short run, and for Portugal both in short and long run. Fourth, the conservation hypothesis is suggested for, Denmark, Germany, Norway, the USA in the long run, for Australia, Canada, Mexico, Portugal, the UK in the short run and for Austria, the Netherlands, and the USA both in short and long run. The third and the fourth results indicate that the causality relationship changes over time. The growth hypothesis dominates for Portugal, Greece and Finland in the long run. Therefore, the economic activity depends on energy usage and increases in energy consumption may stimulate real GDP in these countries. The conservation hypothesis implying that reducing demand for energy can only be expected to have a small short-run or without detrimental effect on overall economic activity dominates for Austria, Norway, Denmark and the Netherlands in the long run. Furthermore, these possible small short-run detrimental effects may well be caused by adjustment costs to long run energy saving technologies being implemented. The neutrality hypothesis dominates for Australia, Canada, the UK in the long run. As a result, there is little or no impact of energy consumption on real GDP and decreases in energy consumption would not negatively affect real GDP for these countries. In this respect, when energy use is not a vital input in the production function, economic policies aiming to increase economic growth should focus on capital, labor and productivity inputs. Our results have several implications both for energy economists and policy authorities. Energy economists must take account of possible changing causalities in examining between these variables. In particular, conventional Granger causality test could be inappropriate to examine dynamic relationship between energy consumption and real GDP. Policy authorities must take account of changing causalities and design policy actions accordingly. In this regard, the energy policies should take into consideration not only the causality direction between economic growth and energy consumption but also whether the direction is temporal or permanent, in other words whether it changes due to the time dimension.