معافیت های ساختاری، مصرف انرژی و بازبینی رشد اقتصادی : شواهد از تایوان

This paper studies the stability between energy consumption and GDP for Taiwan during 1954–2003. We use aggregate as well as various disaggregate data of energy consumption, including coal, oil, gas, and electricity, to employ the unit root tests and the cointegration tests allowing for structural breaks. Our main findings are: First, though gas consumption seems to have structural breaks in the 1960s, after considering the structural breaks, the series is a stationary variable when Taiwan adopted its expansionary export trade policy. Second, we find that different directions of causality exist between GDP and various kinds of energy consumption. The empirical result shows unanimously in the long run that energy acts as an engine of economic growth, and that energy conservation may harm economic growth. Third, the cointegration between energy consumption and GDP is unstable, and some economic events may affect the stability. Overall, we do find the structural breakpoints, and they look to match clearly with the corresponding critical economic incidents.

Many production and consumption activities involve energy as a required input, making it a key source of economic growth. At the same time, economic growth may induce the use of more energy. However, the aspect of owing to structural breaks is a common problem in macroeconomic series as they are usually affected by exogenous shocks or regime changes in economic events. Therefore, this relationship is likely to be subject to variation as a result of changes in the economy's structure like changes in energy policy or economic development regime, reforms in energy regulation, or institutional developments. The direction, strength, and stability of the relationship between energy consumption and GDP have occupied central importance in the conduct of energy policy. For these reasons, this paper fills the gap in the empirical literature on the stability between energy consumption and GDP, by studying the situation of Taiwan. In the course of its whole economic development, within the period 1954–2003, Taiwan has successively experienced quite a few momentous economic incidents, including adopting an expansion export trade policy in the 1960s, and encouraging exporting and expanding into the international market, which established the foundation for its economic development taking off later on. The first oil crisis in the 1970s caused international raw materials and the price of petroleum to soar. The second oil crisis in the 1979 caused an international imbalance of supply and demand of crude oil. Taiwan then carried out financial liberalization and an internationalization policy in the middle part of the 1980s, which liberalized finance controls by a wide margin. Economic incidents in the 1990s included the military crisis across the Taiwan Strait due to the threat from China, the Asian financial crisis (1997–1999), and oil countries' agreement to drop their production to increase oil prices. Although economic structures and economic development stages are not the same for various countries, most empirical research studies show that energy consumption changes have a high correlation with economic growth. However, the information content of energy consumption and GDP is subject to change over time and so far there has been no empirical work for the possible variations and instability in Taiwan. Especially over the past 50 years, Taiwan has developed its economic development miracle with a steady and abundant energy supply playing an important role. Previous works for discussing the relationships between various kinds of energy consumption and economic growth in Taiwan are started by Yang (2000) and show that there are different directions of causes that exist between GDP and various kinds of energy consumption for the period 1954–1997. He finds evidence of bi-directional causality between energy consumption and GDP. Differently, Cheng and Lai (1997) report the unidirectional causality running only from GDP to energy consumption for the period 1955–1993. Hence, like Yang's (1999) revelation that there is a unidirectional causality from GDP to coal consumption with no repercussions, but Hwang and Gum (1992) indicate a bi-directional causality is observed in Taiwan for the period 1955–1993. Based on these mixed or conflicting results, we categorize it as follows: If there does exist a causality running from energy consumption to GDP, then this denotes an energy-dependent economy such that energy is an impetus for GDP (Kraft and Kraft, 1978 and Shiu and Lam, 2004). On the other hand, if there is a reverse chain of causality from GDP to energy, then this denotes a less energy-dependent economy such that energy conservation policies may be implemented with little adverse or no effect on GDP (Oh and Lee, 2004). Finally, the finding of no causality in either direction, the so-called ‘neutrality hypothesis’ (Yu and Choi, 1985 and Altinay and Karagol, 2004), means that energy conservation policies do not affect GDP.1 The instability of the economic system may unfortunately in fact be reflected in the parameters of the estimated models that, when used for inference or forecasting, can induce misleading results. In contrast to previous studies, our study carefully analyzes the specification of the model and draws a conclusion by taking into account possible structural breaks in the model. However, there are three common questions on the Taiwan case in the above-mentioned literature. First, analyzed by way of vector autoregression (henceforth VAR), the causality between economic variables may omit long-term information that is received. Moreover, the VAR method may be improper in the presence of cointegration. Previous studies that have applied this development include Masih and Masih (1996) and Asafu-Adjaye (2000). Second, when there are structural breaks, the various test statistics are biased towards the non-rejection of a unit root. Third, if we neglect the structural break problem, then we are unable to find out whether the parameters are unstable or not within each of the sub-periods, because the bias in parameters presents conventional unit root tests toward accepting the null hypothesis of a unit root, even though the series is stationary. Why do we need to take account of structural breaks in energy consumption series? We respond to this question in that the energy market always has allied relationships with an economic system and is vitally correlated with the economic system.2 Thus, under the impact of external factors is the ineluctable possibility that we should pay attention to it.3 Looking back at past history, the two energy crises had the most violent exogenous shock to the world's energy consumption market, which impacted the economic development of almost all countries by pushing them to sink into a recession, causing each government to adopt an energy conservation policy. However, the change regime of an energy administrative policy means having a heavy impact on the habit of energy use for consumers, implying that changes in energy consumption contribute to one of the factors that cause structural break in the relationship between energy consumption and economic growth. Even Crompton and Wu (2005) also forecast the slower growth of energy consumption in 2010 that reflects the expected slower economic growth because of the structural breaks in the Chinese economy. Therefore, this structural break due to the existence of an energy consumption threshold should be taken into account when constructing estimation and prediction models of economic growth for researchers or policymakers. Fortunately, we also find a few proofs to support the above inferences. For example, Hamilton (2003) reports that oil price increases are much more influential than oil price decreases, creating an asymmetric relation between crude prices and economic output. This suggests that non-linear relationship linkages between oil consumption and economic growth should be revealed. In addition, Hooker (2002) also provides strong evidence of a structural break, with oil price changes making a substantial direct contribution on inflation, and the structural break specification provides a better fit to the data. Hence, Moral-Carcedoa and Viceńs-Oterob (in press) find that both increases and decreases of temperature, linked to the passing of certain threshold temperatures, increase the demand for electricity. The principal studies indicate that the relationship between electricity demand and temperature is clearly non-linear. Based on these previous contributions, we know that once we deal with the data of energy consumption, we should take into account that the series has the characteristic of non-linear adjustment. Continuously, several renowned empirical studies recently have found that the series are actually trend stationary with a structural break, such as Altinay and Karagol (2004) who offer time series data of Turkey's energy consumption and GDP. Schäfer (2005) discusses structural break in final energy use based upon the historical development of the entire energy system in 11 world regions from 1971–1998. He indicates that a structural break in the economy, via the conduction of the middle sector, such as agriculture, industry, and services, will cause similar sector shifts in final energy use. From the above-mentioned previous wisdom, we acquire the essential result that an energy consumption series probably has the phenomenon of structural breaks. Overall, the main purpose and contribution of this paper are as follows. In the beginning, we test Taiwanese data for the last 50 years, which may be affected greatly by economic incidents or energy policy alterations. This may yield structural changes from the data of GDP or energy consumption, which therefore influence the result of the stationary test. Secondly, we use Johansen's (1988) multivariate cointegrated method to test the coinegration relation between GDP and energy consumption. We measure it with aggregate and several disaggregate categories of energy consumption. Furthermore, based on the multivariate cointegration model, we probe into the causality between GDP and energy consumption with the weak exogeneity test. Finally, it is possible that structure breaks have occurred which might affect the result of the cointegration test. Using two kinds of Hansen (1992) and Gregory and Hansen (1996) structural break tests, we re-examine the cointegration relationship altogether and whether there is instability or not between GDP and energy consumption. The rest of the paper is organized as follows. The next section presents the data used. The third section discusses the econometric method, and the empirical findings of the study are reported. The fourth section concludes with remarks of the study.

This paper studies the stability between energy consumption and GDP for the period 1954–2003 in Taiwan. We use aggregate as well as several disaggregate categories of energy consumption, including coal, oil, gas, and electricity, to employ the unit root tests and the cointegration tests allowing structural breaks. The main purpose of this paper is that we test the data of Taiwan, which as an economy has been affected by critical economic incidents or energy policy alterations that led to the data of GDP or energy consumption yielding structural breaks. Some conclusions have been drawn from our empirical tests. First, considering the structural breaks, gas consumption is a stationary variable. The structure change took place in the 1960s, when Taiwan adopted an expansionary export trade policy. The above-mentioned results are inconsistent with Yang (2000). The reason lies in that Yang (2000) did not consider the influence of the structural break, and so the outcome might be distorted. We next find different directions of causes that exist between GDP and various kinds of energy consumption. This indicates that bi-directional causal linkages between GDP and both total energy and coal consumption are identified. However, there is a unidirectional causality running from oil consumption to GDP. Furthermore, there is a unidirectional causality running from gas consumption and electricity consumption to GDP that is detected for these cases. Consequently, energy acts as an engine of economic growth. Third, relative to Yang (2000), this paper finds that except for total energy and gas consumption being the same as with the causality of GDP for Yang (2000), other empirical results are different. Aside from this, the results of the causality test under the different energy case conditions emphasize unanimously that an energy conservation policy will hurt the economic growth of Taiwan. These empirical findings have important implications for Taiwan. From the present time and into the future, energy looks to act as an engine of economic growth for the island, and so the neutrality hypothesis of energy consumption and economic growth is not supported in Taiwan. Fourth, the Hansen (1992) test shows that the relationship between energy consumption and GDP may indeed be unstable. At the same time, the GH (1996) test does suggest that a structural break in the cointegration vector is important and needs to be taken care of in the specification of energy consumption and GDP. We find that the breakpoints in various GDP–ENERGY consumptions match compatibly with critical economic incidents of Taiwan. While analyzing the relation of energy consumption and GDP of Taiwan, one should surely include a structure break into the question.