The determinants of aggregated electricity intensity in ChinaWord文件下载.docx

1、ABSTRACT This study provides technological change, factor demand and inter-factor substitutability measures for china electricity industry. We use individual fuel price data and translog factor cost function approach to estimate total factor cost functions and fuel share equations. The implied price

2、 and Morishima substitution elasticities for inter-factor are obtained. By estimating the substitution between factors, our purpose is to explore the driving forces to aggregate electricity intensity. The results suggest that energy is substitutable for capital and complementary for labor regarding

3、cross-price elasticity. Capital is Morishima substitution to energy and labor is complement to for energy. Factor substitution, budget and technology effects contribute to the year-to-year volatility of electricity intensity in the proportion 48%: 41.99%: 9.89%. Key words: China, Electricity intensi

4、ty, Factor substitution1. IntroductionWith prominent decrease of 69 percent of China energy intensity in 1985-2002 and a reversal since 2003, a vast majority of studies attempt to explain this phenomenon. The debate of the driving forces to the persistent decline of energy intensity has been revolvi

5、ng around structural shifts and efficiency improvements. Just as Crompton and Wu （2005） pointed out that Chinas declining energy intensity is often explained in terms of technological and structural factors. At the beginning of 1990s, researchers argued that the structural shifts away from more ener

6、gy-intensive industrial sub-sectors played the most important role in China （e.g., Smile, 1990; Kambara, 1992）. Since then, the literatures usually accepted the view that the continuous decline in energy intensity was mainly attributed to the enhanced efficiency of energy use of industry sectors, wh

7、ile the structural shifts within the manufacturing sub-sectors or from primary to secondary or tertiary industry play only a nominal role from 1980 to the early 1990s （Siton and Levine 1994; Zhang, 2003）. Garbaccio et al. （1999） even found that structural change actually increased energy intensity b

8、etween 1987 and 1992. Since the mid-1990s, efficiency improvements have been particularly marked in energy-intensive industries such as metallurgy, cement, paper, textiles, oil and coal processing, and electrical power generation （Philip, 2009）. Contrary to the above mentioned studies focusing on th

9、e decline of energy intensity, there are literatures specializing on the unusual phenomenon of energy intensity increasing since 2003. Ma and Stern （2008） indicated that the increase in energy intensity since 2000 could be explained by negative technological progress. Zhao et al. （2010） found that t

10、he most important driver behind Chinas energy intensity increase during 1998-2006 was the rapid development in energy-intensive industries.The common feature the previous studies have is that they usually apply the index decomposition method which is used to track the energy intensity/efficiency tre

11、nd and assess the fulfillment of energy efficiency targets. A shortcoming of purely descriptive decomposition studies is that it can not explain what is behind energy efficiency improvements and energy-saving structural change （Welsch and Ochsen, 2005）. To our understanding, not only the energy effi

12、ciency improvements could be caused by technology advancement and factors substitution, but also be possible by the energy price changing. This view is supported by other researchers. Employing a set of panel data for approximately 2500 of Chinas most energy intensive large and medium-sized industri

13、al enterprises during 1997-1999, Fisher-Vanden （2004） obtained that rising relative energy prices account for 54.4% of the decline in measured aggregate energy intensity. By reviewing of the deregulation of energy prices in China between 1985 and 2004, Hang and Tu （2007） implied that an increase of

14、relative prices of different energy types leads to a decrease in coal intensity, oil intensity, and aggregate energy intensity. Shi and Polenske （2005） found a negative own-price elasticity of energy intensity, a price-inducement effect on energy efficiency improvement, and a greater sensitivity of

15、the industry sector, compared to the overall economy. Zhao et al. （2010） argued that low energy prices have directly contributed to high industrial energy consumption and indirectly to the heavy industrial structure although they didnt empirically get whether the energy prices have effect on energy

16、intensity. This paper also attempts to investigate the determinants of energy intensity. However, we focus on electric power industry. To our understanding, few literatures have focused on power sector on this issue. The rest of the paper proceeds as follows. The first part of this paper outlines th

17、e electricity production and consumption in China. Following this, the methodology and data are developed to calculate the interfactor substitutability and decompose Chinas changing electricity intensity to ascertain the driving forces. The empirical findings and the interpretation of the results ar

18、e presented in the fourth section. The last section provides conclusions and discusses the implications of our empirical results for increasing Chinas electricity efficiency. 2. Electricity production and consumption in ChinaElectricity is one of the most important and basic needs of today community

19、. According to CSY （2008）, net electricity generation in China in 2007 totalled 3281 TWh, 83 percent of which was thermal power, followed by hydropower with 14.8%. The thermal power was generated through the combustion of fossil fuels. Coal plays the largest role in generation, followed by petroleum

20、 and then natural gas, with each representing 97.46 percent, 1.38 percent, and 1.16 percent of 2007s total net generation, respectively. It is obvious that fossil fuel combustion dominates in generating power, despite its negative environmental externalities. Moreover, coal-fired power generation wi

21、ll be in a stage of stable development until at least 2020.Before 2003, the State Power Corporation （SPC） dominated the production and supply of electric power in China and owned all transmission networks and a large proportion of the distribution facilities. At the end of 2002, the generating asset

22、s of the SPC were divested to five independent power producers to increase the competitiveness of the power industry and revamp pricing mechanisms. Despite the remarkable growth of electricity generation experienced from 1985 to 2007, the current balance between power supply and demand in China is s

23、till at a low level. Since 1998, demand for electricity has accelerated beyond what many economists had initially predicted. Consequently, a number of serious economic problems have begun to emerge.The average power network energy loss rate in China was 6.97% in 2007, which means electricity of 229

24、TWh was lost in power grid in 2007. Losses in transmission in 2007 were 206 TWh, accounted for 6.72% of consumption by usage. It indicates that Chinas power transmission system remains under-improved and there are much potential in enhancing the rates of electricity utilization efficiencyWith respec

25、t to electricity consumption, since 1980, it increased fast. From 1985 to 2007, it grew at annual averaging nearly 6.87%, versus 4.43% per annual for coal and 5.46% per annual for petroleum. The reasons are numerous, and mainly led by strong growth in industrial activity and increasing penetrations

26、of electrical equipment, space cooling and other electric appliances in both industrial and residential sectors. Fig. 1 shows the consumption structure of electricity by sectors in 1985 and 2007. Industry contributes most of the nations total electricity demand although it decreased from 79.74% to 7

27、5.3%. The residential sectors contribution to total national electricity demand rose from 5.4% to 11.07% during the same period times （23 years） （CSY）. It was largely due to the increasing penetration of electrical apparatus such as television, space cooling to the family with higher income. Fig1. E

28、lectricity consumption structure3. Methodology and dataThe translog production function （TPF） developed by Christensen et al. （1973） introduces interaction terms and can be estimated in a symmetric system of derived factor share equations that improve estimation properties relative to a single equat

29、ion. A general problem in estimating a production function is that input factors are likely endogenous, thereby violating a basic necessary condition for ordinary least squares to be unbiased. By using factor prices in a cost function framework this particular problem is most likely circumvented, so

30、 most empirical studies estimate a translog cost function instead of a translog production function （Koetse, 2008）. Our point of departure is to estimate the factor substitution. Following the previous studies on factor substitution, we use the typical approach reflecting factor substitutability wit

31、h a translog cost function. The unit cost function can be stated as: （1）Where Ln indicates the natural logarithm; C means the equilibrium total cost; Pit（Pjt） is the price of input factor i （j） at time t; t is a time variable to capture technical change.Linear homogeneity in put prices, an inherent

32、feature of any cost function requires the following regularity conditions:and for j=1,n. （2）Applying the Shephards lemma, we can derive a linear expression of the share of overall cost attributable to each factor i: （3）Here, is appointed as distribution parameters and as substitution parameters （Christensen et al., 1973）. The former measures how the co