This paper chooses the 36 industry sub-sectors as samplings, based on the data sets of added value and end-use energy consumption from 1993 to 2003 of China. By implying the improved index decomposition methods, ADMI and LMDI, the models are formulated. The results obtained show that structure effect which was less than intensity effect decreased year by year before 1998 and turned into steady from 1999. The intensity effect descended during the whole sampling periods. The biggest contributions on average structure effect and intensity effect were from sub-sectors of electric equipment and machines and raw chemical materials and chemical products, and the smallest contributions were from industries of production and supply of gas and petroleum processing and coking. The paper provides the foundation for policy making on improvement of industry energy efficiency.
Soon after the 1973 world energy crisis, there have been a variety of studies investigating energy saving, energy efficient use and energy related environment problems. Many countries have speeded up the research and development plans for saving energy in which energy efficiency plays an important role. Energy intensity which is the essential indicator of energy efficiency, at the country view, means the proportion of energy consumption of the whole country to gross domestic product. And at the industry or sub-sector level, it is defined as the proportion of the energy consumption of industry or sub-sector to value-added.
China has experienced spectacular economic uprising, with 9.79% average rate growth in GDP over the period 1980–2003(NBS, 1981–2004) [1]. Against the upward trend of economy, the energy intensity showed a reversed direction of decline continuously more than 20 years from 14.14 in 1980 to 4.87 in 2001, with 66.2% drop but an upturn since 2002 and reached 4.95 in 2003. Comparing with other developing countries, Chinese energy income elasticity was much lower in these periods [2], which clearly indicated that China used less energy with fixed GDP. The question arises of what drove the decline of energy intensity prior to 2001 and what were the dominant effects to the change of energy intensity since 2002. Researchers demonstrated different issues on theory and empirical results with different samples and methods [3], [4], [5], [6], [7], [8].
Smil [3] and Kamibara [4] conducted that the structural shifts away from energy-intensive industrial sub-sectors to less energy-intensive industrial ones was the major cause for the change of energy intensity. Stinton and Levine [5] examined the structural shift and real intensity change in Chinese industrial sector between 1980 and 1990 and found that the latter accounted for 85% of the country's overall industrial intensity vibration. Based on Chinese input–output tables in 1981 and 1987 for China, Lin and Polenske [6] conducted a structural decomposition analysis to explain China's energy use changes between 1981 and 1987. Their conclusion was that comparing to 1981 all the energy savings in China in 1987 can be attributed to energy efficiency improvements [6]. Zhang [7] proposed decomposition method of giving no residual noted as Laspeyres of choosing the samples of China's industrial sectors from1990 to 1997, showed that 88% of the cumulative energy savings in the industrial sector was attributed to real intensity change and the 12% contribution was from structure shifts. For the decline of energy intensity in China, Fisher-Vanden et al. [8]identified three broad explanations (1) sectoral change (2) sub-sector energy productivity gains and (3) inaccurate statistics. The statistic error was thought mainly from the emission account of coal produced by small coalmines that have been officially shutdown and imported fuels [8].
From the previous literatures, we see although the driving forces influencing the change of energy intensity were obtained, the energy scholars studied on the in-depth industry sub-sector views of China that have, to date, been rare and led to the conclusions and suggestions difficult to carry on. The major difficulty of in-depth research may come from the miscellaneous data sets. Since the revision of the data of energy consumption, the poor data before adjustment may be partly to blame for the swing of previous studies. Moreover, the driving factors would fluctuate as time goes on and the latest study chose the data of China that was just from 2000 [9]. While from 2001 the energy intensity of China terminated the state of 20 years drop and took a signified turn to rise [2]. In order to be high degree of sub-sectoral research and aid in optimizing fuel-mix and industry structure, with the change of the energy intensity of China from 1993 to 2003, the objective of this paper is to reveal the structure effect and intensity effect influencing the viability of industry sub-sector energy intensity from 1993 to 2003 by applying improved index decomposition. The further purpose is to list the contribution degree of every industry sub-sector to the two effects by implying contribution degree indexes.
The article is laid out as follows. Section 2 describes the basic method, Section 3 discusses the data and samples used and Section 4 applies the improved decomposition method to the data to analyze the driving factors for the changes of energy intensity form 1993 to 2003 and estimates the contribution degree of the sub-sectors to the two forces. Section 5 offers conclusions and instruction, gives the suggestions and proposes further studies.
2. Methodology
Decomposition methodology has become a useful and meaningful tool in energy and energy related environmental analysis. It can be seen from the foregoing that the decomposition methodology is a technique that provides a linkage between an aggregate and the original raw data whereby information of interest is captured in a concise and usable form. Divisia [10] proposed Divisia index decomposition and more detailed interpretation can be found in [11]. Boyd et al. [12], [13] proposed the multiplicative and addictive form arithmetic mean Divisia index (AMDI). Meanwhile in late 1970s, the Laspeyres index was developed and then quite widely adopted by academicians in the early 1980s. But after 1995, the Laspeyres index decomposition was seldom used as its main drawbacks are that it fails to pass the time reversal test and the residual arising from the interactions among factors in the decomposition would be too large to be accepted [14]. For example, the residual, brought about by Park [15] applying this method, was as high as 1322%, which led to difficulty in interpreting the result. Otherwise Divisia explains relative value while Laspeyres explicates absolute value. While the samples are time series, the Divisia index decomposition is more adaptive. By the requirement of practice and deep research on theory, some scholars continuously improved the index decomposition. They also compared different methods systematically and developed both in the methodological and application fronts. In 2001, Ang and Liu [16] proposed the logarithmic mean Divisia index (LMDI) which was perfect in decomposition and consistent in aggregation. Ang [17] discussed the properties of Divisia index and the Laspeyres index, concluded by recommending the multiplicative and additive LMDI methods due to their theoretical foundation, adaptability, ease of use and result interpretation, and some other desirable properties in the context of decomposition analysis. A practical guide for the use of LMDI was provided subsequently [18]. Ang and Liu [19]proposed and proved that replacing the zero value by a small number δ = 10–20 would give satisfactory decomposition results for index decomposition method. The settlement of handling zero-problem modified the LMDI method. In this paper, therefore, we first present two general methods for decomposing the energy intensity for industry sub-sectors, one popular conventional decomposition method AMDI and one recently proposed perfect decomposition method LMDI.
If you are interested in this research article, you can download the full length article through the download page using the link below, both PDF file and Word file are provided.
No comments:
Post a Comment