The Great Guzzler? A Longitudinal Analysis of the Impact of Chinese growth on Energy Usein a Globalizing World
Erik Dietzenbacher1,2, Bart Los1*, Frederik Neuwahl3, Robert Stehrer4, Marcel Timmer1 and Alejandro Villanueva3
1 Faculty of Economics and Business, University of Groningen, The Netherlands;
2 Graduate University, Chinese Academy of Sciences, Beijing, China;
3 European Commission, Institute for Prospective Technological Studies, Sevilla, Spain;
4Vienna Institute for International Economic Studies (WIIW), Vienna. Austria;
* Corresponding author. E-mail:
Paper prepared for the 19th International Input-Output Conference, June13-17, 2011
(Alexandria, United States)
Preliminary and incomplete version, please do not quote
Short summary
In this paper, we will use a new World Input-Output database and input-output tools to divide global energy use into four quadrants: the Chinese footprint for non-renewable energy, the Chinese footprint for renewable energy, the footprint of the remaining countries for non-renewable energy and finally the footprint of the remaining countries for renewable energy. Comparing the relative sizes of these footprints over the period 1995-2006 gives an idea of the consequences of China’s increasing purchasing power for total global energy use and its effects on the global dependence on renewable and non-renewable sources of energy.
In the next set of analyses, we quantify the partial effects of a number of sources of change in the relative sizes of the Chinese footprints for renewable and non-renewable energy, using Structural Decomposition Analysis. This will give insights in the extent to which substitution is a consequence of changes in consumer tastes, substitution effects between intermediate inputs, changes in international trade structures, and changes in the amounts of renewable and non-renewable energy per unit of gross output.
Extended outline
Reduced transport costs and improved communication technology have led to an increasingly tight network of trade flows across many parts of the world, as well as lots of foreign direct investment. China's growth is an immediate consequence of its success in ensuring acrucial position in this network. This paper attempts to quantify the impact of China's take-off on worldwide energy use in the period 1995-2006.Three main questions will be addressed.
First, to what extent did the rising purchasing power of substantial parts of the Chinese population lead to increased energy use, in China itself and elsewhere in the world? Second, to what extent did relocation of production activities from North America and Western Europe to China lead to changes in worldwide energy use, taking into account that energy efficiencies vary across countries? Third, given the world-wide efforts to increase the share of renewable energy carriers in total energy use: what is the contribution of China to the overall change in the use of renewable energy? In view of the fact that both growing consumer demand in China and relocation of manufacturing processes will most probably be continuing phenomena for the next decade, answers to these questions shed useful light on policy questions regarding depletion of non-renewable energy sources.
Data
The analysis is based on a preliminary version of the World Input-Output Database (WIOD), which is being developed by an international consortium of researchers and funded by the 7th Framework Program of the European Commission.[1]
The core of the database consists of a series of annual full ‘World Input-Output Tables’ for the period 1995-2006. The tables include information for 40 countries. These countries (all 27 EU members plus 13 countries with large economies, such as the US, Mexico, Brazil, Russia, China, Japan and India) cover about 85% of world GDP. The World Input-Output Tables contain the values of annual transactionsbetween each industry i in each country r to each industry j in each country s. An example of such an intermediate input flow is the value of exports of the Australian mining industry to the Chinese metals industry. The 35 industries cover the entire spectrum of agriculture, extractive industries, manufacturing industries and services industries. Not all economic transactions relate to sales of intermediate inputs, however. Hence, the values of transactions that involve sales of products by each of the industries i in each of the countries r to each of the final demand categories f in each of the countries s have also been included. Examples of such final demand categories are household consumption, government consumption and demand for investment goods. Hence, information about the values of exports of the Chinese transport equipment industry to German consumers and American firms can also be readily found in the World Input-Output Tables. Finally, export values of each of the industries in each of the countries to the Rest of the World are also present in these tables.
WIOD is not the first project that delivers international input-output tables. The OECD, the EXIOPOL consortium, the GTAP consortium and IDE-JETRO (a Japanese research institute) have all estimated one or more of such tables.[2] Setting aside a vast number of often minor differences between the approaches adopted by each of these consortia, WIOD has two defining characteristics. First, it is the first database that offers opportunities for longitudinal analysis, since none of the other projects produced a time-series of tables with an identical industry classification. Second, the methods adopted to link international trade statistics (provided by the United Nations and the World Bank) and data on national production structures (in the form of National Accounts data and Supply and Use tables made publicly available by Eurostat and National Statistical Institutes) are much more transparent than for most other projects. Whereas many of those projects use complicated mathematical algorithms to align conflicting data, all data in WIOD are in agreement with National Accounts statistics, according to well-described procedures. Also, we have improved BEC.
The analysis in this paper also makes use of an environmental satellite account constructed in the WIOD project. This satellite account containsindustry-level data on energy use by carrierswith a clear distinction between non-renewables (coal, oil, natural gas) and renewables(solar energy, hydropower etc.), for each of the countries included in the World Input-Output Tables and each of the years for which these tables have been constructed. The data are based on publicly available data from the International Energy Agency. Since the industry classification for these energy use indicators is identical to the classification used for the input-output tables, these data constitute a very rich source for longitudinal input-output studies focusing on worldwide energy use.
Methods
The toolkit developed by input-output researchers contains a wide array of empirical methods to study the sources of growing energy use in the world. Detailed expositions of the methods used in this paper are beyond the scope of this abstract, but the intuition behind these methods definitely requires some attention.
Input-output analysis revolves around the concept of ‘vertically integrated industries’ or (in an international context) ‘global production chains’. The energy required to produce a Chinese car (a final product) is not limited to the use of electricity in a car manufacturing plant in that country, but should also comprise part of the energy that is used in e.g. the Japanese electronics industry delivering components, the Chinese basic metals industry and the Australian mining industry that exports iron ore to China’s basic metals industry. Using the well-known ‘Leontief inverse’ matrix, detailed information on international, interindustry transactions can be deployed to attribute parts of the gross output levels of each of the 35 industries in each of the 40 countries to final product sales of each of the 35 industries in each of the 40 countries. Hence, one could arrive at results like “0.5% of the gross output of the Japanese electronics industry ends up in Chinese cars, and for the Chinese basic metals industry and the Australian mining industry these shares are 2% and 0.8%, respectively.” After computing the amount of e.g. energy carried in oil per dollar of gross output for each of the industries in each of the countries, worldwide use of energy carried in oil can also be attributed to the deliveries of final products by each of the industries in each of the countries.[3]
By aggregating over a subset of final demand cells in the input-output tables, a number of policy-relevant indicators can be computed. Adding the energy ending up (or ‘embodied’) in all deliveries to Chinese households yields the so-called ‘energy footprint’ of Chinese consumers.[4] It gives an estimate of the energy that would have been used in production globally (in China itself and abroad) to satisfy demand exerted by Chinese households.
Energy footprints change over time, for a number of reasons. Consumption levels change; the composition of consumption can shift across products with varying energy intensities of production,; international trade patterns can change from imports favoring ‘energy-guzzling’ countries to patterns in which energy-efficient countries are dominant; energy-extensive intermediate inputs can be substituted by energy-intensive ones in domestic industries; and energy intensities (energy per unit of gross output) in particular industries can go down due to policy-making and/or technological progress.Structural Decomposition Analysis (the input-output analogon of index number approaches) provides ways to quantify the partial effects of such myriads of sources of change.[5]It offers answers to questions like “what would have happened to the Chinese energy footprint if energy intensities would have changed the way they actually did between 1995 and 2006, while all other determinants of change would have remained as in 1995?”
In this paper, we will first use the WIOD database and input-output tools to divide global energy use into four quadrants: the Chinese footprint for non-renewable energy, the Chinese footprint for renewable energy, the footprint of the remaining countries for non-renewable energy and finally the footprint of the remaining countries for renewable energy. Comparing the relative sizes of these footprints over the period 1995-2006 gives an idea of the consequences of China’s increasing purchasing power for total global energy use and its effects on the global dependence on non-renewable sources of energy. Next we will use Structural Decomposition Analysis to assess the quantitative contributions of several factors to changes in these four footprints between 1995-2006. We will make a systematic distinction between the effects of changes in the Chinese economy, changes in trade patterns between China and the Rest of the World and changes in the Rest of the World. This yields a balanced view on the impact of China’s ascent on the global economic ladder on the global use of renewable and non-renewable energy.
References
Davis, S.J. and K. Caldeira (2010), “Consumption-Based Accounting of CO2 Emissions”, Proceedings of the National Academy of Sciences of the United States of America, vol. 107, pp. 5687-5693.
Dietzenbacher, E. and B. Los (1998), “Structural Decomposition Techniques: Sense and Sensitivity”, Economic Systems Research, vol. 10, pp. 307-323.
Guo, D., C. Webb and N. Yamano (2009), “Towards Harmonised Bilateral Trade Data for Inter-Country Input-Output Analyses: Statistical Issues”, OECD Science. Technology and Industry Working Paper 2009/4, Paris, France.
Hannon, B. (2010), “The Role of Input-Output Analysis of Energy and Ecologic Systems”, Annals of the New YorkAcademy of Sciences, vol. 1185, pp. 30-38.
Hertwich. E. and G.P. Peters (2009), “Carbon Footprint of Nations: A Global, Trade-Linked Analysis”, Environmental Science & Technology, vol. 43, pp. 6414-6420.
Inomata, S. (2010), “Asia beyond the Crisis: Visions from International Input-Output Analyses”, Paper presented at the 18th International input-Output Conference, June 23-27, Sydney, Australia.
Johnson, R.C. and G. Noguera (2010), “Accounting for Intermediates: Production Sharing and Trade in Value Added”, Working Paper, Dartmouth College, Hanover, United States of America.
Tarancon, M.A., P. del Rio and F. Callegas Albinana (2010), “Assessing the Influence of Manufacturing Sectors on Electricity Demand. A Cross-Country Input-Output Approach”, Energy Policy, vol. 38, pp. 1900-1908.
Tukker, A., E. Poliakov, R. Heijungs et al. (2010), “Towards a Global Multi-Regional Environmentally Extended Input-Output Database”, Ecological Economics, vol. 68, pp. 1928-1937.
Vringer, K., K. Blok and B, van Engelenburg (2009), “A Step-Wise Guide for Energy Analysis: How to Calculate the Primary Energy Requirements of Households”, in: S. Suh (ed.), Handbook on Input-Output Analysis in Industrial Ecology, Springer, Heidelberg, Germany, pp. 491-506.
Wiedmann. T., R. Wood, J. Minx, M. Lenzen, D. Guan and R. Harris (2010), “A Climate Footprint Time Series of the UK – Results from a Multi-Region Input-Output Model”,Economic System Research,vol. 22, pp. 19-42.
1
[1] See for detailed information about the project and its contributors.
[2]See, for instance, Guo et al. (2009), Tukker et al. (2010), Johnson and Noguera (2010), Hertwich and Peters (2009) and Inomata (2010) for studies describing and/or using the results of these data construction efforts.
[3]See Vringer et al. (2009), Hannon (2010) and Tarancon et al.(2010) for recent examples of applications of input-output analysis in energy studies, and perspectives thereon.
[4] See, e.g. Davis and Caldeira (2010) and Wiedmann et al. (2010).
[5]See, e.g. Dietzenbacher and Los (1998).