الگوهای نوسانات مشترک انتشار سرانه دی اکسید کربن: نقش بازارهای انرژی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|16669||2013||12 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Energy Economics, Volume 39, September 2013, Pages 1–12
This paper applies principal component analysis to investigate the linkages, or dominant co-fluctuation patterns, of per capita carbon dioxide emissions across countries for the time period 1950–2000. Energy resource world markets are investigated as an offsetting mechanism possibly coordinating emission fluctuations between countries. The results of the analysis provide evidence that world energy resource markets are acting as a coordinating mechanism for emission fluctuations in most cases. The results also suggest that until recently the dominant emission co-fluctuation pattern for developed countries differs from the dominant emission co-fluctuation pattern for developing countries. The common fluctuation pattern found in the 1984–2000 time period suggests that an offsetting mechanism does exist and will help contain global per capita emissions into the future. The strong degree that emissions are linked between countries and energy markets acting as an offsetting mechanism suggests that to be successful a global agreement to address climate change must require emission reductions by all major emitters, not just the developed countries.
There has been a significant amount of research into the statistical characteristics of national per capita carbon dioxide (CO2) emissions. This topic is important for projecting future global emissions and forecasting changes in the distribution among countries. Many studies have focused on tests of convergence of national per capita CO2 emissions (e.g., Aldy, 2006, Nguyen Van, 2005, Ordas Criado and Grether, 2011 and Strazicich and List, 2003). Stationarity at the global and/or national level has been examined by McKitrick et al. (in press) and Romero-Avila (2008). A key question at present is the extent to which emissions growth in one country or region affects emissions elsewhere. McKitrick et al. (in press) find evidence that offsetting effects occur between countries, and may constrain global per capita emissions in the future. The purpose of this paper is to investigate more closely the extent to which national per capita CO2 emissions are linked across countries, and whether those linkages can be explained based on energy markets, openness to trade, and other factors. This paper applies principal component analysis (PCA) to investigate the co-fluctuation patterns of per capita carbon dioxide emissions across countries. PCA allows for extraction of ranked orthogonal vectors from a data matrix, where ranking is by the percentage of underlying explained variance. If all countries' emissions respond linearly to the same external shocks, the first principal component (PC1) will explain a high proportion of variance in the whole data set. If countries' emissions are independent of each other over time, the first principal component will explain relatively little of the underlying variance. Hence we interpret the explained variance associated with the first principal component as an index of homogeneity of national per capita CO2 emissions. Our hypothesis is that energy prices transmit information across borders in such a way as to increase coordination of emission fluctuations. This is tested by examining the effect of energy prices on the index of homogeneity. We find evidence in support of the hypothesis; however, the pattern of emission fluctuations differs between developing and developed countries until the most recent time period (1984–2000). We then examine the effects of openness to trade and government intervention, and find that neither of these factors have an identifiable coordinating effect on emission fluctuations between countries. Overall the evidence suggests that emissions are strongly linked between countries, and we discuss what this may imply about future emission growth and global agreements to address climate change. The next section discusses the statistical characteristics of per capita CO2 emissions. Section 3.1 introduces the data, the analytical methodology, and analyzes a global sample, a developed country sample, and a developing country sample. Section 3.2 applies the methodology to samples of countries defined by region. Section 3.3 investigates the importance of openness to trade and government size. Section 4 concludes the paper.
نتیجه گیری انگلیسی
This paper applied principal component analysis to investigate linkages, in the form of co-fluctuation patterns, of per capita carbon dioxide emissions across countries. The analysis focused on identifying common factors that coordinate the emission fluctuations between countries. The results presented in Section 3.1 indicate a difference in co-fluctuation patterns of emissions between developed and developing countries over the first two time periods, but a common fluctuation pattern in the most recent time period. A possible explanation of this result is that, according to DeVany and Walls (1996), Gulen (1999), and Kleit (2001), among others, regional energy markets became more closely linked throughout the 1980s and 1990s. The common co-fluctuation pattern in the late period is inconsistent with the conclusions found in many of the convergence studies discussed in Section 2 (i.e., the emissions of developing countries behave differently than those of developed countries in relation to convergence). This result may be due, in part, to the convergence studies needing to use the whole time series rather than splitting it into three time periods. If the PCA analysis is conducted on the whole time series, the emissions of developing countries do indeed appear to behave differently than those of developing countries. Energy prices have coordinated the emissions of developed countries in all periods, however, prices only coordinate the emissions of developing countries globally in the most recent time period. The results of this paper support the findings of McKitrick et al. (in press), since evidence of an emission offsetting mechanism was found. Furthermore, the strong degree of emissions co-fluctuation combined with energy resource markets as an offsetting mechanism suggests that any global agreement to address climate change requires emission reduction efforts by all major emitters to be successful. For example, if an agreement only requires emission reductions by developed countries, as the Kyoto Protocol did, then these reductions and the associated reduced use of energy resources in developed countries will result in increased energy use and corresponding increased emissions in developing countries. The results of Section 3.2 suggest that regional common fluctuation patterns are driven by energy prices in most cases (all regions except for Africa in 1967–1983); however, energy prices were found to not always be the dominant common factor. The regional results suggest that another unidentified factor also plays a coordinating role. The regional results also suggest that although the emissions of developing countries were not globally coordinated by energy prices in the first two time periods, they were, for the most part, regionally coordinated by energy prices. The results from the expanded analysis in Section 3.3 indicate that openness to trade and the level of government intervention do not play a coordinating role on emissions in any time period for developed countries. And these are not contributing coordinating factors for developing countries in the 1984–2000 time period. An extension to this research would be to compare the emission fluctuation patterns of China and India with those of the developed countries. This could be done with a similar methodology as the one applied in this paper. Such a study could potentially provide insight on whether future increases of per capita emissions from India and China would be offset by reduced per capita emissions in the developed countries. Another extension would be to apply non-linear principal component analysis in order to consider further moments of the data.