مصرف نفت و بهره وری اقتصادی: تجزیه و تحلیل مقایسه ای پیشرفته، اقتصادهای در حال توسعه و نوظهور
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|21375||2011||9 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Ecological Economics, Volume 70, Issue 7, 15 May 2011, Pages 1354–1362
This paper investigates the economic efficiency–oil consumption relationship in 42 countries during the period 1986–2006. In a first stage by using DEA window analysis countries' economic efficiencies are obtained. In a second stage an econometric analysis based on robust GMM estimators reveals an inverted ‘U’-shape relationship between oil consumption and economic efficiency. In order to capture heterogeneities among countries' development stages the analysis has been separated into two groups (advanced economies and developing/emerging economies). The results show that advanced economies have much higher turning points compared to emerging and developing economies. It appears that oil consumption increases countries' economic efficiency. In addition the consumption patterns of oil products and its derivatives have changed through years and among countries. The different turning points from the econometric analysis indicate the dependence of oil consumption in advanced economies (higher turning points) is driven mainly by household purchasing activities and their standards of living (transport, housing and water, food, etc.). Finally, it appears that oil consumption is the main driver behind the progress of industrialization and urbanization regardless of the country's development stage.
The impact of energy on a countries' economic growth has been the center of research agenda for several years. Kraft and Kraft (1978) were among the first who investigated the relationship between energy consumption and economic growth for the USA. Since then many studies have used Granger causality tests in order to establish the link between energy and income, and energy and economic growth (Abosedra and Baghestani, 1991, Akarca and Long, 1980, Yu and Choi, 1985, Soytas et al., 2007, Soytas and Sari, 2009 and Zhang and Cheng, 2009). However, the results reported vary according to the country and the time period considered. Furthermore, Huang et al. (2008) applied a dynamic panel approach examining the energy consumption-GDP relationship for 82 countries from 1972 to 2002. They divided the data into four categories according to countries' income levels and their results showed absence of evidence that energy consumption leads economic growth in any of the four income groups. Similarly, Ozturk et al. (2010) using a sample of 51 countries, divided into sub-samples according to income groups and for the period 1971 to 2005, found that there is no strong relationship between energy consumption and economic growth indicating the need for further research. Jorgenson (1984) emphasized the fact that much research remains to be done until we are able to establish the relationship between energy utilization in productivity growth. According to Reynolds, 1994 and Reynolds, 1998 energy determines growth. The mechanism which explains how energy creates economic growth is presented and analyzed in Reynolds (2000). This mechanism is the energy utilization chain (EUC) that contains four links (namely energy source, conversion, consumption and service), which in combination can lead to growth. For instance, transportation and large independent mobile machinery (LIMM) operations are the most important energy services, which are strongly connected to countries economic growth. In contrast to the majority of studies investigating the energy consumption–GDP relationship, our paper investigates for the first time (to our knowledge) the economic efficiency1-oil consumption relationship. In the literature and in most of the cases oil consumption is embedded in the ‘total energy consumption’ variable. But none of the studies have approached the economic growth-oil consumption relationship given the strong oil dependence in most of the countries. According to Reynolds (2000) oil is the main energy source of LIMM operations and therefore the main driver behind countries' economic growth. In addition Reynolds (1994) explains the mechanism of how energy resources can contribute to economic growth and the reason why oil consumption is linked to countries' economic growth. Specifically, Reynolds (2000, p. 218) claims that energy resources have four grade types: (1) the weight grade (British thermal units, BTUs, per pound of energy resource); (2) the volume grade (BTUs per cubic foot of energy resource); (3) the area grade (BTUs per acre, where the energy resource is found in its original state); and (4) the state grade (the original physical state of the energy resource as a liquid, gas, solid, or energy). The higher the grade types of the energy resource, the higher its contribution to economic growth. Oil is a liquid state grade and it has high area grade, weight grade and volume grade characteristics that no other energy resources can match. Therefore, it appears that oil has the highest characteristic grades from all energy resources making difficult its substitution from other resources (Reynolds, 1994 and Reynolds, 1998). After 100 years the price of oil has been constant in real terms and any additional alternatives haven't been found in order to replace oil dependence of economic growth (Reynolds, 2000, p. 216). Additionally Goldemberg (2007) claims that fossil fuels (oil, coal, and gas) represent 80.1% of the total world energy supply, nuclear energy 6.3%, and renewables 13.6%. From that total of 80%, oil represents 45% of consumption, coal 30% and gas 25%. As can be observed oil consumption is still the dominant source of energy. Furthermore, Pimentel et al. (2009) over the past century indicated that affordable supplies of fossil fuels have mainly powered growth, industrialization, and prosperity of the USA and other countries' economies. In order to capture the effect of energy consumption on economic growth, the energy-economic growth studies have used GDP as a proxy of economic growth. Following the analysis by Ayres, 1996a and Ayres, 1996b our paper, instead of GDP as a proxy of economic growth, uses countries' economic efficiency. The standard growth theory (Solow, 1956 and Solow, 1957) assumes that the production function is a function of capital and labor, but the most important growth factor is an unexplained exogenous driver, the “technological progress” (Ayres and Warr, 2005 and Ayres and van den Bergh, 2005). As explained by Ayres, 1998 and Li and Ayres, 2008 “technological progress” is the efficiency with which resources are converted into final services. Generally, efficiency, where inputs are converted to outputs can provide us with more useful information than an absolute measure.2 As such our paper measures countries' economic efficiency over the period 1986–2006 by applying a variation of Data Envelopment Analysis (hereafter DEA) in order to handle panel data and allow for dynamic effects. This methodology is called DEA window analysis and has been introduced by Charnes et al. (1985). It works on the principle of moving averages and helps to establish efficiency measures over time by treating each country in a different period as if it was a different unit (Asmild et al., 2004, Cooper et al., 2007, Halkos and Tzeremes, 2009a and Halkos and Tzeremes, 2009b). Thus country's economic efficiency in a particular period is contrasted with its own efficiency in other periods in addition to the economic efficiency of other countries. In this way robustness related problems associated with DEA measurement techniques can be avoided. In a second stage our paper introduces a partial adjustment model formulation using Dynamic Panel Data Analysis with one- and two-step Arellano and Bond, 1991 and Arellano and Bover, 1995 GMM estimates. These GMM estimators are robust, as they do not require information of the exact distribution of the disturbances. The estimators rely on the assumption that the disturbances in the equation are uncorrelated with a set of instrumental variables. At the same time we also apply a random coefficients model formulation assuming that each parameter is a random variable. Countries are heterogeneous with different stochastic regression coefficients, which arise from a k-variate normal distribution. In this way we calculate the effect of oil consumption in the obtained countries' economic efficiencies and test the existence of an inverted U-shape relationship between economic efficiency and oil consumption. Li and Ayres (2008) suggest that oil consumption is a good energy proxy when the relationship between energy and economic growth needs to be tested. In addition to other studies our paper uses oil consumption since oil is an exhaustible nonrenewable resource and the consumption behavior of the current generation affects the economic development and welfare of future generations (Abdel and Sabry, 2005). The paper is organized as follows. Section 2 presents the data used in the analysis whereas, Section 3 discusses the proposed methodology and the econometric methods adopted. Section 4 presents the empirical results derived while the last section concludes the paper.
نتیجه گیری انگلیسی
This paper using a sample of 42 countries separated in 28 advanced economies and 14 developing and emerging economies for the period 1986–2006 explores empirically the economic efficiency-oil consumption relationship. In order to handle panel data, in a first stage a variation of the DEA methodology is applied in order to compute countries' economic efficiencies. Then in a second stage an econometric analysis has been contacted. Following the econometric methods adopted fixed and random effect models produced insignificant results. Similarly, the random coefficient model formulation used indicates the absence of an inverted ‘U’-shape relationship between economic efficiency and oil consumption. The reason is that there are enormous cross-country variations indicating the dynamic nature of the economic efficiency–oil consumption relationship which couldn't be captured by the previous econometric approaches. However the GMM used supports the ‘U’-shape relationship between economic efficiency and oil consumption producing turning points well within the sample for each case (total, advanced and emerging and developing countries). The turning point for the advanced economies is much higher than these points for emerging and developing countries. The fact that advanced economies have higher turning points of the contribution of oil consumption to economic efficiency (compared to emerging and developing economies) is an indication that oil consumption patterns have changed in countries over the years and among their development stages. In the total sampled countries the rate of adjustment with which economic efficiency adjust to equilibrium values is approximately 33% per annum implying that adjustment of economic efficiency is effected within almost three periods. In the case of the advanced economies the rate of adjustment of adjustment becomes 45% and 46% in the case of first differences and orthogonal deviations respectively. Similarly for emerging and developing economies the adjustment rate is similar to the total sample (33.5% as 1–0.675) in the case of first differences but lower and equal to 24% in the case of orthogonal deviations. According to Solow (1978) these variations depend on the structure of the economy and the stage of economic growth of the countries examined. It appears that for economies which tend to move their production structure towards services (advanced economies) oil consumption has greater turning points in relation to their economic efficiencies. This is mainly based on household purchasing activities and their standards of living. Finally, the results indicate that oil consumption causes economic efficiency indicating the oil consumption dependence among the countries.