Bulletin of Energy Economics http://www.tesdo.org/JournalDetail.aspx?Id=4 The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia Muhammad Shahbaz a,  , Md. Mahmudul Alam b, Gazi Salah Uddin c, Loganathan Nanthakumar d a COMSATS Institute of Information Technology, Lahore, Pakistan. National University of Malaysia (UKM), Malaysia c Linköping University, Sweden b d Universiti Teknologi Malaysia, Malaysia. Abstract: This paper investigates the impact of scale and technique, composition, trade openness and urbanization on energy consumption in the case of Malaysia. The study covers the sample period of 19702011 using quarter frequency data. We applied the bounds testing approach in the presence of structural breaks to examine the long run relationship between the variables. The VECM Granger causality is used to detect the direction of causality between the variables. Our findings indicate that growth effect (scale and technique effect) has a positive (negative) impact on energy consumption whereas composition effect stimulates energy demand in Malaysia. Energy consumption is positively influenced by both from trade openness and urbanization. This study opens new policy insights for policy making authorities to articulate a comprehensive energy and trade policy to sustain economic growth and improve the environmental quality of Malaysia. Keywords: Trade Openness, Energy Demand, Malaysia JEL Classification: O17 I. Introduction Energy utilization plays a crucial role in any modern industrial economy, since it can influence the level of productivity of other underlying production inputs. The debate between energy utilization and economic growth has been extensive, predominantly since the occurrence of the oil shocks in the 1970s. The International Energy Agency has projected that the global primary energy demand will rise by 40% from its 2007 level by 2030 [1]. The nonOECD nations collectively will account for over 90% of this increase; their share of global primary energy demand will rise from 52% to 63%. The consumption of fossil fuels by these countries will continue to dominate the energy scenario, accounting for 77% of the global rise; their demand for fossil oil is projected to increase from the 85 million barrels per day in 2008 to 105 million barrels per day in 2030, rising 24% of the demand. Trade liberalization promotes economic growth through the industrialization. In order to build a sound export-oriented industrial sector, energy is the key resources to fulfil its demand. It is required for the newly industrialized countries to increase their exports growth through the specialization in production that will stimulate a higher degree of competition, scale of economics and technology transfer. On the other hand, the role of imports allows the technology transfer and factors of production that will create energy demand in the country. The discussion explained that trade openness creates energy demand in three channels such as, growth effect, scale effect and technology effect. In terms of trade performance, Malaysia is one of the successful countries in the Southeast Asian region, which has utilized export with high rate of economic growth for the past 3 decades. Malaysia’s trade performance has slums in 1980s, 1998 and 2009 cause by global and Asian financial crises. Malaysia economic transformation begins in 1985 after the implementation of industrial plan. Manufacturing sector becomes an important driving instrument for Malaysia’s economic performance and contribute to exports, production and increase for energy utilization [2]. Total trade recorded for Malaysia in 2012 is RM1.31 trillion compared to RM1.27 trillion in 2011. Where, exports grew by 0.6% to RM702.19 billion while imports expended by 5.9% to RM607.36 billion. Among the top trading  Corresponding Author Email: shahbazmohd@live.com - 280 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. partners of Malaysia are ASEAN countries, China, Japan, European Union, United States of America, Australia, India, United Arab Emirates and Republic of Korea. The establishment of ASEAN Free Trade Area in 2003 also plays an important role on Malaysia export performance. Malaysia has a steady and consistent economic performance, although there is a fluctuation cause of global economic instability on trade, but the utilization of energy shows and upward trend since 1990s. There is a positive link between Malaysia’s trade and energy utilization for the past 3 decades. This is mainly because of the manufacturing sector development and increased in urban areas in Malaysia. Among export-oriented manufacturing sectors in Malaysia is palm oil, electric and electronic products, crude petroleum and rubber products. Recent statistic figures of industrialization development show an upward trend where Malaysia actually has emerged due to shift towards market-based policies and industrial policies introduced since 1990s [3]. Trade volume increased dramatically in the 1990s after the Malaysian government decide to reduce tariffs, especially in industrial and agriculture products. Another trade formulation appeared in 2000 where trade liberalization policies (trade openness) take into account by the government and this phenomenon has captured high volume of foreign direct investment. The recession in major export destinations in 1007-1998 affected Malaysia’s electric and electronic products' performance. However, the slower growth in manufacturing sector was mitigated by better export earnings from higher commodity prices, namely from palm oil, crude petroleum and liquefied natural gas. Malaysia also has continued this trade openness until today and the rising volume of trade has significant effects on energy supply and demand. Furthermore, energy demand also reflected on environmental issues as energy is generated from fossil fuels in Malaysia. This is key issue happens in most of developing nation in the past 3 decades and this finally will give negative impact sustainable environment. Numbers of empirical studies can be found on the issues of economic growth and environment hazards, but small attention has been given on trade and energy demand. Here, we have formulated a comprehensive literature and empirical models to capture the link between trade and energy use for Malaysia. This study may have a comprehensive effort on this topic of the economy of Malaysia and it will five ways contribution to the existing literature by applying: (i) both conventional and structural break unit root tests; (ii) The ARDL bounds testing approach to cointegration for the long run relationship between the variables. (iii) OLS and ECM for the long run and short run impacts (iv) The VECM Granger causality approach for a causal relationship is applied. II. Literature Review In earlier literature on energy economics, there is a large number of empirical studies on the positive association between energy consumption and stage of economic growth of the country. Economic development is influenced by the amount of energy usage as well as primary inputs usage [4] in production function. The intensified interest by the major economic powers of the world to gain a firm foothold on energy-based regions across the globe is a testimony to the fact that energy will remain a major focus in the foreseeable future. The battle for such control will also increase, as more energy will be needed to meet the demand for future economic growth. On the other hand, there has been an increased pressure from the developed economics to reduce CO2 emissions in order to reduce the rate of global warming and climate change. Therefore, emerging and industrialized countries are concerned about their negative effects on economic growth caused by the restricted use of energy. There are a large number of empirical literatures on the relationship between energy consumption and economic growth. Using the annual data from 1948 to 1994, [5] examined the relationships between income, energy consumption, labour market and capital stocks in the United States of America (USA). He concluded that there is a mutual causality between energy consumption and GDP in USA. On the other hand, [6] found conflicting evidences with the neutrality hypothesis supported in a substantial number of countries, with little support for the hypothesis that energy consumption causes economic growth. According to the eighteen developing countries, [7] found that causality running from energy consumption to economic growth but not vice versa applying the panel cointegration and Granger causality approaches. A group of six Gulf Cooperation countries, [8] found a unidirectional causality running from economic growth to energy consumption. [9] included energy consumption in their analysis and found a relationship between income and energy consumption and emissions. Using a tri-variate model, [10] failed to identify a significant Granger causality link between any of the variables. Recent study by, [11] used capital, labour, technology, and energy as separate inputs to test the existence of long-run relationship between output growth and electrical energy usage and suggested long run relationship between growth and electricity consumption and bidirectional causality between electrical energy consumption and real GDP in long run. However, the causality is - 281 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. unidirectional from energy to output in short run. A study by Latin America, [12] found both short run and long run causality from energy consumption to economic growth applying the panel cointegration method. [13] supported that unidirectional causality runs from energy consumption to economic in the case of Bangladesh. On the contrary, [14] found that economic growth Granger causes energy consumption in Tunisia. Recently, [15] reported feedback hypothesis between energy consumption and economic growth in Pakistan. The effect of trade liberalization on energy consumption is relatively new in the energy-income literature. [16] argued that trade liberalization may have effect on energy consumption, because it induces changes in trade policies that are related to energy usage, such as reduction in tariff and non-tariff barriers on energy efficient products. Liberalization may also influence energy consumption indirectly through changes in economic growth, environmental regulations, implementation of ecologically beneficial management practices, reallocation of resources, etc. It is also anticipated that liberalization brings about institutional changes, which affects the transfer of energy-saving technologies that can help to improve energy efficiency. Trade liberalization may also induce the technique effect indirectly through the increase in income. Higher income is believed to change consumer preferences, inducing the government to reform environmental and energy regulations. Similarly, in another recent study, [17] explained how exports and imports are linked to energy consumption. Exports expansion increases the demand for the factors of production (capital, labour, energy) used to make the exports. Once exports are produced, machinery and equipments must be used to load and transport the exports to seaports, airports or other docking stations where the exports are then offloaded and re-loaded for voyages abroad. The machinery and equipments used in the production, processing and transportation of goods for export require energy to operate. An increase in exports represents an increase in economic activities in export-oriented sectors and this should increase the demand for energy. Likewise, imported goods can affect the demands for energy in two ways. Firstly, imports are trade flows into a country and this necessitates a well-functioned transportation network to move goods around, which requires energy. As a consequence, an increase in trade flows is expected to increase energy consumption. Secondly, the composition of imports can affect energy consumption especially if the imports are energy intensive products like automobiles, dishwashers, air conditioners, etc. It is also possible that energy consumption can affect the flow of imported goods if the imported goods are machinery or equipments that require energy to operate. Energy conservation policies or lack of accessible energy may reduce the usefulness and efficiency of energy-dependent imported goods, making it less likely that such goods will be imported. In both cases, there is also a possibility of a feedback relationship or no statistically significant relationship between imports and energy, and exports and energy. The empirical studies, however, demonstrated inconclusive results on the relationship between trade openness and energy consumption. For instance, [18] for a panel of six Middle-Eastern countries (Iran, Israel, Kuwait, Oman, Saudi Arabia, and Syria) showed short-run Granger causality running from electricity consumption to real GDP and from economic growth to exports. They also found evidence in favour of a long run Granger causality relationship running from exports and electricity consumption to real income earned from exports and real income to electricity consumption. In two research papers on almost the similar topic, studying electricity generation and consumption in Malaysia, [2] found evidence for Granger causality running from electricity generation to exports; however, [19] did not find any evidence for a Granger causal relationship between exports and electricity consumption. While [2], [19] and [18] focused on the relationship between exports and electricity; [20] focussed on the more general relationship between energy consumption and trade (measured by either exports or imports). [20] for a panel of Middle-Eastern economies (Bahrain, Iran, Jordan, Oman, Qatar, Saudi Arabia, Syria, and United Arab Emirates) found that short run dynamics show causality running from exports to energy consumption and the feedback relationship between imports and energy consumption. In the case of Pakistan, [21] investigated the relationship between natural gas consumption and economic growth by incorporating exports in production function. Their results indicated that the variables are integrated for a long run relationship. Natural gas consumption, exports, capital and labour add in economic growth. Natural gas consumption Granger causes economic growth and exports. Several studies have been conducted recently considering the technique effect, labour-capital composition effect and urbanization effect to determine the energy demands in an economy. Utilizing the theoretical model of [22], [23] has modelled empirically the mechanisms through which trade liberalization affects the pattern of national energy consumption on 32 developed and developing countries. The results suggest that per capita energy consumption is subject to scale effect which, for the mean country, outweighs the negative technique effect, indicating that - 282 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. regulations and technological improvements are not keeping pace with the growth of GNP. With regard to the trade induced composition effect, evidences have been found to suggest that energy intensive industries are subject to conflicting forces as postulated by the factor endowment and the pollution haven hypotheses. These results indicate that trade liberalization and per capita energy consumption have a positive correlation. Working on the annual data from 1971 to 2007, [24] empirically examined the dynamic causal relationships between CO2 emission, energy consumption, economic growth, trade openness and urbanization for a panel of nine newly industrialized countries. The results of the study have supported the hypothesis that there is no evidence for long run causal relationship, but there is a unidirectional short run causal relationship from economic growth and trade openness to CO2 emission, from economic growth in energy consumption, from trade openness to economic growth, from urbanization to economic growth and from trade openness to urbanization. These results support the study of [20] on eight MiddleEastern countries for the period of 1980-2007, showing short-run causality from exports to energy consumption, and the feedback effect between imports and energy consumption is found. Studying in seven South American countries for the period of 1980 to 2007; [17] showed a short run bi-directional feedback relationship between energy consumption and exports, output and exports and output and imports. Evidences also suggest that there is a one way short run relationship from energy consumption to imports. In long run, a causal relationship has been established between trade (exports or imports) and energy consumption. [16] working for 54 developing countries, showed that trade liberalization per se does not affect the growth of energy consumption of developing countries, but its interaction with capital per labour reduces the growth of energy consumption as capital per labour increases. However, the effect is only significant after a certain minimum threshold level capital per labour is reached. On other hand, economic growth increases energy consumption and its effect is not conditional upon trade liberalization. Based on these mixed empirical findings, this paper determines the impacts of trade openness, income effect, scale and technique effects, labour-capital composition effect, and urbanization on energy consumption in the case of Malaysia. Malaysia, an important economic player in East Asia, has successfully pursued a policy of robust economic growth enviable for any emerging nation. However, a significant spurt in energy consumption followed by a concomitant rise in pollutant emissions in recent times has made the choice of the study for this country not only timely but also of much significance. According to the United Nations Development Report, CO2 emission in Malaysia has increased by 221% in 2004 compared to the emissions level of 1990. The report lists Malaysia at 26th among the top 30 green house gas emitting nations. If the current rate of emissions persists and other top greenhouse gas emitters improve their energy consumption efficiency, Malaysia may move up the ladder. The fact that Malaysia is a signatory to Kyoto Protocol did little to alter the pattern of the rapid growth in emissions [25]. However, the several initiatives taken by the government to promote renewable energy and to reduce CO2 emissions are reassuring. Therefore, the findings of the energy demand model will help policymakers choose appropriate strategies for sustainable economic growth in Malaysia. Recent study by, [26] working in Malaysia suggested that energy consumption is influenced by economic growth and financial development, both in short run and long run, but the population-energy relationship holds only in the long run1. III. Data, Model Construction and Estimation Strategy The data on energy consumption per capita (kt of oil equivalent), CO2 emissions per capita (metric tons per capita), real capital stock, real trade (exports + imports), labor, urban population has been obtained from world development indicators. The present study has covered the time period of 1970-2011. We have used series of population to transform the data into per capita. The series are converted into a natural log form. We follow [23], [27], [16] and [28] to examine the impact of trade openness on energy demand, we use quarter frequency data over the period of 1970-2011 in the case of Malaysia. The empirical equation is modeled as following: E t  f (Y t , Y t 2 , KL t , TR t , U t ) (1) 1 [29] also reported bidirectional causality between financial development and energy consumption in case of Malaysia - 283 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. We apply log-linear model2. Our empirical model is articulated as follows: ln E t   1   2 ln Yt   3 ln Yt 2   4 ln KL t   5 ln TR t   6 ln U t   t (2) where, ln Et , ln Y t , ln Yt 2 , ln KLt , ln TRt and ln U t is natural log of energy consumption per capita, real GDP per capita, real GDP per capita square, capital-labor ratio per capita, trade openness (exports + imports) per capita and urbanisation per capita. Yt indicates economic growth effect, scale and technique effect is captured by Y t 2 , KLt represents composite effect, trade openness is shown by TRt , urbanization effect is indicated by U t and E t is for energy use .  t is error term expected to be independently identically distributed. We expect that ln Y t > 0, ln Yt 2 < 0, ln K t .Lt > 0, ln TRt > 0 and ln U t < 0. II.I Zivot-Adndrews Unit Root Test Historically, in order to test stationarity properties of the variables unit root tests like ADF by [25]; P-P by [30]; KPSS by [31]; DF-GLS by [32] and [34] have been used extensively. However, due to lack of information on structural breaks stemming in the series, these tests produce unreliable results. To remove this anomaly [35] suggested another model that allows to accommodate single structural break point in the variables at level form, in slope of trend component, and in intercept and trend function. Using [35] model the structural break in the series can be tested as:  x t  a  ax t 1  bt  cDU t   x t  b  bx t 1  ct  bDT t  k d j 1 k d j 1 j j  xt  j   t xt j   t  x t  c  cx t  1  ct  dDU t  dDT t  k d j 1 j  xt  j   t (3) (4) (5) where DU t denotes dummy variable and gives the mean shift incurred at each point while DTt denotes trend shift variable. DU t  t  TB ...if t  TB  1 ... if t  TB and DU t      0 ... if t  TB  0 ...if t  TB The null hypothesis of unit root break date is c  0 which indicates that series is not stationary with a drift not having information about structural break stemming in the series while c  0 hypothesis implies that the variable is found to be trend-stationary with one unknown time break. Zivot-Andrews unit root test fixes all points as potential for possible time break and does estimation through regression for all possible structural breaks successively. Then, this unit root test selects that time break which decreases one-sided t-statistic to test cˆ(  c  1)  1 . Zivot-Andrews intimate that in the presence of end points, asymptotic distribution of the statistics is diverged to infinity point. It is necessary to choose a region where the end points of sample period are excluded. Further, Zivot-Andrews suggested the trimming regions i.e. (0.15T, 0.85T) are followed. II.II The ARDL Bounds Testing Approach Since traditional approaches to cointegration have certain demerits, we have used the structural break autoregressive distributed lag model or the ARDL bounds testing approach to cointegration in the presence of structural break 2 See [14], [33] and [15] for more details - 284 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. stemming in the series. The ARDL bounds testing approach to cointegration has certain merits like it is flexible regarding integrating order of the variables whether variables are found to be stationary at I(1) or I(0) or I(1) / I(0). In addition, Monte Carlo investigation confirms that this approach is better suited for small sample size [36]. Moreover, a dynamic unrestricted error correction model (UECM) can be derived from the ARDL bounds testing through a simple linear transformation. The UECM integrates the short run dynamics with the long run equilibrium without losing any information for the long run. The empirical formulation of the ARDL bounds testing approach to cointegration is given below:  ln E t   1   T T   E ln E   Y ln Yt 1   Y 2 ln Yt 21   K / L ln KL t 1   TR ln TRt 1 p q r s j 0 k 0 l 0   U ln U t 1    i  ln E t  i    j  ln Yt  j    k  ln Yt 2 k    l  ln KL t  l i 1 t t m 0 n0 (6)    m  ln TRt  m    n  ln U t  n   t Where, ln E t , ln Yt , ln K / Lt , ln TR t and ln U t natural log of energy consumption per capita, natural log of real GDP per, natural log of capital-labour ratio, natural log of trade openness per capita and natural log of urbanisation per capita.  is for difference operator and  t denotes residual term. F-statistics are computed to compare with upper and lower critical bounds generated by [37] to test for existence of cointegration. The null hypothesis to examine the existence of the long run relationship between the variables is H 0 :  E   Y   K / L   TR   U  0 against alternate hypothesis ( H a :  E   Y   K / L   TR   U  0 ) of cointegration for equation-6. Using [37] critical bounds, if computed F-statistic is more than upper critical bound (UCB) there is cointegration between the variables. If computed F-statistic does not exceed lower critical bound (LCB) the variables are not cointegrated for a long run relationship. If computed F-statistic falls between lower and upper critical bounds then decisions regarding cointegration between the variables is uncertain. However, since our sample size is large (160 observations) and critical bounds generated by [37] may be suitable. Therefore, we use lower and upper critical bounds developed by [37] rather than [38]. II.III The VECM Granger Causality Approach We should apply the vector error correction model (VECM) to investigate the causal relationship between the variables once cointegration relationship exists between the series. It is argued by [47] that the VECM is an appropriate approach to examine causality between the variables when series are integrated at I(1). The empirical equation of the VECM Granger causality approach is modelled as follows:  ln  ln   ln (1  L )   ln  ln   ln   a1   a  Yt   2  2 a3  Yt      KL t  a4    TR t    5 U t    6             Et  b 11 i b  21 i p  b 31 i   (1  L )  i 1  b 41 i b  51 i  b 61 i  1t       2t    3t     ECT t  1    4t       5t      6 t  b 12 i b 13 i b 14 i b 15 i b 16 i   ln E t  1    b 22 i b 23 i b 24 i b 25 i b 26 i   ln Y t  1  b 32 i b 33 i b 43 i b 53 i b 36 i   ln Y t 2 1     b 42 i b 43 i b 44 i b 45 i b 46 i   ln KL t  1  b 52 i b 53 i b 54 i b 55 i b 56 i   ln TR t  1     b 62 i b 63 i b 64 i b 65 i b 66 i   ln U t  1  (7) - 285 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. where (1  L) indicates difference operator and lagged residual term is indicated by ECTt-1 which is obtained from long run relationship while  1t ,  2t ,  3t ,  4t ,  5t and  6t are error terms. These terms are supposed to be homoscedastic i.e. constant variance. The statistical significance of the coefficient of lagged error term i.e. ECTt 1 shows the long run causal relationship causal relationship between the variables. The short run causality is shown by statistical significance of F-statistic using Wald-test by incorporating differences and lagged differences of independent variables in the model. Moreover, the joint significance of the lagged error term with differences and lagged differences of independent variables provides joint long-and-short runs causality. For example, b12,i  0i implies that economic growth Granger-causes energy consumption and economic growth is Granger cause of energy consumption shown by b 21 , i  0  i . IV. Results and their Discussions The primary step is to test the unit root properties of the variables to proceed for the ARDL bounds testing approach to cointegration. The ARDL bounds testing approach is free from pre-unit root testing but we ensure that none of the variables is integrated at I (2). The bounds testing approach assumes that variables should be stationary at I(0) or I(1) or I(0)/I(1). So to overcome this issue, we have applied traditional unit root tests such as ADF test by [39] PP test by [30] and DF-GLS test by [32]. The results of these unit root tests are reported in Table-1. The results show that energy consumption, income, capital-labor ratio, trade openness and urbanization have a unit root problem at level with intercept and trend. This series is found to be stationary at 1st difference. This shows that the variables have unique order of integration. Table-1. Unit Root Analysis Variables ADF PP DF-GLS ln Et –2.0780 (9) –2.0807 (9) –2.6542 (3)  ln E t –4.5032 (8)*** –4.5032 (9)*** –10.4149 (3)*** ln Yt –1.9093 (6) –1.0571 (6) –1.3670 (5)  ln Yt –4.6320 (6)*** –4.6320 (6)*** –4.6135 (6)*** ln KLt –1.7590 (4) –1.7852 (12) –1.8756 (1)  ln KLt –4.2119 (4)*** –4.1736 (6)*** –6.1110 (6)*** ln TRt –0.0854 (9) 0.0914 (1) –0.1797 (4)  ln TRt –3.5589 (8)** –6.0791 (6)*** –3.3142 (4)** ln U t -0.3421 (8) 1.0792 (3) -1.4042 (9)  ln U t -4.8665 (10)*** -4.6162 (12)*** -3.5619 (3)*** Note: *** and ** denote the significance at 1% and 5% levels respectively. Figure in the parenthesis is the optimal lag structure for ADF and DF-GLS tests, and bandwidth for the PP test. These unit root tests may provide misleading results regarding stationary properties of the variables. The reason is that these unit root tests do provide information about structural breaks arising in the series. The information about structural breaks in the series would enable policy makers to analyze the impact of major economic policies and help them in articulating a comprehensive economic, energy as trade policy to sustain long run economic growth. This issue is resolved by applying Z-A unit root test, which accommodates information about one unknown structural break occurring in the series. The results of Z-A unit root test are pasted in Table-2 the results indicates an I(1) level stationarity for all variables with structural break appeared in late 1970s, 1980s and 1990s. Several reasons cause on this structural breaks. First, economic crisis faced my Malaysia in 1985-1987, where the government faced budget - 286 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. Variable T-statistic ln Et -4.705 (1) ln Yt -4.311 (2) ln KLt -4.700 (2) ln TR t -3.231 (1) ln U t -3.102 (2) Table-2: Zivot-Andrews Structural Break Unit Root Test At Level At 1st Difference Time Break Decision T-statistic Time Break Unit root exists 1993Q2 -12.317 (3)* 1978Q3 Unit root exists 1991Q2 -8.271 (3)* 1986Q3 Unit root exists 1997Q2 -7.181 (2)* 1997Q2 Unit root exists 1992Q2 -8.534 (3)* 1987Q2 Unit root exists 2000Q2 9.723 (3)* 1989Q2 Stationary Stationary Stationary Stationary Stationary Stationary Note: * represents significant at 1% level. Critical T-values are -5.57 and -5.08 at 1% and 5% levels respectively. Lag order is shown in parenthesis. deficits caused by heavy industrialization more and adverse trends in price of Malaysia’s major export products which put an end to the state-led heavy industrialization push [40]. Besides that, high oil prices, which push upward shift in energy cost also, have provided fresh impetus to energy usage in 1980 in Malaysia. Secondly, trade and energy used in manufacturing sector reflected by the Promotion of Investment Act introduced by the Malaysian government to increase the involvement of private investors to boom-up the trade volume of manufacturing sector. Thirdly, breaks in late 1980s generally involve with tariff reduction and removal of quantitative import restrictions, especially in the manufacturing sectors [41]. Finally, breaks in late 1990 are generally come from the Asian financial crisis. This crisis has slowdown most of major manufacturing sectors in Malaysia and trade volume reduce dramatically in this period because most of vendor industries in Malaysia connected with foreign investors from Japan and Singapore. The decline in manufacturing sector also reflected on the demand for energy resources and this has slow down the demand for energy in Malaysia. The results disclose that variables are non-stationary at level but integrated at order 1 i.e. I(1). This exposes that we should apply the ARDL bounds test approach to examine long run association between the variables. Table-3: Lag Order Selection VAR Lag Order Selection Criteria Lag LogL LR FPE 0 2176.286 NA 6.66e-20 1 4456.584 4361.070 4.36e-32 2 4736.099 513.6086 2.08e-33 3 4761.973 45.6022 2.38e-33 4 4772.083 17.0603 3.31e-33 5 4938.858 268.9250 6.55e-34 6 5046.284 165.1682* 2.74e-34* 7 5056.361 14.7371 3.91e-34 8 5062.527 8.5551 5.91e-34 * indicates lag order selected by the criterion LR: sequential modified LR test statistic (each test at 5% level) FPE: Final prediction error AIC: Akaike information criterion SC: Schwarz information criterion HQ: Hannan-Quinn information criterion AIC -27.1285 -55.1823 -58.2262 -58.0996 -57.7760 -59.4107 -60.3035* -59.9795 -59.6065 SC -27.0132 -54.3757 -56.7270* -55.9086 -54.8930 -55.8358 -56.0367 -55.0208 -53.9559 HQ -27.0817 -54.8545 -57.6174 -57.2099 -56.6053 -57.9590 -58.5709* -57.9659 -57.3120 According to the ARDL approach, lag order of the variables is important for the model specification. Table-3 indicated the lag length criterion. In this paper, we followed Akaike information criteria to select an appropriate lag length. It is pointed by [42] that AIC has superior power properties for small sample data compared to any lag length - 287 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. criterion. Akaike information criterion provides efficient and consistent results as compared to final prediction error (FPE), Schwarz information criterion (SBC) and Hannan-Quinn information criterion (HQ). Based on empirical evidence provided by AIC, we find that the optimum lag is 6 in such quarter frequency data over the period of 19702011 in the case of Malaysia. The results of bounds testing are detailed in Table-4. Our empirical evidence opines that the calculated F-statistic exceeds the upper critical bound at the 5% level of significance. This concludes that we may reject the null hypothesis of no cointegration. The results reported in Table-4 show that there is evidence of cointegration once we treated income, capital-labor ratio, trade openness and urbanization as predictor variables. This shows that there exists a long run relationship between trade openness and energy consumption in the case of Malaysia over the period of 1970-2011. Log run model fulfills the assumptions of the classical linear regression model (CLRM) such as serial correlation, autoregressive conditional heteroskedasticity as well as white heteroskedasticity and specification of the model. Our results find that evidence of no serial correlation is found. The autoregressive conditional heteroskedasticity and white heteroskedasticity are not found. Finally, the bounds testing model is articulated well as confirmed by Ramsey RESET tests. Tabel-4: The ARDL Bounds Testing Analysis Estimated Model FE ( E t / Yt , KLt , TRt ,U t ) Optimal Lag Length F-statistics Critical values# 1 per cent level 5 per cent level 10 per cent level (6, 6, 5, 6, 6, 5) 4.757** Lower Critical Bound 3.60 2.87 2.53 0.7521 R2 2 Adjusted- R F-statistics Durbin-Watson Diagnostic Tests Test  2 SERIAL Upper Critical Bound 4.90 4.00 3.59 0.6715 9.33666* 1.9626 F-statistic 1.4828 Prob-value 0.2311  2 ARCH  2 WHITE 0.4813 0.6188 0.2332 0.7668  REMSAY 1.2363 0.2684 2 Note: * denotes the significance at 1% level respectively. The optimal lag structure is determined by AIC. The next turn is to find the impact of independent actors on the dependent variable. The results show that income has a positive impact on energy consumption. This shows that income effect is more elastic for energy consumption and it is statistically significant at the 1% level of significance. A 1% increase in income is linked with 1.9980% energy consumption. Due to the vision 2020 to become a developed nation, the Malaysian economy is growing very fast. Like other developed countries, consumption patterns of people are moving more towards the technological usage. Moreover, better financial growth and credit facilities are leading to more use of energy. Further, due to the high subsidy and lower price of fuel, people tend to use more vehicles and household machineries. Therefore, energy reduction policies will have less impact on economic growth here. The impact of scale and technique effects on energy consumption is negative and significant at 5% significance level. This implies that economies of scale and adoption of advanced technology saves energy by 0.3036%. Therefore, energy consumption policies should be - 288 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. Table-5: Long-and-Short Runs Analysis Dependent Variable = ln E t Long-Run Results Coefficient -2.2585* Std. Error 0.7545 T-Statistic -2.9932 1.9980* 0.6764 2.9538 -0.3036** 0.1321 -2.2966 ln KLt 0.0474*** 0.0263 1.8015 ln TRt 0.0617** 0.0283 2.1800 ln U t 0.8188* 0.1976 4.1437 Variable Constant ln Yt ln Yt R 2 2 0.9909 F-statistic 35.6064* Short-Run Results Variable Constant Coefficient -0.0024 Std. Error 0.0020 T-Statistic -1.2098 ln E t 1 0.4548* 0.0634 7.1654 ln Yt -6.5693* 2.4965 -2.6313 0.3837* 0.1336 2.8718 ln KLt 0.0954** 0.0413 2.3095 ln TRt -0.1276*** 0.0704 -1.8115 ln U t 2.5121 1.8945 1.3259 -0.1774* 0.0320 -5.5437 ln Yt 2 ECM t 1 R 2 F-statistic D. W Test Diagnostic Tests 0.4483 18.1096* 2.0688 Test  2 NORMAL F-statistic 0.6539 Prob. Value 0.7211  2 SERIAL  2 ARCH 0.6447 0.4232 1.3588 0.2454  WHITE  2 REMSAY 1.0538 0.4042 0.7043 0.4026 2 Note: *, ** and *** represent significance at 1%, 5% and 10% levels respectively. driven based on the adoption of advanced technologies and achieving economies of scales, especially in the industrial and manufacturing production level. The impact of capital-labor ratio (composite effect) has a positive impact on energy demand and it is statistically significant at 10% level of significance. It implies that keeping other things constant, a 1% increase in capital-labor ratio is linked with energy use by 0.0474%. Due to this highly inelastic nature of a composite effect on energy demand, emphasizing on the usage of more capital in terms of labor - 289 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. or capital intensive production will lead to a very low increase of energy usage. Based on the technique and composition effects, Malaysia should increase and focus more on advanced energy saving technologies to control its high usage of energy. The positive and statistically relationship is found between trade openness and energy consumption. We find that a 0.0617% increase in energy consumption is related to 1% increase in trade openness, all else is same. Generally, in a low-income country an increase in trade openness will increase their energy usage, but in a high-income country energy usage falls in response to trade liberalization. As Malaysia is a growing economy, the energy consumption shows an inelastic nature in response to trade openness. Therefore, emphasizing on more openness in trade will lead to very low level of increase in energy usage. Thus, trade expansion policies like export promotion policies designed to increase exports will not increase energy consumption here. Another implication of the findings is that environmental policies which reduce energy consumption will not affect the growth in exports. The impact of urbanization on energy consumption is positive and significant at 1% significance level. We infer that a 1% urbanization raises energy demand by 0.8188% by keeping other things constant. Due to the vision 2020 to become a developed nation, the countryside is growing rapidly with the urban facilities. At the same time, infrastructural advancement causes the expansion of urban areas. The growing income facilities in rural areas are also leading to more use of technology. Eventually, the urbanization effect causes an increase in the energy consumption in Malaysian society. Figure-1. Plot of Cumulative Sum of Recursive Residuals 40 30 20 10 0 -10 -20 -30 -40 1980 1985 1990 1995 CUSUM 2000 2005 2010 5% Significance The straight lines represent critical bounds at 5% significance level Figure-2. Plot of Cumulative Sum of Squares of Recursive Residuals 1.2 1.0 0.8 0.6 0.4 0.2 0.0 -0.2 1980 1985 1990 1995 CUSUM of Squares 2000 2005 2010 5% Significance The straight lines represent critical bounds at 5% significance level The short run results are also reported in the lower segment of Table-5. The results opine that future energy demand is positively affected by energy demand in the current period at 1% level of significance. Economic growth declines but the scale and technique effects increase energy consumption. These findings are statistically significant at the 1% level of significance. The capital-labor ratio has positive and statistically significant impact on energy demand. Trade openness declines energy consumption. Urbanization is also positively linked with energy consumption but it - 290 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. is statistically insignificant. The long run relations between trade openness and energy consumption established through the lagged error correction term which is negative and significant. The negative sign of the error correction term confirms the expected convergence process in the long-run dynamics of trade openness and energy consumption model in Malaysia. In fact, 18 percent of the last year’s disequilibria are corrected in the current year based on these results, suggesting a speed of adjustment in the relationship process following a shock. In order to achieve the stable long run equilibrium energy demand in Malaysia, it would take around five years and two quarters. The results of the diagnostic tests suggest that the underlying desirable assumptions are fulfilled. The short run findings of the empirical results are consistent in order to implement environmental policy in Malaysia. The stability of the ARDL bounds testing estimates is investigated by applying the CUSUM and CUSUMsq tests. The results are shown in Figure-1 and 2. The plots of the CUSUM statistics are well within the critical bounds. Table-6. Chow Forecast Test Forecast from 1983Q1 to 2011Q4 F-statistic 0.7191 Probability Log likelihood ratio 0.1847 Probability 0.9100 0.5101 The plot of the CUSUMsq of squares statistics are not well within the critical bounds. Furthermore, we apply Chow forecast test to examine the significance structural breaks in an economy for the period 1983-1984 due to industrialization in terms of consumer goods and restrictive trade policy. In this study, F-statistics computed in Table-6 suggests that there is no significant structural break in the economy during the sample period. The chow forecast test is more reliable and preferable than graphs [43]. This confirms that the ARDL estimates are reliable and efficient. The VECM Granger Causality Analysis The existence of long run (cointegration) relationship between the variables such energy consumption, trade openness, economic growth, capital-labour ratio and urbanisation intends us to apply the VECM Granger causality approach. The appropriate information about the direction of causality between the variables provides a clearer picture for policymakers to formulate trade, energy, economic and technological policy to sustain long run economic growth. The VECM Granger causality approach provides information about the causality between the variables both for long-and-short runs. The results of Granger causality test are reported in Table-7. The long run causality is indicated by the significance of coefficient of the one period lagged error-correction term ECTt 1 in equation (7) using t-test. The short run causality can be detected by the joint significance of LR test of the lagged explanatory variables in the equation. Our empirical results suggest that the ECTt 1 is having negative sign and statistically significance in all the VECM equations except in urbanisation VECM equation. The long run causality reveals that feedback effect is found between energy consumption and economic growth. This shows that energy consumption and economic growth are complementary. This shows that reductions in energy supply would retard economic growth. The energy exploration policies should be encouraged to sustain economic growth in long run. The bidirectional causality is found between trade openness and economic growth. This implies that energy exploration polices should be on priority and adoption of energy conservation policies is not only reduce trade but also it would inversely affect economic growth. Capital-labor ratio is Granger cause of energy consumption and same is true from opposite side. The feedback effect exists between capital-labor ratio and trade openness is found and same inference is drawn between economic growth and capital-labor ratio. Urbanization Granger causes economic growth, energy consumption, capital-labor ratio and trade openness in long run. In short run, we find that feedback effect is found between economic growth and energy consumption and same inference is drawn for energy consumption and capital-labour ratio. The bidirectional causality is found between trade openness and economic growth. Energy consumption and capital-labour ratio are Granger cause of each other. Trade openness and urbanisation are interdependent i.e. bidirectional causality exists. Urbanisation is Granger cause of economic growth. - 291 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. Table-7: The VECM Granger Causality Analysis Dependent Variable  ln Et  lnYt  ln KLt  lnTRt  lnUt Direction of Causality Short Run Long Run Joint Long-and-Short Run Causality  ln Et 1  ln Yt 1 lnKLt1 lnTRt 1  lnUt 1 ECTt 1 …. 14.6984* [0.0000] 3.0565** [0.0499] 0.0221 [0.9781] 0.7991 [0.4516] 14.2796* [0.0000] 0.5716 [0.5658] 22.4189* [0.0000] 3.8974** [0.0223] 0.4143* [0.6615] 3.3125** [0.0390] 10.9102* [0.0000] 0.0337 [0.3547] 5.8463* [0.0036] 10.1231* [0.0001] 0.5716 [0.5658] 0.4255 [0.6542] 0.1881 [0.8286] 2.7656*** [0.0660] -0.1573* [-5.1946] -0.0598* [-2.8605] -0.0733* [-3.5960] -0.0300** [-2.0700] …. 11.1644* [0.0000] 29.8893* [0.0000] 2.4387*** [0.0667] …. …. …. …. 9.5412* [0.0000] 0.2122 [0.8090] …. 0.1382* [0.8710]  ln Et 1 , ECTt 1  lnYt 1 , ECTt 1 18.4623* [0.0000] …. 9.6336* [0.0000] 4.4250* [0.0051] ….  ln KLt 1 , ECTt 1 10.6803* [0.0000] 16.9208* [0.0000] …. 8.1813* [0.0000] ….  lnTRt 1, ECTt 1  lnUt 1 , ECTt 1 9.0925* [0.0000] 6.6542* [0.0003] 15.6563* [0.0000] 9.5420* [0.0000] 3.2364* [0.0023] 8.1809* [0.0000] 3.2064* [0.0249] …. …. …. Note: *, ** and *** show significance at 1, 5 and 10 per cent levels respectively. 292 Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. V. Conclusion and Policy Implications Trade liberalization policies reduce or remove the tariff and non-tariff barriers in order to allow the exchange of goods and services without any barriers. In order to implementation of trade liberalization policies in the emerging and industrialized countries, it will attract foreign direct investment that accelerates economic growth via the creating the employment opportunities, transfer technology and improve the welfare in the country. In these processes, there are two types of gains from trade such as static and dynamic. Trade liberalization process helps to create a more competitive market through the reduction of cost due to economies of scale and efficiency. This process is known as the static gains from trade. On other hand, dynamic gains from trade are the transfer and adoption of better management practices and energy-efficient technologies. Trade openness consists of both exports and imports activities can influence the economic growth through the primary inputs in the production process such as energy. Exports-oriented countries will increase the demand for energy utilization for machineries, equipments and transportation of the exports-oriented products. The technology based energy intensive imported products like television, computer, automobile, air-conditioner and washing machines are increasing the demand for energy. According to the endogenous growth theory, imports-trade can be a potential channel to influence the economic growth by transferring technology and factors of production. This implies that trade openness enhances energy consumption via exports, imports and foreign direct investment (adoption of advanced energy-efficient technology) etc. This paper investigated the how much energy demand is affected by trade openness by incorporating economic growth and urbanization in energy demand function in the case of Malaysia. Theoretically trade openness affects energy consumption via scale and technique effects, composition effect. We applied the ARDL bounds approach to examine the long run relationship between the variables while dummy variable is included to capture the structural break arising in the series. Our results exposed that in the presence of structural breaks, cointegration exists between the variables. The findings unveiled that scale and technique effects have positive and negative impact on energy demand. The composition of factor of production also adds in energy consumption. Trade openness raises energy demand. A positive relationship from urbanization to energy consumption is found. In short run, trade openness declines energy consumption and urbanization increases energy demand. The causality analysis showed the bidirectional causality between trade openness and energy consumption. Economic growth and energy consumption are interdependent. The feedback effect is found between economic growth and trade openness. Energy consumption, economic growth, capital-labour ratio and trade openness are Granger cause of urbanisation. The Malaysian economy is growing very fast with the vision of a developed nation by 2020. Here economic growth will have high impacts on energy consumption due to more elastic relationship from income to energy consumption. However, due to the inelastic nature of energy consumption in response to trade openness, emphasizing on more openness in trade will lead to very low level of increase in energy usage. Thus, trade expansion policies like export promotion policies designed to increase exports will not increase energy consumption. Therefore, this paper suggests that in Malaysia the energy consumption policies should be driven based on the adoption of advanced energy saving technologies and achieving economies of scales, especially in the industrial and manufacturing production level. Government trade policies have cause directly on trade and energy use in Malaysia for the last 3 decades. The main factor is tariff, which affect Malaysia’s imports and exports trends. In the late of 1990s, almost 5% of all tariff lines had non ad valorem tariffs and this declined to 0.2% in early year 2000 because of the rationalization of the tariff structure by the government, followed by World Trade Agreement (WTO) in the middle of 1990s. This tariff features have directly increased energy for major economic sectors in Malaysia, which directly dealing with export based production activities, mainly manufacturing sectors related with export-promotion policies control by the government. Clearly, it can be seen that energy demand has increased dramatically along with trade volumes in the entire period of 2000 until 2010. On the other hand, increased on trades also has increased numbers of urban and export based industrial areas along the west coast of Peninsular Malaysia. Besides trade policy, Malaysian government also formulated numerous energy related policies in order to ensure the long-run reliability and security of energy supply for the nation [47]. - 293 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. Policies related to energy and environment in Malaysia are National Energy Policy (1979), National Depletion Policy (1980), Fuel Diversification Policy (1981, 1990), National Policy on the Environment (2002), National Green Technology Policy (2009) and Climate Change Policy (2009). All these policies have one fundamental objective, which is focusing on effective energy utilization along with environmental protections. Besides domestic policies, the Malaysian government also joins the ASEAN Power Grid and the Trans-ASEAN Pipeline Infrastructure Project that specially design to ensure the reserves for energy resources for ASEAN region. As been explained earlier through the findings of this study, trade is an important indicator that able to increase demands for energy in Malaysia. While the demand for energy is rapidly increased year by year, the Malaysian government should go for mix energy resource and not depending on single source of energy generation. Therefore, through this study we would like to suggest mix energy contribution for future energy resources for Malaysia. This energy resource formulation has been discussed in the previous Five-Fuel Diversification Strategy energy mix implementation in 1999, namely focusing on coal, gas, oil, hydro and renewable energy. The first 4 energy mixed implementation has moved smoothly through the sustainable energy development program, but renewable energy resources faced numerous challenges. For instance, renewable energy such as biomass, solar, hydrogen fuel cells and landfill gas not easy to be computed and obtained in Malaysia. [44] estimated more than 70 countries would use renewable energy technologies in the power sector by 2017. Demand for renewable energy utilization in Malaysia is low and there is also weak awareness among manufacturers on renewable energy compare to non-renewable energy. These are the main issues faced in energy utilization in Malaysia and therefore, we would suggest to policymakers in Malaysia to wider the usage of non-renewable energy to fulfill the demand from manufacturing sectors related to trades. Besides that, many policy implications could be drawn from this paper. First, Malaysian government should improve trade policies which based on energy efficiency and green energy products to reduce CO2 emissions. Secondly, government should appreciate green energy utilize manufacturer and emphasis tax reduction for them as an incentive. This may attract local and foreign investors in manufacturing sectors to increase the usage of non-renewable along with green energy in future. The Malaysian government should make effort to improve energy efficiency and the same moment reduces CO2 emissions. Although Malaysia already emphasis several energy policies, but future comprehensive energy efficiency policy focusing on supplying more renewable energy with clean and low carbon energy is needed. In future, impact of trade openness i.e. scale effect, technique effect, composition effect, comparative advantage on CO2 emissions can be examined following [23] in the case of Malaysia. More than 70% population of Malaysia is living in urban area due to industrialisation. This huge level of urbanisation and industrialisation has increased energy demand as well as level of CO2 emissions. The study would be augmented to investigate the impact of industrialisation and urbanisation on energy demand and CO2 emissions following [14] and, [45]. Other than that, there are several panel based cointegration and nonlinear approach can be undertaken by region using multivariate time series panel framework on this topic in future. Reference [1]. [2]. [3]. [4]. [5]. [6]. [7]. International Energy Agency, 2009.World Energy Outlook. IEA, Paris. Lean, H. H., and Smyth, R. (2010). Multivariate Granger causality between electricity generation, exports, prices and GDP in Malaysia. Energy, 35(9), 3640-3648. Siwar, C., Huda, N., and Hamid, A. (2009). Trade, Economic Development and Environment: Malaysian Experience. The Bangladesh Development Studies, 32(3), 19-39. Beaudreau, B. C. (1995). The impact of electric power on productivity: a study of US manufacturing 1950– 1984. Energy economics, 17(3), 231-236. Stern, D. I. (2000). A multivariate cointegration analysis of the role of energy in the US macroeconomy. Energy Economics, 22(2), 267-283. Wolde-Rufael, Y. (2005). Energy demand and economic growth: the African experience. Journal of Policy Modeling, 27(8), 891-903. Lee, C. C. (2006). The causality relationship between energy consumption and GDP in G-11 countries revisited. Energy Policy, 34(9), 1086-1093. - 294 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. [8]. [9]. [10]. [11]. [12]. [13]. [14]. [15]. [16]. [17]. [18]. [19]. [20]. [21]. [22]. [23]. [24]. [25]. [26]. [27]. [28]. [29]. [30]. [31]. [32]. [33]. Al-Iriani, M. A. (2006). Energy–GDP relationship revisited: an example from GCC countries using panel causality. Energy policy, 34(17), 3342-3350. Richmond, A. K., and Kaufmann, R. K. (2006). Is there a turning point in the relationship between income and energy use and/or carbon emissions? Ecological economics, 56(2), 176-189. Soytas, U., and Sari, R. (2006). Energy consumption and income in G-7 countries. Journal of Policy Modeling, 28(7), 739-750. Lorde, T., Waithe, K., and Francis, B. (2010). The importance of electrical energy for economic growth in Barbados. Energy Economics, 32(6), 1411-1420. Apergis, N., and Payne, J. E. (2009). Energy consumption and economic growth in Central America: evidence from a panel cointegration and error correction model. Energy Economics, 31(2), 211-216. Uddin, M. G. S., Alam, M. M., and Murad, M. W. (2011). An empirical study on income and energy consumption in Bangladesh. Energy Studies Review (ISSN: 0843-4379), 18, 1. Shahbaz, M., and Lean, H. H. (2012). Does financial development increase energy consumption? The role of industrialization and urbanization in Tunisia. Energy policy, 40, 473-479. Shahbaz, M., Zeshan, M., and Afza, T. (2012). Is energy consumption effective to spur economic growth in Pakistan? New evidence from bounds test to level relationships and Granger causality tests. Economic Modelling, 29(6), 2310-2319. Ghani, G. M. (2012). Does trade liberalization effect energy consumption?. Energy Policy, 43, 285-290. Sadorsky, P. (2012). Energy consumption, output and trade in South America. Energy Economics, 34(2), 476488. Narayan, P. K., and Smyth, R. (2009). Multivariate Granger causality between electricity consumption, exports and GDP: evidence from a panel of Middle Eastern countries. Energy Policy, 37(1), 229-236. Lean, H. H., and Smyth, R. (2010). On the dynamics of aggregate output, electricity consumption and exports in Malaysia: evidence from multivariate Granger causality tests. Applied Energy, 87(6), 1963-1971. Sadorsky, P. (2011). Trade and energy consumption in the Middle East. Energy Economics, 33(5), 739-749. Shahbaz, M., Lean, H. H., and Farooq, A. (2013). Natural gas consumption and economic growth in Pakistan. Renewable and Sustainable Energy Reviews, 18, 87-94. Werner, A., Brian, C., and Scott, T. (2001). Is free trade good for the environment. American Economic Review, 91(4), 877-908. Cole, M. A. (2006). Does trade liberalization increase national energy use?. Economics Letters, 92(1), 108112. Hossain, M. S. (2011). Panel estimation for CO 2 emissions, energy consumption, economic growth, trade openness and urbanization of newly industrialized countries. Energy Policy, 39(11), 6991-6999. Liebman, B. (2007). Malaysia leads the world in greenhouse gas emission growth rate. Curb Global Warming. Nov, 29. Islam, F., Shahbaz, M., Ahmed, A. U., and Alam, M. M. (2013). Financial development and energy consumption nexus in Malaysia: a multivariate time series analysis. Economic Modelling, 30, 435-441. Kneller, R., Morgan, C. W., and Kanchanahatakij, S. (2008). Trade liberalisation and economic growth. The World Economy, 31(6), 701-719. Ghani, G. M. (2011). The impact of trade liberalisation on the economic performance of OIC member countries. Journal of Economic Cooperation and Development, 32(1), 1. Tang, C. F., and Tan, B. W. (2014). The linkages among energy consumption, economic growth, relative price, foreign direct investment, and financial development in Malaysia. Quality and Quantity, 48(2), 781797. Phillips, P. C., and Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335346. Kwiatkowski, D., Phillips, P. C., Schmidt, P., and Shin, Y. (1992). Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root?. Journal of econometrics, 54(1-3), 159-178. ELLIOrr, G., Rothenberg, T. J., and Stock, J. H. (1996). Efficient tests for an autoregressive unit root. Econometrica, 64(4), 813-836. Shahbaz, M., and Lean, H. H. (2012). The dynamics of electricity consumption and economic growth: A revisit study of their causality in Pakistan. Energy, 39(1), 146-153. - 295 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296. Shahbaz et al., / Bulletin of Energy Economics, 4(3), 280-296. [34]. Ng, S., and Perron, P. (2001). Lag length selection and the construction of unit root tests with good size and power. Econometrica, 69(6), 1519-1554. [35]. Zivot, E., Andrews, D., 1992. Further evidence of great crash, the oil price shock and unit root hypothesis. Journal of Business and Economic Statistics 10, 251–270. [36]. Pesaran, M. H., and Shin, Y. (1998). An autoregressive distributed-lag modelling approach to cointegration analysis. Econometric Society Monographs, 31, 371-413. [37]. Pesaran, M. H., Shin, Y., and Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of applied econometrics, 16(3), 289-326. [38]. Narayan, P. K. (2005). The saving and investment nexus for China: evidence from cointegration tests. Applied economics, 37(17), 1979-1990. [39]. Dickey, D. A., and Fuller, W. A. (1981). Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica: Journal of the Econometric Society, 1057-1072. [40]. Corden, W. M. (1996). Pragmatic Orthodoxy: macroeconomic policies in seven East Asia economies (Vol. 60). International Center for economic growth. [41]. Athukorala, P. C. (2005). Trade policy in Malaysia: Liberalization process, structure of protection, and reform agenda. ASEAN Economic Bulletin, 21(1), 19-34. [42]. Lütkepohl, H. (2006). Structural vector autoregressive analysis for cointegrated variables. In Modern Econometric Analysis (pp. 73-86). Springer Berlin Heidelberg. [43]. Ghin, Y. L. (2004). A reexamination of the role of exports in Malaysia's economic growth: after Asian financial crisis, 1970-2000. International Journal of Management Studies (IJMS), 11, 79-104. [44]. International Energy Agency, 2012. Renewable Energy Outlook, IEA, Paris. [45]. Poumanyvong, P., and Kaneko, S. (2010). Does urbanization lead to less energy use and lower CO 2 emissions? A cross-country analysis. Ecological Economics, 70(2), 434-444. [46]. Engle, R. F., and Granger, C. W. (1987). Co-integration and error correction: representation, estimation, and testing. Econometrica: journal of the Econometric Society, 251-276. [47]. Mohamed, A. R., and Lee, K. T. (2006). Energy for sustainable development in Malaysia: Energy policy and alternative energy. Energy policy, 34(15), 2388-2397. [48]. Granger, C. W. (1969). Investigating causal relations by econometric models and cross-spectral methods. Econometrica: Journal of the Econometric Society, 37, 424-438. [49]. Tenth Malaysia Plan (2011-2015). Government Printers, Kuala Lumpur. - 296 - Citation: Shahbaz, M., Alam, M., Uddin, G. S. and Nanthakumar, L. (2016). The Effect of Scale, Technique, Composition and Trade Openness on Energy Demand: Fresh Evidence from Malaysia. Bulletin of Energy Economics, 4(3), 280-296.