We investigate the relation between financial development, energy consumption, and economic growth in the economy of Lebanon over the period 2000M2–2010M12. Our findings confirm the existence of cointegration among the variables. The results indicate that financial development and energy consumption contribute to economic growth in Lebanon. The impact of energy consumption on economic growth is positive showing the significance of energy as a main stimulant of economic growth. Financial development is also found to play a vital role in enhancing economic growth. Financial development and economic growth also result in further increase in energy consumption. We offer some policy implications specific to Lebanon considering the recent discovery of large oil and gas reserves in the country and the historical importance of its banking sector which remains a center of Lebanon’s service-oriented economy.
1. Introduction
Energy economics has drawn substantial attention from academicians in recent times. A number of studies have investigated the causal relationship between energy consumption and economic growth. This issue is important because energy drives the wheels of economic growth since it is a key factor of production, along with capital, and labor. In addition, the higher the GDP per capita, the more the energy demand, which is a relation that is intuitively appealing. The pioneering study by J. Kraft and A. Kraft [1] confirms this by providing evidence of unidirectional causality from GNP to energy use for the US over the period 1947–1974.
Much of the research aimed at exploring the long run relationship and the direction of causality between economic growth and energy use has included several other variables, for example, population, urbanization, financial development, and so forth, to better understand the underlying dynamics of the relation. Lee et al. [2] included capital stock and labor to explain energy use for some Asian nations. They found that the positive link between economic growth and energy demand gets stronger as relevant variables are included. Ang [3] explored the dynamic causal relationships between GDP and energy consumption in France. He found that economic growth influences energy consumption (and pollution) in the long run, but the relation reverses in the short run in case of France. Apergis and Payne [4–6] and Wolde-Rufael [7] argued that rise in energy demand in African countries is closely linked to income. Population growth creates pressure on rural resources, forces people to move to urban areas, and thus increases energy demand. For sustained economic growth, the increased energy demand over a long period must be met from new sources, or by developing cost-effective alternative energy. Using both bi- and multivariate models for New Zealand, Bartleet and Gounder [8] found causality running from real GDP to energy use. They also found that real GDP and employment exert significant impact on capital formation, where capital stock plays an important role in determining the direction of causality.
The objective of this paper is to examine the existence of long run relationship among energy consumption (lnEt), financial development (lnFt), and economic growth (lnYt) for Lebanon using monthly data over the period of 2000M2–2010M12. The ARDL bounds testing approach to cointegration is used to examine a long run relation among the series. The innovative accounting technique (variance decomposition approach and impulse response function) (the methodological framework is based on the studies by Shan [9] and Shahbaz et al. [10]) shows the response of dependent variable to shocks arising in independent variables. We use this method to examine the short run dynamics and the direction of causality. This method is helpful in determining the relative strength of causal relation beyond the chosen time frame [9, 11] and the magnitude of the feedback among the series.
The rest of the paper is organized as follows. Section 2 reviews the relevant literature with focus on the link between economic growth and energy consumption, economic growth and financial development, and financial development and energy consumption. Section 3 discusses the motivation of this study. Section 4 describes the data and estimation strategy. Results of the study are reported in Section 5. Conclusion and policy implications are reported in Section 6.
2. Literature Review
The question of whether energy consumption and/or financial development sway the rate of economic growth of a country or a region has shaped an important query among economists in the literature for some time. This interest is driven primarily from the important policy implications that can be obtained from such studies in relation to the desired action(s) that can accelerate the rate of economic growth and prosperity of a country. Empirical studies in this regard, however, provided conflicting results so economists’ views on this issue have not been unanimous. Below, we will provide a brief review of the studies that addressed the impacts of these variables on economic growth.
2.1. Economic Growth and Energy Consumption
Energy forms the lifeblood of the world economy as it is an essential input to producing almost all of the goods and services of the modern global economy. It contributes to economic growth directly as it creates jobs and value associated with extracting, transforming, and distributing of energy. Furthermore and more importantly, this sector’s activities relate to and strengthen the rest of the economy as energy forms an input for almost all production processes of goods and services. Supply interruptions of many sources of energy are known to have a great impact as they can harshly impact the economies of almost all countries. In addition, stable and lower energy prices are known to help stimulate the growth rate of any economy. This is because lower energy prices result in increasing disposable income for consumers and lowering costs for firms. The resulting improved profit margins for firms and higher disposable income for consumers provide incentives for accelerated rates of growth.
One can distinguish between four different hypotheses explaining relationship between the energy consumption and economic growth based on the type of the relationship between both variables [12].
Neutrality hypothesis implies no causality between energy consumption and economic growth. This means that energy conservation may not adversely affect economic growth.
Conservation hypothesis advocates for an implementation of conservative energy policy, as economic growth would not be slowed down by conservation. Economically, this means that causality is unidirectional running from economic growth to energy consumption.
Growth hypothesis is supported when unidirectional causality is running from energy consumption to economic growth. The latter indicates that energy conservation may reduce investment and negatively influence economic growth.
Feedback hypothesis confirms the interdependence between energy consumption and economic growth since both variables affect each other. This encourages the implementation of energy expansionary policies for long run sustainable economic growth.
The energy growth nexus has been extensively investigated since the pioneer work of J. Kraft and A. Kraft [1]. The authors examined the relationship between energy consumption and economic growth in case of the United States. The empirical results revealed that economic growth Granger causes energy consumption. Later on, a large number of empirical studies using different approaches, time periods, and proxy variables have tested this causal relationship in a number of countries. Abosedra and Baghestani [13], Cheng and Lai [14], Soytas and Sari [15], Oh and Lee [16], Jumbe, [12], Fatai et al. [17], Lee et al. [2, 18–20], Chontanawat et al. [21], Narayan and Smyth [22], Apergis and Payne [4], and Bowden and Payne [23] among others have examined this issue for different countries and over various sample periods. Lines of evidence from these empirical studies are still mixed at best and controversial results in terms of the direction of causality and the strength of impact of energy use on economic growth are reported. Some papers documented unidirectional causality from energy consumption to economic growth (growth hypothesis). Narayan and Smyth [22], Apergis and Payne [4], Odhiambo [24], Bowden and Payne [23], Tsani [25], Wang et al. [26], and Yazdan and Hossein [27] found evidence supporting this view.
Other researchers have documented unidirectional causality running from economic growth to energy consumption (conservation hypothesis). This hypothesis is supported by Lise and van Montfort [28], Erdal et al. [29], Huang et al. [30], Mallick [31], Sa’ad [32], Binh [33], Qazi et al. [34], and Soile [35]. Next, the feedback hypothesis is supported if bidirectional causality between energy consumption and growth is found. This is supported by Glasure [36], Lee and Chang [20], Belke et al. [37], Eggoh et al. [38], Marques et al. [39], Kaplan et al. [40], Shahbaz et al. [41], Fuinhas and Marques [42], Zeshan [43], and Shahbaz et al. [10]. The final hypothesis is that of neutrality, where no causality between energy consumption and growth is found, and it is supported by Soytas et al. [44] and Gross [45]. In the case of Lebanon, Dagher and Yacoubian [46] applied the cointegration developed by Johansen [47] and Granger causality by Toda and Yamamoto [48] as well as the VECM Granger causality to examine cointegration and causality relationship between energy consumption and economic growth. Their results indicated that long run relationship exists between the variables and energy consumption and economic growth are bidirectional Granger cause.
2.2. Financial Development and Economic Growth
Since pioneering works of Schumpeter [49], of Goldsmith [50], and recently of McKinnon [51] and Shaw [52], the relation between financial development and economic growth has attracted interest of both theorists and practitioners. Demirguc-Kunt and Levine [53] and Levine [54] provided an extensive literature survey on this topic. Güryay et al. [55], Wolde-Rufael [7, 56], and Shahbaz [57], among others, have used both cross-country and time series data to investigate the relationship between financial development and economic growth.
Financial development has a positive effect on economic growth as it may increase the efficiency of capital accumulation [50] and/or augment the level of saving and consequently investment [51, 52]. By increasing the size of savings and improving the efficiency of investment, financial development leads to higher economic growth [54, 58, 59]. Furthermore, financial development would support financial innovation and promote the adoption of advanced technology [57, 60]. In the related literature, this is known as “supply-leading.” In other words, financial development causes economic growth in the sense of Granger. Ibrahim [61] observed that financial development stimulates economic growth in Malaysia. Jalil and Ma [62] found that financial development contributes to economic growth by increasing capital formation in Pakistan and China. Shahbaz [57] showed the same results in case of Pakistan. Coccorese [63] reported that economic growth is Granger cause of financial development. Masih et al. [64] applied long run structural modeling (LRSM) to examine the causality between financial development and economic growth in Saudi Arabia. They validated the existence of supply-side hypothesis. Kar and Mandal [65] also noted that financial development promotes economic growth by enhancing capitalization in India.
“Demand-following” hypothesis suggests that financial development follows economic growth. Patrick [66] pointed out that “demand-following” relationship indicates that real economic activity Granger causes financial development by generating demand for financial services as an economy develops. Liang and Teng [67] investigated the relationship between financial development and economic growth in China using VAR approach. They noted cointegration between the variables and economic growth Granger causes financial development. Fung [68] finds that positive impact of economic growth on financial development is due to productivity boost. Chukwu and Agu [69] also supported demand-side hypothesis in case of Nigeria. Amarathunga [70] also confirmed the presence of demand-side hypothesis in case of Sri Lanka. In case of South Africa, Odhiambo [71] investigated the relationship between financial development and economic growth by applying the ARDL bounds testing and the VECM Granger causality approaches. The results indicated that the variables are cointegration and financial development is Granger cause of economic growth. Odhiambo [72, 73] investigated the causality between financial development, foreign capital inflows, and economic growth for Tunisia. The empirical results revealed that financial development follows economic growth and that financial development and foreign capital inflows are interdependent. Hassan et al. [74] also reported unidirectional causality running from economic growth to financial development in developing economies. Odhiambo [72, 73] used trivariate model to examine causality between financial development and economic growth by incorporating foreign capital inflows. The empirical evidence validated the existence of demand-side hypothesis in Tanzania.
The feedback effect between financial development and economic growth is also found in the existing literature. For example, Ilhan [75] applied cointegration and error correction method for the relationship between financial development and economic growth. He noted that cointegration exists and that the feedback effect is validated in case of South Africa. Zheng et al. [76] used bivariate framework model and reported that financial development and economic growth are complementary in case of China. The same exercise was conducted by Al-Malkawi et al. [77] in case of UAE and results found a feedback effect between financial development and economic growth. Eslamloueyan and Sakhaei [78] applied generalized least square (GLS) method with cross-section seemingly unrelated regression (SUR) to probe the relationship between financial development and economic growth and confirmed the findings of Husam-Aldin et al. [79]. On the contrary, Bakhouche [80] reported that financial development does not promote economic growth and that economic growth does not contribute to financial development, that is, neutral hypothesis. Gantman and Dabós [81] also supported the view by Bakhouche [80] that financial development and economic growth are independent.
2.3. Financial Development and Energy Consumption
Love and Zicchino [82] reported that financial development impacts on real interest can possibly result in an increase in investment. This in turn can promote economic growth and generate employment opportunities which further increase income. Such impact will increase purchases by consumers especially of durable items [83, 84] which add further to energy use. This shows that the linkages between energy use and economic growth are better understood when we go beyond a simple bivariate framework. Karanfil [85] suggested using stock market capitalization, liquid liabilities, and domestic credit to the private sector, each as share of GDP among the financial variables. Yandan and Lijun [86] examined the impact of financial development on primary energy consumption in Guangdong (China). Their study finds Granger causality running from energy consumption to financial development, while the reverse is insignificant. Sadorsky [83] examined 22 emerging economies (1990–2006) using different indicators of financial development. This included FDI, bank deposits as share of GDP, stock market capitalization as share of GDP, stock market turnover ratio, and total stock market value traded as a share of GDP. His results confirmed that energy consumption is positively linked to economic growth but the impact is small.
Sadorsky [87] investigated the impact of financial development on energy consumption using data of 9 Central and Eastern European frontier economies. He reported that financial development increases energy demand once deposit money and bank assets to GDP, financial system deposits to GDP, liquid liabilities to GDP, and stock market capitalization are used as measures of financial development. Similarly, Shahbaz and Lean [88] examined energy demand for Tunisia and reported results showing that financial development increases energy demand resulting from economic growth. In the case of Malaysia, Tang and Tan [89] examined the relationship between financial development and energy consumption by incorporating relative prices and foreign direct investment (FDI) in energy demand function. They report bidirectional causality between financial development and energy consumption both in the short and the long runs. Islam et al. [90] reported that financial development, economic growth, and population are driving forces to increase energy demand in Malaysia. The feedback effect is also reported between financial development and energy consumption in the long run but financial development Granger causes energy demand in short run.
3. Motivation of the Study
This paper provides an investigation of the relationship between economic growth, financial development, and energy consumption using monthly data over the period of 1993–2010 in case of Lebanon. This is a country with a sectarian-based parliamentary republic located in the Middle East, with a population of approximately 4.2 million. The Lebanese economy is service oriented, with tourism and banking sectors being the main driving force, contributing over 70% of GDP and therefore considered the primary sectors for growth. The banking sector is one of the main pillars of Lebanese economy with a size equivalent to 350% of GDP as of 2009. Lebanese banks benefit from a strong net inflow from both expats and the Gulf States; domestic private sector credit growth has been 19% until October 2011 and the banking system’s foreign assets have also grown rapidly, supporting the fact that the banking industry is not affected by the political unrest. About 18% of GDP is contributed by the industrial sector and about 5% by the agriculture. Net remittances from the Lebanese Diaspora living abroad, mainly in the Gulf region, also contribute 5% to the GDP. Tourism industry development in Lebanon dates back to the 1960s when Lebanon’s capital, Beirut, was known as “the Paris of the Middle East.” This sector contributes notably to the employment in the economy. Employment in that sector as a share of total employment stood at 31.2% in 2005 and is estimated at 38% in 2010. This is not surprising given that the percentage of tourism and travel in GDP stood at 31.2% in 2005 and is estimated at 37.6% in 2010 [91]. This is not unexpected in a country that is known for its diverse atmosphere, earliest history, ancient Roman ruins, preserved castles, and notable mosques and churches, as well as its stunning beaches in the Mediterranean Sea and rugged ski resorts.
The country’s economy has faced much challenge owing to its continuous political unrest. The civil war (1975–1990) had a heavy unconstructive impact on the nation, causing the country to have a high budget deficit. Even more recent, the assassination of ex-Prime Minister Rafik Hariri in February 2005, the July 2006 war between Lebanon and Israel, the sit-ins, protests and clashes between the opposing government alliances in 2006 till 2008, and the constant instability and corruption within Lebanon contributed to the huge deficit and the increase in sectarianism. In 2010, growth in Lebanon was stimulated by rising nonresident deposits, an elevated number of tourist arrivals, and a vigorous real estate market. However, due to the regional political chaos, both tourist arrivals and expat’s housing demand in the real estate sector have slowed down in 2011. Furthermore, the continued unrest in Syria seems to challenge Lebanon’s economic prospects in 2012, especially when 25% of Lebanon’s exports are to Syria and 11% of its imports are from Syria.
We think that examining the possible linkages between economic growth, financial development, and energy consumption in Lebanon is justified and needed for three reasons. First, the above features of Lebanese economy justify the need for this study as it will assist the policy makers in the country in assessing their priorities for resource allocations for the country’s development. Second, the clear interface of tourism and energy use, in Lebanon, is another important factor for policy makers to consider given the outages of electricity and shortages of some fuels in the country. Third, the authors are not aware of any study of this issue for Lebanon with the exception of that of Dagher and Yacoubian [46]. Our study improves upon theirs, however, since we do not employ a bivariate framework as they did and since our sample is larger than theirs (1980–2009) and excludes the years of the civil war in Lebanon (1975–1990).
The findings should help better understand this relationship that underlies energy use, financial development, and economic growth nexus for Lebanon which will help identify an appropriate policy mix for these sectors in future economic planning for economic growth of that country.
4. The Data and Estimation Strategy
In this study our primary interest lies in the energy consumption-economic growth nexus, financial development-economic growth nexus, and energy consumption-financial development nexus.
The prime hypothesis of energy consumption-economic growth causality postulates that economic growth is impeded by energy conservation policies if causality runs from energy consumption to economic growth. Energy conservation policies do not have adverse impact on economic growth if causality is running from economic growth to energy consumption or no causality is found between both of the variables.
The second hypothesis deals with financial development and economic growth. Financial development boosts investment activities by directing financial resources to new and existing potential ventures which not only enhances domestic production but also raises the rate of economic growth. This implies that financial development drives economic growth. This unidirectional causality running from financial development to economic growth is called supply-side hypothesis. The rise in per capita income or economic growth will increase the demand of financial services for both customers (consumption purpose) and producers (investment purpose) which as a result raises financial development. This shows that economic growth leading to financial development is called demand-side hypothesis.
The third hypothesis deals with energy consumption and financial development nexus. Financial development results in an increase in funds availability to investors for new and existing investment and to consumers to purchase ticket items which directly can increase energy demand. In turn, a rise in demand for financial services associated with financial development can lead to increased demand for energy. On the other hand, an increase in energy consumption as a result of an increase in the level of income may also lead to further financial development as a larger percentage of the population is exposed to the need to use financial services. Finally financial development and energy use may be complementary in nature as a feedback effect may exist.
4.1. The Data
We have monthly frequency data over the period of 2000M2–2010M12. We use M2 as a measure of financial development (see Adu et al. [92] for more details). Economic growth is measured by the index of coincident indicator as a measure for real economic activity in Lebanon. The coincident indicator is a broad measure for real economic activity in Lebanon. It is based upon a linear combination of a set of indirect indicators. These consist of imports of petroleum products (18.2 per cent), electricity production (18.6 per cent), cement deliveries (16.5 per cent), number of foreign passengers (11 per cent), total international trade (11.8 per cent), value of checks clearance (12 per cent), and M3 money supply (12 per cent). Finally, we use energy use in millions of KWH as proxy energy consumption. We convert all series into natural logarithm to avoid sharpness and variations in the data. The log-linear specification provides efficient results as compared to simple linear specification. Ft denotes financial development, energy consumption is indicated by Et, and Yt is for economic growth in time period t. The data of M2 and index of coincident indicator are obtained from The Central Bank of Lebanon while energy consumption data is obtained from the Central Administration of Statistics-Lebanon.
4.2. The ARDL Bounds Testing Approach
We employ the autoregressive distributed lag (ARDL) bounds testing approach to cointegration developed by Pesaran et al. [93] to explore the existence of long run relationship between economic growth, financial development, and energy consumption in the presence of structural break. This approach has multiple econometric advantages. The bounds testing approach is applicable irrespective of whether the variables are I(0) or I(1). The results of the ARDL approach are insensitive small sample data test. 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 long run information. The UECM is expressed as follows:
(1)ΔlnEt=α1+αTT+αElnEt-1+αYlnYt-1+αFlnFt-1+∑i=1pαiΔlnEt-i+∑j=0qαjΔlnYt-j+∑k=0rαkΔlnFt-kαDD1+μt,ΔlnFt=α1+αTT+αElnEt-1+αYlnYt-1+αFlnFt-1+∑i=1pϑiΔlnFt-i+∑j=0qϑjΔlnEt-j+∑k=0rϑkΔlnYt-k+ϑDD3+μt,ΔlnYt=α1+αTT+αElnEt-1+αYlnYt-1+αFlnFt-1+∑i=1pβiΔlnEt-i+∑j=0qβjΔlnYt-j+∑k=0rβkΔlnFt-k+βDD2μt,
where Δ is the first difference operator, lnYt is natural log of index of coincident indicator (economic growth), lnFt is natural log of M2 (financial development), and lnEt is natural log of energy consumption. D is a dummy for structural break point and μt is error term assumed to be independently and identically distributed. The optimal lag structure of the first differenced regression is selected by the Akaike information criteria (AIC). Pesaran et al. [93] suggest F-test for joint significance of the coefficients of the lagged level of variables. For example, the null hypothesis of no long run relationship between the variables is H0:αE=αF=αY=0 against the alternative hypothesis of cointegration Ha:αE≠αF≠αY≠03. Accordingly, Pesaran et al. [93] compute two sets of critical values (lower and upper critical bounds) for a given significance level. Lower critical bound is applied if the regressors are I(0) and the upper critical bound is used for I(1). If the F-statistic exceeds the upper critical value, we conclude in favor of a long run relationship. If the F-statistic falls below the lower critical bound, we cannot reject the null hypothesis of no cointegration. However, if the F-statistic lies between the lower and upper critical bounds, inference would be inconclusive. When the order of integration of all the series is known to be I(1), then decision is made based on the upper critical bound. Similarly, if all the series are I(0), then the decision is made based on the lower critical bound. To check the robustness of the ARDL model, we apply diagnostic tests. These tests are used to check for normality of error term, serial correlation, autoregressive conditional heteroskedasticity, white heteroskedasticity, and the functional form of the empirical model.
4.3. The VECM Granger Causality Test
After examining the long run relationship between the variables, we use the Granger causality test to determine the causality between the variables. If there is cointegration between the series, then the vector error correction method (VECM) can be developed as follows:
(2)ΔlnEtΔlnYtΔlnFt=b1b2b3b4+B11,1B12,1B13,1B21,1B22,1B23,1B31,1B32,1B33,1×ΔlnEt-1ΔlnYt-1ΔlnFt-1+⋯+B11,mB12,mB13,mB21,mB22,mB23,mB31,mB32,mB33,m×ΔlnEt-1ΔlnYt-1ΔlnFt-1+ζ1ζ3ζ3×ECMt-1+μ1tμ2tμ3t,
where difference operator is Δ and ECMt-1 is the lagged error correction term, generated from the long run association. The long run causality is found by examining the significance of coefficient of lagged error correction term using t-test statistic. The existence of a significant relationship in first differences of the variables provides evidence on the direction of short run causality. The joint χ2 statistic for the first differenced lagged independent variables is used to test the direction of short run causality between the variables. For example, a12,i≠0∀i shows that economic growth Granger causes energy consumption and economic growth is Granger of cause of energy consumption if a11,i≠0∀i.
5. Results of the Study
The results of descriptive statistics and correlation matrix are reported in Table 1. The Jarque-Bera test statistics reveal that the series of economic growth, energy consumption, and financial development have normal distributions. Our empirical evidence finds that correlations between the variables are positive and strong. For instance, a positive correlation is found between energy consumption and economic growth. Financial development and economic growth are positively correlated and financial development and energy consumption also have positive correlation. The normal distribution of the series leads us to proceed for further analysis.
Descriptive statistic and correlation matrix.
Variable
Mean
Median
Maxi.
Mini.
Skewness
Kurtosis
Jarq.Bera
lnEt
lnYt
lnMt
lnEt
6.7773
6.7719
7.0707
6.5132
0.2122
2.7366
1.3722
1.0000
lnYt
5.1574
5.1439
5.5861
4.7858
0.4778
2.3847
0.7105
0.7197
1.0000
lnMt
10.1706
10.1084
10.9920
9.6700
0.9432
3.0181
1.9575
0.7017
0.9118
1.0000
The next step is to test the unit root properties of the variables. The stationarity level of the variables is very important for policy implications. For example, if energy consumption series is stationary at level, it shows that innovations in energy use have transitory effects and series returns to its trend path; otherwise, innovations show permanent effect on energy consumption if energy consumption series is nonstationary. Similarly, the impact of financial policies adopted to improve financial sector efficiency has temporary effect on financial development if financial development series is stationary at level. Financial policies will have permanent impact on financial development if series is found to be integrated at I(1). The shocks to economy by economic policies have permanent effects if economic growth series is nonstationary which implies that fiscal and/or monetary or any other stabilization policies would only have permanent effects on the real output levels. If economic growth series is stationary, then shocks to economy have transitory effect. Therefore, it is necessary to check the order of integration of the variables before applying the ARDL bounds testing to investigate the long run relationship among the series of interest. ADF and Ng-Perron unit root tests were therefore applied and results are disclosed in Table 2. The results reveal that energy consumption, economic growth, and financial development are nonstationary at level. All the variables are stationary at 1st difference with intercept and trend. This indicates that the series have unique order of integration, that is, I(1).
Unit root analysis.
Variable
ADF unit root at level
ADF unit root at 1st difference
T-statistic
Prob. value
T-statistic
Prob. value
lnEt
−2.9832 (3)
0.1394
−11.3974 (2)*
0.0000
lnFt
−2.2992 (4)
0.4321
8.8743 (4)*
0.0000
lnYt
−1.9300 (4)
0.2547
−17.4153 (3)*
0.0000
Variable
Ng-Perron unit root test
MZa
MZt
MSB
MPT
lnEt
−8.0367 (5)
−1.9982
0.2486
11.3579
lnFt
−4.2960 (4)
−1.4656
0.3411
21.2112
lnYt
−9.6893 (5)
−2.1405
0.2209
9.6774
ΔlnEt
−46.2855 (2)*
−4.8075
0.1038
1.9851
ΔlnFt
−25.3303 (4)
−3.5588
0.1405
3.5975
ΔlnYt
−24.4338 (5)*
−3.4612
0.1416
3.9365
Note: * represents significance at 1% levels. Lag order is shown in parentheses.
However, the results of these traditional tests may be biased. These unit root tests do not have information about unknown structural break occurring in the series. The appropriate information about unknown structural breaks would help policy makers in designing a comprehensive energy, financial, and economic policy to enhance economic growth for the long run by considering these structural breaks. This issue is resolved by applying Clemente-Montanes-Reyes unit root test with single and two unknown structural breaks arising in the variables. Our empirical exercise indicated that all the series are nonstationary at level with single structural break in energy consumption, financial development, and economic growth in 2008M1, 2009M1, and 2006M6, respectively. (In this regard, we note that the 2006 Lebanon War started on July 12, 2006, and continued until a United Nations-brokered cease fire on August 14, 2006. Furthermore, Lebanon witnessed a series of protests and sit-ins that began on December 1, 2006. This was led by groups in Lebanon that opposed the US and Saudi-backed government of Prime Minister Fouad Siniora. This ended on May 21, 2008, following the Doha Agreement. On January 25, 2008, a bombing in the Lebanese capital, Beirut, killed a senior intelligence officer, who was involved in the investigation of assassination of former Prime Minister Rafiq Hariri who was killed in 2005. This was followed by a series of bombings and assassinations which have struck Lebanon, most of them occurring in and around the capital, Beirut, during the last few years. Finally, we note that 27 January 2009 marked a historical event in which Syria accepted Lebanon’s first ambassador ever to Damascus.) We conclude that all the variables are stationary at first difference accommodating single and two unknown structural breaks confirmed by Clemente-Montanes-Reyes unit root test (see Table 3).
Clemente-Montanes-Reyes structural break unit root analysis.
Model: trend-break model
Variable
Level data
First difference data
TB1
TB2
Test statistics
K
TB1
TB2
Test statistics
K
lnEt
2008M4
—
−2.053
3
2006M6
—
−8.346*
2
2004M4
2008M4
−2.909
1
2006M6
2009M8
−7.510*
1
lnFt
2009M1
—
−2.372
3
2005M1
—
−4.224**
5
2003M1
2008M9
−3.487
4
2005M1
2008M1
−5.643**
6
lnYt
2006M6
—
−2.450
3
2006M6
—
−6.465*
5
2003M9
2008M3
−5.270
4
2006M5
2006M9
−5.711**
6
Note: TB1 and TB2 are the dates of the structural breaks; K is the lag length; * and ** show significance at 1% and 5% levels, respectively.
Given that, the computation of ARDL bounds testing is known to be sensitive to lag length selection. As such, inappropriate selection of lag length may produce biased results. Therefore, it is necessary to have exact information about lag order of the series to avoid the problem of biasedness of ARDL F-statistics [94]. We follow AIC criteria for selection of lag length where we found that lag order 4 is suitable for our data sample. The information about lag order is given in Table 4 following AIC criterion.
VAR lag order selection criteria.
VAR lag order selection criteria
Lag
logL
LR
FPE
AIC
SC
HQ
0
245.5326
NA
4.02e-06
−3.911816
−3.843584
−3.884099
1
585.0314
657.0944
1.94e-08
−9.242442
−8.969512*
−9.131571
2
600.3055
28.82361
1.76e-08
−9.343637
−8.866008
−9.149613
3
615.0278
27.07002*
1.60e-08
−9.435932
−8.753605
−9.158755*
4
624.1486
16.32916
1.60e-08*
−9.437880*
−8.550856
−9.077550
5
627.3032
5.495216
1.76e-08
−9.343600
−8.251878
−8.900117
6
633.6844
10.80681
1.85e-08
−9.301361
−8.004941
−8.774725
7
635.0608
2.264400
2.10e-08
−9.178400
−7.677282
−8.568610
8
642.5061
11.88850
2.16e-08
−9.153324
−7.447509
−8.460382
* 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.
Table 5 reports the results of the ARDL bounds testing approach to cointegration. We followed the critical bounds produced by Pesaran et al. [93]. The critical bounds generated by Pesaran et al. [93] are suitable for large sample size (T=500 to T=40,000). Our findings reveal that calculated F-statistics seem to exceed upper critical bounds at 1% and 5%, respectively, once we treated energy consumption, financial development, and economic growth as dependent variables. This shows that there are three cointegrating vectors confirming the existence of long run relationship among the series in the presence of structural breaks. The ARDL models fulfill the assumptions of normality, ARCH, and functional forms of models. The findings note that error terms are normally distributed, there is no evidence of ARCH, and models are well articulated. (Although the ARDL bounds testing is widely used to analyze the relation between energy consumption and economic growth while ignoring the role of structural breaks in the series, we have overcome this issue by including dummy for structural break points. This confirms that our finds are more reliable and consistent than previous ones.)
The results of ARDL cointegration test.
Bounds testing to cointegration
Diagnostic tests
Estimated models
Optimal lag length
F-statistics
Structural break
χNORMAL2
χARCH2
χRESET2
FE(E/F,Y)
4, 4, 4
6.577**
2008M4
0.8936
[1]: 0.3263
[3]: 1.7586
FF(F/E,Y)
4, 4, 4
4.730**
2009M1
0.1449
[1]: 0.2831
[2]: 0.4828
FY(Y/E,F)
4, 4, 4
8.719*
2006M6
0.1229
[1]: 0.2005
[1]: 0.0237
Significant level
Critical values (T=132)
Lower bounds I(0)
Upper bounds I(1)
1 per cent level
4.40
5.70
5 per cent level
3.47
4.57
10 per cent level
3.03
4.06
Note: * and ** denote the significance at 1 and 5 per cent levels, respectively. The optimal lag length is determined by AIC. [] is the order of diagnostic tests.
The next step is to examine the long run relationship between the variables of which results are shown in Table 6. In energy consumption demand function, we find that financial development adds in energy consumption. A 1 per cent increase in financial development is linked with 0.1272 per cent increase in energy demand. The impact of economic growth on energy consumption is positive as expected and statistically significant at 5 per cent level of significance. We find that 0.2003 per cent energy consumption is increased due to 1 per cent increase in economic growth. Energy consumption has positive effect on financial development and it is statistically significant at 10 per cent level. A 1 per cent increase in energy demand raises financial development by 0.30 per cent. Economic growth has positive and statistically significant impact on financial development. A 1 per cent increase in economic growth is positively linked with financial development by 1.6 per cent. In economic growth empirical model, we find that energy consumption stimulates economic growth and it is statistically significant at 1 per cent level of significance. We note that a 1 per cent increase in energy consumption boosts economic growth by 0.2620 per cent. The positive relationship exists from financial development to economic growth at 1 per cent significance level. A 0.4239 increase in financial development leads to economic growth by 1 per cent. We find that energy consumption and financial development are complementary but energy consumption has stronger impact on financial development and the same inference is for economic growth to financial development. Energy consumption and economic growth are interdependent but economic growth depends on energy consumption. The assumptions of classical linear regression model (CLRM) are fulfilled by energy consumption, financial development, and economic growth models successfully. There is no evidence of nonnormality of error terms or serial correlation and no evidence of ARCH or white heteroskedasticity. The functional form of all the models is well specified.
Long run results.
Variable
Dependent variable = lnEt
Dependent variable = lnMt
Dependent variable = lnYt
Coefficient
T-statistic
Coefficient
T-statistic
Coefficient
T-statistic
Constant
4.4502*
22.1637
−0.0714
−0.0876
−0.9305**
−2.2596
lnEt
⋯
⋯
0.2968***
1.8336
0.2620*
3.3233
lnMt
0.1272*
2.6974
⋯
⋯
0.4239*
16.4255
lnYt
0.2003**
2.2226
1.5958*
16.4255
⋯
⋯
R2
0.5111
0.8358
0.8440
Adj-R2
0.5035
0.8332
0.8416
F-statistic
67.4413*
32.8348*
34.9205*
Diagnostic test
F-statistic
Probability
F-statistic
Probability
F-statistic
Probability
χ2 NORMAL
0.2536
0.8808
0.5604
0.6066
4.0656
0.1309
χ2 SERIAL
1.0040
0.8870
0.9786
0.8900
0.6545
0.5507
χ2 ARCH
1.5812
0.1704
1.0796
0.4490
0.1962
0.9860
χ2 WHITE
1.3436
0.2576
0.9372
0.4405
0.5576
0.5060
χ2 REMSAY
0.1598
0.6899
0.8042
0.7659
2.6240
0.1077
Note: *, **, and *** denote the significance at 1%, 5%, and 10% levels, respectively. χ2 NORMAL is for normality test, χ2 SERIAL for LM serial correlation test, χ2 ARCH for autoregressive conditional heteroskedasticity, χ2 WHITE for white heteroskedasticity, and χ2 REMSAY for Remsay Reset test.
In the short run (Table 7), we find that financial development and economic growth have positive impact on energy consumption at 5 per cent significance level. The impact of energy consumption and economic growth is also positive and it is statistically significant at 5 per cent level of significance. Finally, energy consumption stimulates economic growth at 1 per cent level of significance. Financial development also adds in economic growth at 5 per cent level. The negative and statistically significant estimates for each of the ECMt-1, 0.6450, 0.0419, and −0.2693 (energy consumption, financial development, and economic growth), lend support to a long run relationship among the series. The short run deviations from the long run equilibrium are corrected by 66.50%, 4.19%, and 26.93% towards long run equilibrium path each month. The diagnostic tests show that error terms of short run models are normally distributed and are free of serial correlation, heteroskedasticity, and ARCH problems for all models. The Ramsey reset test suggests that functional form for the short run models is well specified.
Short run results.
Variable
Dependent variable = ΔlnEt
Dependent variable = ΔlnMt
Dependent variable = ΔlnYt
Coefficient
T-statistic
Coefficient
T-statistic
Coefficient
T-statistic
Constant
−0.0037
−0.5813
0.0073*
2.3969
0.0023
0.5188
ΔlnEt
⋯
⋯
0.0639**
2.4847
0.1311*
2.6318
ΔlnMt
0.4539**
2.5264
⋯
⋯
0.3246**
2.6012
ΔlnYt
0.2735**
2.3395
0.1442**
2.5082
⋯
⋯
ECMt-1
−0.6450*
−7.9533
−0.0419*
−2.9695
−0.2693*
−4.5419
R2
0.3756
0.0904
0.2002
Adj-R2
0.3608
0.0689
0.1813
F-statistic
25.4650*
4.2105*
10.5974*
Diagnostic test
F-statistic
Probability
F-statistic
Probability
F-statistic
Probability
χ2 NORMAL
0.9044
0.6362
0.3808
0.6776
0.1352
0.9212
χ2 SERIAL
0.2916
0.7416
0.2260
0.8500
0.8703
0.4212
χ2 ARCH
0.0291
0.8646
1.6659
0.1931
0.5830
0.6065
χ2 WHITE
0.7954
0.5752
0.2217
0.9691
0.4754
0.4012
χ2 REMSAY
2.2072
0.1142
0.5671
0.5681
1.6040
0.1175
Note: * and ** denote the significance at 1% and 5% levels, respectively. χ2 NORMAL is for normality test, χ2 SERIAL for LM serial correlation test, χ2 ARCH for autoregressive conditional heteroskedasticity, χ2 WHITE for white heteroskedasticity, and χ2REMSAY for Resay Reset test.
The cumulative sum (CUSUM) and the cumulative sum of squares (CUSUMsq) tests suggest stability of the long and short run parameters (Figures 1 and 2). The graphs of the CUSUM and CUSUMsq test lie within the 5 per cent critical bounds which confirm stability of parameters [95] of the model which is also well specified.
Plot of cumulative sum of recursive residuals. The straight lines represent critical bounds at 5% significance level.
Plot of cumulative sum of squares of recursive residuals. The straight lines represent critical bounds at 5% significance level.
5.1. The VECM Granger Causality
If cointegration is confirmed, there must be uni- or bidirectional causality among the series. We examine this relation within the VECM framework. Knowledge about causality can help craft appropriate energy and financial policies for sustainable economic growth. Table 8 reports results on the direction of long and short run causality. We find feedback relation between energy consumption and economic growth. This implies that economic growth depends upon energy consumption and rise in income per capita further increases energy demand. Therefore, adoption of energy conservation policies will have detrimental impact on economic growth. Our findings suggest the importance of encouraging energy exploring policies. In this regard, we praise the government of Lebanon in concluding its first offshore oil and gas rights auction which has drawn interest from about 100 companies. They have bought geophysical data in preparation for an upcoming bid round for Lebanese offshore energy production rights, which should occur soon (Lebanon Explores Offshore Energy, By April Yee, 2/25/2013, http://www.moneyshow.com/investing/article/1/TheNational-30654/Lebanon-Explores-Offshore-Energy/). Furthermore, given that Lebanon is very close to some of the major producers of LNG in the world (Qatar, Nigeria, and Egypt), Lebanon should explore this option to meet its energy needs via this source in the short term given the close geographical location of these sources and the low cost of shipment as well as the good relations with these countries.
VECM Granger causality analysis.
Dependent variable
Direction of causality
Short run
Long run
Joint long-and-short run causality
ΔlnEt-1
ΔlnFt-1
ΔlnYt-1
ECTt-1
ΔlnEt-1,ECTt-1
ΔlnFt-1,ECTt-1
ΔlnYt-1,ECTt-1
ΔlnEt
⋯
3.7677*[0.0068]
2.5274**[0.0444]
−0.8753*[-6.2656]
⋯
9.7481*[0.0000]
8.5766*[0.0000]
ΔlnFt
2.2240***[0.0707]
⋯
3.0709**[0.0191]
−0.0491**[-2.2649]
2.3876**[0.0422]
⋯
2.5597**[0.0301]
ΔlnYt
2.8445**[0.0256]
1.5706 [0.1868]
⋯
−0.2735*[-3.6323]
4.2855*[0.0013]
3.5162*[0.0054]
⋯
Note: *, **, and *** show significance at 1, 5, and 10 per cent levels, respectively.
Financial development and energy consumption Granger cause each other, that is, supporting bidirectional causal relationship. Financial development results in lowering the cost of borrowing to consumers and producers for big durable items and for setting up new businesses. This raises energy demand which leads to further economic expansions. These further attract consumers and raise the demand for financial services and hence financial development. The bidirectional causality between financial development and economic growth also could show the importance of directing monetary policy in Lebanon to enhance capitalization, especially in the energy sector which is highly capital intensive. This is important given our earlier observation of the importance of encouraging energy exploring policies.
In the short run, bidirectional relationship between financial development and energy consumption is found. The feedback effect exists between energy consumption and economic growth and economic growth Granger causes financial development.
5.2. Variance Decomposition Method (VDM)
We have used the generalized forecast error variance decomposition method (GFEVDM) using vector autoregressive (VAR) system to test the strength of causal relationship between energy consumption, financial development, and economic growth in case of Lebanon. This is due to the limitations associated with VECM Granger causality test which cannot capture the relative strength of causal relation between the variables beyond the selected time period. The GFEVDM indicates the magnitude of the predicted error variance for a series accounted for by innovations from each of the independent variables over different time-horizons beyond the selected time period. The main advantage of this approach is that it is insensitive with ordering of the variables because such ordering is uniquely determined by VAR system. Further, the GFEVDM estimates the simultaneous shock effects. Engle and Granger [96] and Ibrahim [97] argued that, with VAR framework, variance decomposition approach produces better results as compared to other traditional approaches.
The results of variance decomposition approach are reported in Table 9. The results indicate that a 67.16 per cent portion of energy consumption is explained by its own innovative shocks while innovative shocks of financial development and economic growth contribute to energy consumption by 21.44 per cent and 11.39 per cent, respectively. The innovative shocks stemming in energy consumption contribute to financial development by 23.37 per cent. The contribution of economic growth to financial development is 15.68 per cent and the rest is explained by innovative shocks on financial development. Economic growth is 23.35 per cent and 39.78 per cent explained by innovative shocks in energy consumption and financial development.
Variance decomposition method (VDM).
Horizon
Variance decomposition of lnEt
Variance decomposition of lnFt
Variance decomposition of lnYt
lnEt
lnFt
lnYt
lnEt
lnFt
lnYt
lnEt
lnFt
lnYt
1
100.0000
0.0000
0.0000
2.0421
97.9578
0.0000
5.1366
1.4692
93.394
5
83.1278
14.6550
2.2170
4.0701
95.7476
0.1821
11.9018
11.2447
76.8533
10
77.8984
14.5835
7.5180
14.1938
81.9566
3.8495
17.8004
23.7391
58.4604
15
74.4234
16.3723
9.2042
18.5622
72.7660
8.6716
20.3784
30.3791
49.2424
20
71.7926
18.1021
10.1052
20.9817
66.9868
12.0313
21.6981
34.7860
43.5157
25
69.3898
19.8089
10.8012
22.4392
63.3360
14.2246
22.6609
37.7032
39.6358
26
68.9305
20.1451
10.9243
22.6571
62.7784
14.5644
22.8160
38.1738
39.0101
27
68.4776
20.4773
11.0450
22.8580
62.2635
14.8784
22.9631
38.6143
38.4224
28
68.0320
20.8044
11.1634
23.0436
61.7872
15.1691
23.1020
39.0285
37.8693
29
67.5941
21.1265
11.2793
23.2154
61.3459
15.4386
23.2326
39.4193
37.3480
30
67.1634
21.4439
11.3926
23.3747
60.9362
15.6890
23.3553
39.7885
36.8561
Overall our results show that the feedback effect is found between financial development and energy consumption but it is stronger from energy consumption to financial development. The unidirectional causality is found running from energy consumption and financial development to economic growth, that is, energy-led growth and finance-led growth hypothesis.
The results of IRF are shown in Figure 3 which reveals that the response of energy consumption is positive but minimal after 10th and 6th time horizon due to one standard deviation shock stemming from financial development and economic growth. The response in financial development is positive and strong due to one standard deviation shock in energy consumption and economic growth. Energy consumption and financial development seem to contribute to economic growth. Overall our results are consistent with findings of variance decomposition approach.
Impulse response function.
6. Conclusion and Policy Implications
This paper explored the relationship between economic growth, financial development, and energy consumption in Lebanon. Using data for 2000M2–2010M12, we have applied unit root test accommodating single unknown structural break stemming in the series. The ARDL bounds testing is applied to find out cointegration among the variables in the presence of structural breaks. The directions of causal relationships between economic growth, financial development, and energy consumption were examined by applying the VECM Granger causality and robustness of causality results was tested by using innovative accounting approach.
Our results found cointegration between the series in the presence of structural breaks arising in the variables. We found that economic growth and financial development raise energy demand. Energy consumption boosts economic growth and financial development also contributes to economic growth by capitalization enhancing effect. Furthermore, energy consumption and economic growth increase the rate of financial development. The VECM Granger causality analysis revealed that bidirectional causality is found between energy consumption and economic growth. The feedback effect also exists between financial development and energy consumption. Economic growth and financial development are Granger causes of each other. The results by innovative accounting approach are different from the VECM Granger causality test. This may be due to difference in the methodological backgrounds of both techniques. The empirical analysis by innovation accounting approach shows the bidirectional causal relationship between financial development and energy consumption. Economic growth is Granger cause of energy consumption and financial development.
The study, therefore, recommends that, in the short run, policy makers should put more emphasis on developing strategies that would result in achieving higher mobilization of savings in order to boost Lebanese investors’ confidence and to also attract more foreign investment in Lebanon. Furthermore, desired financial policy to encounter the rising demand for energy by enhancing the process of capitalization of the energy sector is also very desirable. Our results further caution of the use of policy tools geared towards restricting energy consumption in short run that is a part of national energy policy, as these may result in lower economic growth. Such conservation policies should be taken gradually and carefully as to not negatively impact the growth of the economy. Furthermore, given that Lebanon is very close to some of the major producers of LNG in the world (Qatar, Nigeria, and Egypt), it should explore this option to meet its energy needs via this source in the short term given the close geographical location of these sources and the low cost of shipment as well as the good relations with these countries. However, in the long run, the Lebanese government should shift its focus towards achieving higher economic growth, in order to boost its financial development and to sustain a steady flow of needed energy. In this regard, policy makers should put emphasis on the development of domestic energy resources to protect the country from any undesirable external energy shock given its extensive dependence on energy imports.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
KraftJ.KraftA.On the relationship between energy and GNP19783401403LeeC.-C.ChangC.-P.ChenP.-F.Energy-income causality in OECD countries revisited: the key role of capital stock20083052359237310.1016/j.eneco.2008.01.0052-s2.0-48049120591AngJ. B.CO2 emissions, energy consumption, and output in France200735104772477810.1016/j.enpol.2007.03.0322-s2.0-34447542422ApergisN.PayneJ. E.Energy consumption and economic growth: evidence from the Commonwealth of Independent States200931564164710.1016/j.eneco.2009.01.0112-s2.0-67650718489ApergisN.PayneJ. E.CO2 emissions, energy usage, and output in Central America20093783282328610.1016/j.enpol.2009.03.0482-s2.0-67349121256ApergisN.PayneJ. E.The causal dynamics between coal consumption and growth: evidence from emerging market economies20108761972197710.1016/j.apenergy.2009.11.0352-s2.0-77950858018Wolde-RufaelY.Energy consumption and economic growth: the experience of African countries revisited200931221722410.1016/j.eneco.2008.11.0052-s2.0-58549109603BartleetM.GounderR.Energy consumption and economic growth in New Zealand: results of trivariate and multivariate models20103873508351710.1016/j.enpol.2010.02.0252-s2.0-77953686957ShanJ. Q. J.Does financial development lead economic growth? The case of China?20061231250ShahbazM.LeanH. H.FarooqA.Natural gas consumption and economic growth in Pakistan201318879410.1016/j.rser.2012.09.0292-s2.0-84868258519ShahbazM.ShamimS. M. A.AamirN.Macroeconomic environment and financial sector's performance: econometric evidence from three traditional approaches20101103123JumbeC. B. L.Cointegration and causality between electricity consumption and GDP: empirical evidence from Malawi2004261616810.1016/S0140-9883(03)00058-62-s2.0-0347930477AbosedraS.BaghestaniH.New evidence on the causal relationship between U.S. energy consumption and gross national product199114285292ChengB. S.LaiT. W.An investigation of co-integration and causality between energy consumption and economic activity in Taiwan199719443544410.1016/S0140-9883(97)01023-22-s2.0-0031251716SoytasU.SariR.Energy consumption and GDP: causality relationship in G-7 countries and emerging markets200325133372-s2.0-003721806610.1016/S0140-9883(02)00009-9OhW.LeeK.Causal relationship between energy consumption and GDP revisited: the case of Korea 1970–19992004261515910.1016/S0140-9883(03)00030-62-s2.0-0347300440FataiK.OxleyL.ScrimgeourF. G.Modelling the causal relationship between energy consumption and GDP in New Zealand, Australia, India, Indonesia, The Philippines and Thailand2004643-443144510.1016/S0378-4754(03)00109-5MR21093572-s2.0-0942266801LeeC.-C.Energy consumption and GDP in developing countries: a cointegrated panel analysis200527341542710.1016/j.eneco.2005.03.0032-s2.0-18744405664LeeC.-C.The causality relationship between energy consumption and GDP in G-11 countries revisited20063491086109310.1016/j.enpol.2005.04.0232-s2.0-32044448488LeeC.-C.ChangC.-P.Energy consumption and GDP revisited: a panel analysis of developed and developing countries20072961206122310.1016/j.eneco.2007.01.0012-s2.0-34548451054ChontanawatJ.HuntL. C.PierseR.Does energy consumption cause economic growth? Evidence from a systematic study of over 100 countries200830220922010.1016/j.jpolmod.2006.10.0032-s2.0-39149107077NarayanP. K.SmythR.Energy consumption and real GDP in G7 countries: new evidence from panel cointegration with structural breaks20083052331234110.1016/j.eneco.2007.10.0062-s2.0-48049083113BowdenN.PayneJ. E.The causal relationship between U.S. energy consumption and real output: a disaggregated analysis200931218018810.1016/j.jpolmod.2008.09.0012-s2.0-59149087288OdhiamboN. M.Energy consumption and economic growth nexus in Tanzania: an ARDL bounds testing approach200937261762210.1016/j.enpol.2008.09.0772-s2.0-58149195160TsaniS. Z.Energy consumption and economic growth: a causality analysis for Greece201032358259010.1016/j.eneco.2009.09.0072-s2.0-77950518844WangS. S.ZhouD. Q.ZhouP.WangQ. W.CO2 emissions, energy consumption and economic growth in China: a panel data analysis20113994870487510.1016/j.enpol.2011.06.0322-s2.0-79960728364YazdanG. F.HosseinS. S. M.Causality between oil consumption and economic growth in Iran: an ARDL testing approach20122678686LiseW.van MontfortK.Energy consumption and GDP in Turkey: is there a co-integration relationship?20072961166117810.1016/j.eneco.2006.08.0102-s2.0-34548539870ErdalG.ErdalH.EsengünK.The causality between energy consumption and economic growth in Turkey200836103838384210.1016/j.enpol.2008.07.0122-s2.0-51349128976HuangB.-N.HwangM. J.YangC. W.Causal relationship between energy consumption and GDP growth revisited: a dynamic panel data approach2008671415410.1016/j.ecolecon.2007.11.0062-s2.0-38349078543MallickH.Examining the linkage between energy consumption and economic growth in India200943249280Sa’adS.Energy consumption and economic growth: causality relationship for Nigeria201034152410.1111/j.1753-0237.2010.00173.xBinhP. T.Energy consumption and economic growth in Vietnam: threshold cointegration and causality analysis2011111172-s2.0-84868582404QaziA. Q.AhmedK.MudassarM.Disaggregate energy consumption and industrial output in Pakistan: an empirical analysis2012Wuhan University of TechnologySoileI. O.Energy-economy nexus in Indonesia: a bivariate cointegration201226205218GlasureY. U.Energy and national income in Korea: further evidence on the role of omitted variables200224435536510.1016/S0140-9883(02)00036-12-s2.0-0036639668BelkeA.DobnikF.DregerC.Energy consumption and economic growth: new insights into the cointegration relationship201133578278910.1016/j.eneco.2011.02.0052-s2.0-80052166799EggohJ. C.BangakeC.RaultC.Energy consumption and economic growth revisited in African countries201139117408742110.1016/j.enpol.2011.09.0072-s2.0-80054862936MarquesA. C.FuinhasJ. A.MansoJ. P.A quantile approach to identify factors promoting renewable energy in European countrie201149335136610.1007/s10640-010-9436-82-s2.0-79958144437KaplanM.OzturkI.KalyoncuH.Energy consumption and economic growth in Turkey: co-integration and causality analysis201123141ShahbazM.ZeshanM.AfzaT.Is energy consumption effective to spur economic growth in Pakistan? New evidence from bounds test to level relationships and Granger causality tests2012296231023192-s2.0-8486532740910.1016/j.econmod.2012.06.027FuinhasJ. A.MarquesA. C.Energy consumption and economic growth nexus in Portugal, Italy, Greece, Spain and Turkey: an ARDL bounds test approach (1965–2009)201234251151710.1016/j.eneco.2011.10.0032-s2.0-84857366789ZeshanM.Finding the cointegration and causal linkages between the electricity production and economic growth in Pakistan201331134435010.1016/j.econmod.2012.11.0212-s2.0-84871998667SoytasU.SariR.EwingB. T.Energy consumption, income, and carbon emissions in the United States2007623-448248910.1016/j.ecolecon.2006.07.0092-s2.0-34247608102GrossC.Explaining the (non-) causality between energy and economic growth in the U.S.—a multivariate sectoral analysis20123424894992-s2.0-8485736691910.1016/j.eneco.2011.12.002DagherL.YacoubianT.The causal relationship between energy consumption and economic growth in Lebanon20125079580110.1016/j.enpol.2012.08.0342-s2.0-84866950474JohansenS.Statistical analysis of cointegration vectors1988122-323125410.1016/0165-1889(88)90041-3MR986516TodaH. Y.YamamotoT.Statistical inference in vector autoregressions with possibly integrated processes1995661-222525010.1016/0304-4076(94)01616-8MR13230602-s2.0-0000383532SchumpeterJ. A.1932Cambridge, Mass, USAHarvard University PressGoldsmithR.1969New York, NY, USAYale University PressMcKinnonR.1973Washington, DC, USAThe Brookings InstitutionShawE.1973New York, NY, USAOxford University PressDemirguc-KuntA.LevineR.Stock markets, corporate finance, and economic growth: an overview199610222323910.1093/wber/10.2.2232-s2.0-0001033880LevineR.Financial development and economic growth: views and agenda19973526887262-s2.0-0000329790GüryayE.ŞafakliO.TüzelB.Financial development and economic growth: evidence from Northern Cyprus200785761Wolde-RufaelY.Re-examining the financial development and economic growth nexus in Kenya20092661140114610.1016/j.econmod.2009.05.0022-s2.0-69849105007ShahbazM.A reassessment of finance-growth nexus for Pakistan: Under the investigation of FMOLS and DOLS techniques200916580TownsendR. M.Optimal contracts and competitive markets with costly state verification197921226529310.1016/0022-0531(79)90031-0ZBL0446.900072-s2.0-0009424011GreenwoodJ.JovanovicB.Financial development, growth, and the distribution of income19909851076110710.1086/261720Abu-BaderS.Abu-QarnA. S.Financial development and economic growth: the Egyptian experience200830588789810.1016/j.jpolmod.2007.02.0012-s2.0-50149098789IbrahimM. H.The role of the financial sector in economic development: the Malaysian case200754446348310.1007/s12232-007-0023-4JalilA.MaY.Financial development and economic growth: time series evidence from Pakistan and China200829229682-s2.0-51349145192CoccoreseP.An investigation on the causal relationships between banking concentration and economic growth200817355757010.1016/j.irfa.2006.11.0022-s2.0-44349186759MasihM.Al-ElgA.MadaniH.Causality between financial development and economic growth: an application of vector error correction and variance decomposition methods to Saudi Arabia200941131691169910.1080/000368407013202332-s2.0-67650766710KarS.MandalK.Reexamining the finance—growth relationship for a developing economy: a time series analysis of post-reform India2012313PatrickH. T.Financial development and economic growth in underdeveloped countries1966141174189LiangQ.TengJ.-Z.Financial development and economic growth: evidence from China20061743954112-s2.0-3375095207110.1016/j.chieco.2005.09.003FungM. K.Financial development and economic growth: convergence or divergence?2009281566710.1016/j.jimonfin.2008.08.0012-s2.0-58349116212ChukwuJ. O.AguC. C.Multivariate causality between financial development and economic growth in Nigeria2009721AmarathungaH.Finance-growth nexus: evidence from Sri Lanka201040136OdhiamboN. M.Finance-investment-growth nexus in South Africa: an ARDL-bounds testing procedure20104332052192-s2.0-7795486440110.1007/s10644-010-9085-5OdhiamboN. M.Financial deepening, capital inflows and economic growth nexus in Tanzania: a multivariate model2011286571OdhiamboN. M.Financial deepening, capital inflows and economic growth nexus in Tanzania: a multivariate model20112816571HassanM. K.SanchezB.YuJ.-S.Financial development and economic growth: new evidence from panel data20115118810410.1016/j.qref.2010.09.0012-s2.0-78650772406IlhanO.Finance-growth nexus: empirical evidence from South Africa20074317ZhengC.HanX.HeL.Financial development and economic growth: evidence from cointegration and granger causality testsProceedings of the 2nd IEEE International Conference on Advanced Management Science (ICAMS ’10)July 2010Chengdu, China23724110.1109/ICAMS.2010.55532472-s2.0-77957269321Al-MalkawiH.-A. N.MarashdehH. A.AbdullahN.Financial development and economic growth in the UAE: empirical assessment using ARDL approach to co-integration20124510511510.5539/ijef.v4n5p105EslamloueyanK.SakhaeiE. A.The short run and long run causality between financial development and economic growth in the Middle East2011166176Al-MalkawiH.-A. N.MarashdehH. A.AbdullahN.Financial development and economic growth in the UAE: empirical assessment using ARDL approach to co-integration20124510511510.5539/ijef.v4n5p105BakhoucheA.Does the financial sector promote economic growth? A case of Algeria200731123442-s2.0-35248853788GantmanE. R.DabósM. P.A fragile link? A new empirical analysis of the relationship between financial development and economic growth201240451753210.1080/13600818.2012.7285822-s2.0-84869461152LoveI.ZicchinoL.Financial development and dynamic investment behavior: evidence from panel VAR200646219021010.1016/j.qref.2005.11.0072-s2.0-33646892264SadorskyP.The impact of financial development on energy consumption in emerging economies20103852528253510.1016/j.enpol.2009.12.0482-s2.0-77649183123MankiwN. G.ScarthW.2008New York, NY, USAWorthKaranfilF.How many times again will we examine the energy-income nexus using a limited range of traditional econometric tools?20093741191119410.1016/j.enpol.2008.11.0292-s2.0-63449104491YandanLijunZ.Financial development and energy consumption: an empirical research based on Guangdong ProvinceProceedings of the International Conference on Information Management, Innovation Management and Industrial Engineering (ICIII ’09)December 200910210510.1109/ICIII.2009.3342-s2.0-77649302372SadorskyP.Financial development and energy consumption in Central and Eastern European frontier economies201139299910062-s2.0-7865041686110.1016/j.enpol.2010.11.034ShahbazM.LeanH. H.Does financial development increase energy consumption? The role of industrialization and urbanization in Tunisia201240147347910.1016/j.enpol.2011.10.0502-s2.0-84855749375TangC. F.TanB. W.The linkages among energy consumption, economic growth, relative price, foreign direct investment, and financial development in Malaysia201448278179710.1007/s11135-012-9802-42-s2.0-84895909937IslamF.ShahbazM.AhmedA. U.AlamM. M.Financial development and energy consumption nexus in Malaysia: a multivariate time series analysis201330143544110.1016/j.econmod.2012.09.0332-s2.0-84869209135LanquarR.Tourism in the Mediterranean: scenarios up to, 203020111CASE—Center for Social and Economic ResearchAduG.MarbuahG.MensahJ. T.Financial development and economic growth in Ghana: does the measure of financial development matter?20133419220310.1016/j.rdf.2013.11.0012-s2.0-84891840326PesaranM. H.ShinY.SmithR. J.Bounds testing approaches to the analysis of level relationships200116328932610.1002/jae.6162-s2.0-0035588799ShahbazM.Income inequality-economic growth and non-linearity: a case of Pakistan20103786136362-s2.0-7795463191010.1108/03068291011060652BrownR. L.DurbinJ.EwansJ. M.Techniques for testing the constance of regression relations overtime197537149172EngleR. F.GrangerC. W.Co-integration and error correction: representation, estimation, and testing198755225127610.2307/1913236MR882095IbrahimM.Sectoral effects of monetary policy: evidence from Malaysia2005191831022-s2.0-1714438401510.1111/j.1467-8381.2005.00205.x