Intercourse across economic growth, trade and renewable energy consumption in developing and developed countries

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.com/locate/rser

Intercourse across economic growth, trade and renewable energy consumption in developing and developed countries

crossmark

Fethi Amri University of Chartage, Faculty of Economic Sciences and Management of Nabeul, Street Hammamet, Mrezgua, Nabeul 8000, Tunisia

A R T I C L E I N F O

A BS T RAC T

Keywords: Economic growth Trade Renewable energy consumption Developed countries Developing countries

This article exploits a dynamic simultaneous-equation panel data approach to research the bond across economic growth, renewable sort of energy, and trade for information of 72 countries since 1990 until 2012. We likewise analyze this relationship by separating all countries into three groups as per the level of development: whole, developing, high-income developing, upper middle-income developing, lower middle-income developing, lower-income developing, developed, major developed, and others developed countries. The outcomes demonstrate a feedback linkage between income and renewable energy consumption, between trade and renewable energy consumption and between trade and income. This implies that the three variables considered are interdependent.

1. Introduction Current studies inspecting the connection through economic growth, trade, and consumption of renewable energy are composed of three divisions. The first division of researches investigated in the renewable energy consumption- economic growth nexus. The findings of this literature demonstrate the absence of consensus about the sense of causality among them. The results provide four types of assumption [1,2]. At first, the neutrality assumption sustained the absence of causality betwixt renewable energy and economic growth. It involves the absence of any link among them. For example, Payne [3] used yearly data United States during the phase 1949–2006 and deduced that the neutrality hypothesis is accepted. Analogously, Menegaki [4] examined the bond betwixt renewable energy consumption and gross domestic product (GDP) and in Europe in 11 years period (1997–2007). They rejected the possible association across them. In the same way, Omri et al. [1] implanted panel data models to explore the association among renewable energy, nuclear energy and gross domestic product in the case of 17 developing and developed countries. Their results confirmed the neutrality hypothesis in Switzerland, Brazil and Finland, between 1990 and 2011. Also, according to Yildirim et al. [5] there is no relationship between different categories of renewable energy and GDP in USA in the course of epoch moving from 1949 to 2010. Then, the feedback assumption sustained bidirectional links amongst renewable energy and output. It implicates that gross domestic product and renewable energy consumption are interrelated. In the Latin American countries study, Al-mulali et al. [6] used a multivariate panel data

model by integrating trade, non-renewable energy, labor and capital in Cob-Douglas production function. The outcomes indicate bi-directional causality among output and renewable energy. It implicates that policy makers should promote the renewable energy sources. Apergis and Payne [7] focused in Eurasian countries to study the link amongst renewable energy and output between 1990 and 2007. The results denote feedback linking between the two variables by building panel data models with additional variables (capital and labor). Apergis and Payne [8] used statistics of 80 countries to test the linkage among renewable energy and economic growth. The findings provide feedback relationship between them into panel data growth model. It illustrates that those countries should enforce the renewable energy promotion. Lin and Moubarak [9] interrogate the link amongst renewable energy and GDP by incorporating labor and CO2 issuances in the multivariate model. They found that the feedback assumption is found in China study from 1977 to 2011. It implicates that the improvement of renewable energy is needful. Pao and Fu [10] substantiate the feedback causality between renewable energy and GDP in Brazil during the term 1980–2010. Shahbaz et al. [11] investigated the linking among renewable energy and output by means of the multivariate framework together with capital and labor forces as additional variables. They advanced bidirectional association among renewable energy and output in the example of Pakistani data from 1972 to 2011. Ohler and Fetters [12] found feedback linking between different forms of renewable energy and gross domestic product in OECD economies between 1990 and 2008. Also, the conservation assumption sustained unidirectional assumption moving from gross domestic product to renewable energy.

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.rser.2016.11.230 Received 12 June 2015; Received in revised form 6 October 2016; Accepted 18 November 2016 1364-0321/ © 2016 Elsevier Ltd. All rights reserved.

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

and Stiakakis [36] find a bidirectional connection between exports and output in the short and the long run. Fourth, trade and economic growth are not related [21,24,32,37,38]. By using data for 21 African countries over the period 1965–2008, Menyah and al. [37] find the absence of connection among trade and GDP. In 23 Asian countries study between 1950 and 2010, Trejos and Barboza [38] find that there is no evidence association between trade and economic growth. Fifth, trade negatively affects economic growth [24,34,39]. By using data for Japan and Korea for 1970–1997, Jin [24] found that trade has negative impacts on Korean economic growth and insignificant effects in Japan. In developing countries study Grennway et al. [39], prove the presence of negative connection among trade and GDP. In 82 countries study over the period 1960–2000, Chang et al. [34] find that effects of trade on economic growth is conditioned by the complementary reforms undertaken. These effects will be positive or negative. The third strand of researches analyzes the relationship between trade and renewable energy consumption. Ben Aïssa et al. [40] look into the link amongst GDP, trade, and renewable energy consumption by employing data of 11 African countries between 1980 and 2008. Within a long run analysis, the results reveal the presence of bidirectional links between GDP and trade variables and unidirectional relationship running from renewable energy and trade to GDP. In the short term analysis, the findings confirm the bidirectional relationship among trade and GDP and reject the possible association across GDP and renewable energy and between trade and renewable energy. Sebri and Ben Salha [41] examined the links amongst trade, CO2 emissions, renewable energy and the gross domestic product in the example of BRICS countries between 1971 and 2010. The results accept the bidirectional hypothesis between renewable energy and output between trade and output and between trade and renewable energy. They illustrate that the importance of trade and economic growth to mature the renewable energy. Ben Jebli and Ben Youssef [42] examine the linking between renewable energy, non-renewable energy, trade and economic growth in Tunisian study in the period going from 1980 to 2010. The results reveal the appearance of unidirectional nexus extending from the gross domestic product and trade to renewable energy. They propose that Tunisia should stimulate the employment of renewable energy, make a method for augmenting its advantage from renewable energy innovation exchange happening when importing capital products and achieve the establishment of renewable energy ventures oriented to export. To the best of our knowledge, there is no article, attempting to comprehend the linkage amongst renewable energy, trade and GDP in developing and developed countries. In the current study, we scan the nexus amongst GDP, renewable form of energy consumption, and openness in developed and developing countries. We employ a multivariate framework to study the linkage across renewable energy, trade and GDP for 72 countries during the epoch 1990–2012. Especially, we look into three following causality: (1) among GDP and renewable energy; (2) among trade and renewable category of energy; and (3) among GDP and trade. Our article is divided into subsequent parts. Section 2 combines data and econometric method. Section 3 merges results and discussions. Section 4 amalgamates final conclusions and some policy implications.

It demonstrates that economic growth helps to develop the renewable energy consumption. Tugcu et al. [13] examined the linking among GDP and renewable and non-renewable sorts of energy by means multivariate model. They concluded that conservation hypothesis is confirmed in Germany over the duration from 1980 to 2009. Ocal and Aslan [14] stipulated unidirectional causality going from GDP to renewable energy in Turkish study during the epoch 1990– 2010. Salim et al. [15] presented unidirectional relationship amongst GDP and renewable energy consumption in OECD economies from 1980 to 2011. At last, the growth assumption sustained unidirectional assumption going from renewable energy to GDP. It implicates that renewable category of energy consumption is a fundamental factor of output. Fang [16] studied the linkage among renewable energy and gross domestic product in the example of China from 1980 to 2010. The estimation of a multivariate model with labor and capital factors proves the positive impact of renewable energy on GDP. Pirlogea and Cicea [17] show unidirectional link from GDP to renewable energy in Romania during the epoch 1990–2010. Inglesi-Lotz [18] audited the relationships amongst renewable energy consumption and gross domestic product from 34 OECD economies during the interval 1990– 2010. The attainments prove that there is a positive association between renewable energy and GDP. It implies that OECD countries should focus on renewable energy improvement strategy to ameliorate their output. The second division of researches centered the attention on the relationship between trade and economic growth. The theoretical literature shows varied relationships between trade and economic growth. Firstly, there is evidence that trade stimulates economic growth through the transmission of technical Knowledge, the transfer of technology, the foreign direct investment and through the increase of economies of scale [19–21]. Secondly, there is evidence that economic growth boosts trade through the creation of a competitive advantage [22]. Thirdly, there is an absence of the relationship between trade and economic growth [23]. Fourthly, there is an existence of negative incidence of trade and economic growth [24]. On the empirical side, the literature shows mixed results in this area [22]. Four results are identified. First, trade positively affects economic growth [21,25–31]. In 12 African economies study, Onafowora and Owoye [27] find that there are significant effects of trade on economic growth. Using a data of 69 countries over the period 1986–1991, Greenaway et al. [28] find that higher trade will lead to more economic growth. Yanikkaya [29] confirms the hypothesis that growth increases if the trade increases by technology transfers, by scale economies and by the comparative advantage. In Malaysian study from 1970 to 2003, Chandran and Munusamy [30] suggest that trade contributes to enhance economic growth in the long time. In 75 countries study upon the epoch 1960–2003, Falvey et al. [31] suggest that trade enhances output in the long term. Eris and Ulasan [32] find that there is no evidence that trade affects positively economic growth. In Slovakia study over the period 2001–2010, Szkorupová [33] found that there is a positive action of trade (as measured by export) on economic growth. Second, economic growth positively affects trade [21,34,35]. In Indian study from 1981 to 2010, Sahoo and al [35] prove the presence of causality moving from GDP to trade. Third, trade and GDP are interconnected [21,26,36]. Harrison [26] found a positive connection between growth and different measures of openness in developing countries. The outcomes depend upon the types of specifications (cross-section or panel data) and on the time period. Also, the results demonstrate that there is feedback link among output and trade. By using cointegration, Granger causality tests, and the innovative accounting approach, Shahbaz [21] investigates the influence of trade openness on income in the long run. Shahbaz [21] find that trade boosts output. The outcomes demonstrate the emergence of bidirectional nexus across trade and gross domestic product. By employing yearly data in Croatia between 1994 and 2012, Dritsaki

2. Data and methodology 2.1. Methodology The purpose of this specific research is to provide the links between renewable energy consumption (RW), trade (TD) and economic growth (GW) for the case of developing and developed countries. To reach this objective, we start by considering the Cobb-Douglas production structure [40–44]: 528

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

GWit = e μi + ξit Ait Kit αK L it αL RWitαRW TDitαTD

Rwanda, Madagascar, United Republic of Tanzania, Uganda, Botswana, Mozambique, Gambia, Bangladesh, and Sri Lanka. The developed countries are Belgium, Austria, Finland, Denmark, Greece, France, Italy, Ireland, Portugal, Netherlands, Sweden, Spain, Bulgaria, United Kingdom, Romania, Norway, Poland, Australia, Switzerland, Japan, Canada, United States and New Zealand. This group is partitioned into two distinct categories by selecting the countries which are adopted at least one sort of renewable energy target and which are classified by the World Bank as high-income countries. The major developed countries including France, Italy, United Kingdom, Japan, Canada, and the United States. The others developed countries including Belgium, Austria, Finland, Denmark, Greece, Ireland, Portugal, Netherlands, Sweden, Spain, Norway, Poland, Australia, Switzerland, and New Zealand. We collect five variables: The first variable is the total renewable electricity Consumption in Billion Kilowat-thours (RW). This variable is collected from the U.S. Energy Information Administration. The second variable is the gross domestic product in constant 2005 US dollars (GW). The third variable is the trade (TD) considered as a total of exports and imports in constant 2005 US dollars. The fourth variable is the capital stock considered as the gross fixed capital formation in constant 2005 US dollars. The fifth variable is population (P). The statistics of GW, TD, K and P statistics are extracted from World Bank Indicators. Some statistics of various interested variables in whole, developed and developing countries are introduced in Table 1. The developed countries recorded the maximum level of gross domestic product per capita (67804.55), capital stock (3.11e+12) and trade (4.12e+12). In contrast, the developing countries reported the most level of population (1.35e+09) and renewable energy consumption (1003.515). Within the developed countries, the major ones, mentioned the largest level of renewable energy (527.4897), capital stock (3.11e+12), trade (4.12e+12), and population (3.14e+08). In contrast, the others developed countries registered the highest level of gross domestic product per capita (67804.55). Within the developing countries group, the high-income countries announced the greatest level of gross domestic product per capita (36110.13). The upper middle income countries recorded the largest level of renewable energy (1003.515), capital stock (2.08e+12), trade (3.40e+12), and population (1.35e+09). The developed countries marked the best performance average value of gross domestic product per capita (31020.96), capital stock (2.61e+11), trade (4.86e+11) and renewable energy consumption (62.43886). Contrariwise, the developing countries registered the highest achievement average value of population (8.20e+07). Within the developed countries, the major developed countries have the top result in all variables. Within the developing countries, the highincome group marked the first average value of Gross domestic product per capita (16036.8) and trade (2.00e+11), the upper middle income marked the largest average value of capital stock (7.63e+10), renewable energy (50.05511), and the lower middle income marked the top average value of population(1.26e+08). Also, the developing countries are more changeful in all variables. They have the top coefficient of variation value of Gross domestic product per capita (1.484), capital stock (3.364), renewable energy consumption (3.286), trade (2.538), and population (2.792). Within the developed countries, the others developed countries have the highest coefficient of variation of gross domestic product per capita (0.419), and trade (0.758). The major developed countries reported the greatest coefficient of variation of capital stock (0.991), population (0.892), and renewable energy consumption (1.135). Within the developing countries, the upper middle income announced the greatest coefficient of variation of capital stock (2.909), trade (2.411), renewable energy (2.391), and population (2.740). The lower-income group recorded the highest variability of gross domestic product per capita (1.591).

(1)

where, i (i :1,…,72), t (t=1990,…,2012), ξ and μ represent the country, the time period, the error component and the fixed specific components, respectively. GW, A, K, L, RW, and TD represent the gross domestic product, the total factor productivity, the capital, the labor, the renewable energy consumption and the trade respectively. αK ,αL , αRW and αTD represents the coefficients of capital, of labor, of renewable energy consumption and of trade, respectively. Dividing the both side of the production function by labor and taking them in logarithmic form as follows:

gwit = α0 + α1 kit + α2 rwit + α3 tdit + μi + ξit

(2)

gw, k, rw and td represent the log of gross domestic product per capita, the log of capital stock per capita, the log of renewable energy consumption per capita and the log of trade per capita, respectively. The interactions among GDP, renewable energy consumption, and trade can be expressed by as follows.

gwit = α0 + α1 kit + α2 rwit + α3 tdit + μi + ξit

(4)

rwit = β0 + β1 kit + β2 gwit + β3 tdit + μi + ξit

(5)

tdit = γ0 + γ1 kit + γ2 gwit + γ3 rwit + μi + ξit

(6)

Eq. (4) illustrates that the renewable energy consumption, trade and capital stock have an effect on economic growth [6– 10,14,15,18,45]. Eq. (5) formulates the importance of capital stock, growth and trade on renewable energy consumption [1,6–10,45]. Eq. (6), enunciate that the renewable energy consumption, growth and capital have an influence on trade [6,41,44]. The dynamic form of Eqs. (4)–(6) can be formulated as follows:

gwit = α0 + α1 gwit −1 + α2 kit + α3 rwit + α4 tdit + μi + ξit

(7)

rwit = β0 + β1 rwit −1 + β1 kit + β2 gwit + β3 tdit + μi + ξit

(8)

tdit = γ0 + γ1 tdit −1 + γ2 kit + γ3 gwit + γ4 rwit + μi + ξit

(9)

In order to eliminate the bias related to the correlation between the lagged of the dependent variable and error term, we use the method of moment's generalized (GMM) estimator [47]. This method makes the transformation of the variables to exclude the individual effects. 2.2. Data We exploit yearly information about 23 developed countries and 49 Developing countries over a 23-year period (1990–2012). The Developing countries are Mauritania, Egypt, Cameroon, Morocco, Kenya, Gabon, Rwanda, Madagascar, United Republic of Tanzania, Uganda, Lesotho, Botswana, Mozambique, Mauritius, South Africa, Namibia, Nigeria, Gambia, Brunei Darussalam, Senegal, Hong Kong, China, Malaysia, Indonesia, Republic of Korea, Philippines, Thailand, Singapore, India, Bangladesh, Sri Lanka, Pakistan, Turkey, Jordan, Dominican Republic, Guatemala, Mexico, El Salvador, Honduras, Bolivia, Brazil, Panama, Colombia, Chile, Costa Rica, Peru, Ecuador, Venezuela and Uruguay. These nations are partitioned into four distinct groups by selecting the countries which are adopted at least one category of renewable energy target. The high-income group including 6 countries: Brunei Darussalam, Hong Kong, Republic of Korea, Singapore, Chile, and Uruguay. The upper middle-income group including 18 countries: Gabon, Mauritius, South Africa, Namibia, China, Malaysia, Thailand, Turkey, Jordan, Dominican Republic, Mexico, Brazil, Panama, Colombia, Costa Rica, Peru, Ecuador, and Venezuela. The lower middle-income group including 14 countries: Mauritania, Egypt, Morocco, Lesotho, Nigeria, Senegal, Indonesia, Philippines, India, Pakistan, Guatemala, El Salvador, Honduras, Bolivia. The lower -income group including 10 countries: Kenya, 529

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

Table 1 (continued)

Table 1 Descriptive statistics of variables. M Whole countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Developed countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Major developed countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Others developed countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Developing countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population High-income developing countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Upper middle-income developing countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population Lower middle-income developing countries Gross domestic product per capita Capital stock Trade Renewable energy

SD

COV

MI

M

SD

COV

MI

MA

1.26e+08

2.68e+08

2.126

1597534

1.24e+09

889.3718

1415.147

1.591

143.0367

6697.425

3.06e+09 7.19e+09 2.112624

4.14e+09 7.78e+09 3.27931

1.352 1.082 1.552

4.18e+07 2.82e+08 0

2.66e+10 4.85e+10 16.781

2.93e+07

3.69e+07

1.259

916811

1.55e+08

MA consumption Population

12777.65

15752.46

1.232

143.0367

67804.55

1.13e+11 2.26e+11 35.84571

3.32e+11 4.47e+11 87.53677

2.938 1.977 2.442

1.31e+07 2.82e+08 0

3.11e+12 4.12e+12 1003.515

6.75e+07

1.93e+08

2.859

256929

1.35e+09

31020.96

14295.39

0.460

2460.654

67804.55

2.61e+11 4.86e+11 62.43886

5.17e+11 6.16e+11 101.9949

1.980 1.267 1.663

1.20e+09 1.34e+10 0.69

3.11e+12 4.12e+12 527.4897

3.67e+07

6.00e+07

1.634

3329800

3.14e+08

34156.7

4609.718

0.134

26476.1

45417.4

7.98e+11 1.20e+12 164.6818

7.91e+11 8.14e+11 151.8869

0.991 0.678 0.922

1.37e+11 3.49e+11 5.321

3.11e+12 4.12e+12 527.4897

1.03e+08

8.63e+07

0.873

2.78e+07

3.14e+08

33396.92

14011.69

0.419

4411.412

67804.55

8.00e+10 2.61e+11 28.62117

6.58e+10 1.98e+11 32.5014

0.822 0.758 1.135

1.01e+10 2.96e+10 0.69

3.88e+11 1.05e+12 142.412

1.30e+07

1.16e+07

0.892

3329800

4.68e+07

4214.458

6256.002

1.484

143.0367

36110.13

4.31e+10 1.04e+11 23.36322

1.45e+11 2.64e+11 76.78759

3.364 2.538 3.286

1.31e+07 2.82e+08 0

2.07e+12 3.40e+12 1003.515

8.20e+07

2.29e+08

2.792

256929

1.35e+09

16036.8

9636.208

0.600

3959.596

36110.13

6.08e+10 2.00e+11 6.796257

9.72e+10 2.72e+11 8.338162

1.598 1.360 1.226

8.46e+08 4.38e+09 0

3.19e+11 1.12e+12 30.276

1.40e+07

1.73e+07

1.253

256929

5.00e+07

4220.949

1764.917

0.418

462.7287

8532.348

7.63e+10 1.53e+11 50.05511

2.22e+11 3.69e+11 119.7108

2.909 2.411 2.391

4.43e+08 2.97e+09 0

2.08e+12 3.40e+12 1003.515

1.04e+08

2.85e+08

2.740

946703

1.35e+09

1164.645

615.5968

0.528

399.3269

3036.451

2.60e+10 5.28e+10 13.23215

6.36e+10 9.70e+10 26.48677

2.446 1.837 2.001

Lower-income developing countries Gross domestic product per capita Capital stock Trade Renewable energy consumption Population

M denotes mean. SD denotes the standard deviation. COV denotes the coefficient of variation. MI denotes the minimum value. MA denotes the maximum value.

3. Results and discussions Firstly, the stationary of the entirety variables is audited before applying any estimation method. For this reason, we have utilized Im et al. [48] test to check the stationarity of panel variables in level and logarithmic structure. The results of panel data stationarity tests in whole, developed, and developing countries of each variable with individual intercept and trend are reported in Table 2. The outcomes demonstrate that the assumption of the unit root of the four variables (gross domestic product per capita, trade per capita, renewable energy consumption per capita, capital stock per capita) is accepted at level and rejected at logarithmic structure. It implies that the variables are stationary at a logarithmic structure. Secondly, we expose the outcomes for the links between income, trade, and renewable energy consumption. Those results are obtained through the two-step GMM estimator propounded by Arellano and Bond [47] in order to avoid the possible correlation among the error term and independent variables in an econometric model. Also, we present the Sargan and AR2 tests to inspect the reliability of instruments and to check the absence second of order autocorrelation when the model is transformed in first difference. The estimated outcomes of Eq. (7) in whole, developed, and developing countries are presented in Table 3. The results provide uni-directional linking going from renewable energy, trade and capital stock to economic growth. Firstly, we notice that renewable energy consumption exerts a positive effect on GDP at 1% in whole, developed and developing countries. This benefit is more remarkable in developed countries (0.144) than in the developing ones (0.104). Regarding the panel of developed countries, this advantage is more noteworthy in major developed countries group (0. 152) than in others developed countries group (0.134). For developing countries, the renewable energy consumption terms are varied within a range of 0.113% in upper-income countries to 0.064% in lower-income countries. This outcome is conforming to the discoveries of Sebri and Ben-Salha [45], Salim et al. [15], and Inglesi-Lotz [18]. Secondly, trade has a positive and significant impact on economic growth at 1% in the three groups of countries. A 1% increase in trade helps to develop economic growth for whole, developed, and developing countries by 0.167%, 0.194%, and 0.144% respectively. Within the developed countries group, this coefficient is more important in major developed countries group (0.175) compared with others developed one (0.157). Concerning the developing countries, the high-income group has the most important trade impact comparing to others developing groups. It implicates that developed countries monopolize higher income than developing countries by trade. The result is appropriate in with the exposures of Kyophilavong et al. [49]. Thirdly, capital stock contributes positively to grow GDP at 1%

1.31e+07 4.68e+11 1.40e+09 7.65e+11 0 172.973 (continued on next page)

530

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

Table 2 Results for stationarity tests.

GW gw TR tr RW rw K k

Whole

Developed countries

Major developed countries

Others developed countries

Developing countries

High-income developing countries

Upper middleincome developing countries

Lower Middleincome developing countries

Lower-income developing countries

2.510 (0.994) −10.860* (0.000) 1.171 (0.879) −15.552* (0.000) 0.190 (0.575) −14.610* (0.000) 2.103 (0.982) −10.940* (0.000)

3.703 (0.999)

2.684 (0.996)

3.075 (0.998)

0.506 (0.693)

−1.433** (0.075)

2.829 (0.997)

−1.791** (0.036)

1.013 (0.844)

−5.029 (0.000) 0.865 (0.806)

−3.411 (0.000)

−3.930 (0.000)

−9.719 (0.000)

−4.199 (0.000)

−5.066 (0.000)

−5.669 (0.000)

−4.607* (0.000)

2.215 (0.986)

0.023 (0.505)

0.827 (0.795)

0.511 (0.695)

0.744 (0.771)

1.717 (0.957)

2.343 (0.990)

−4.060* (0.000)

−6.910* (0.000)

−5.234* (0.000)

−6.674* (0.000)

−6.715* (0.000)

−6.680* (0.000)

2.240 (0.987)

1.290 (0.901)

−13.174* (0.000) −0.781 (0.217)

−0.421 (0.336)

−0.464 (0.321)

0.143 (0.557)

−0.563 (0.286)

*

−4.894 (0.000)

−6.553 (0.000)

−3.889 (0.000)

−8.845 (0.000)

−6.771 (0.000)

−2.871* (0.002)

−0.220 (0.412)

−1.086 (0.190)

−12.114 (0.000) 3.150 (0.999)

1.076 (0.859)

2.913 (0.998)

0.990 (0.839)

7.075 (0.999)

−2.892* (0.001)

−4.313* (0.000)

−9.619* (0.000)

−2.393* (0.008)

−6.334* (0.000)

−5.238* (0.000)

−2.135** (0.016)

*

−8.288* (0.000) 1.436 (0.924) −8.173 (0.000) −8.876 (0.190) *

−5.320* (0.000)

*

*

*

*

*

*

*

*

*

*

*

P-values are in (). *** indicates significance level with 10%. * indicates significance level with 1% ** indicates significance level with 5%

The gross domestic parameters are changed at an interval of 0.896% in major developing countries to 0.463% in lower-income countries. This outcome is fitting with the aftereffects of Ben Jebli and Ben Youssef [46]. Next, trade provides a positive and significant effect on renewable energy consumption at 1%. A 1% increase in trade consolidates renewable energy for a whole, developed and developing countries by 0.115%, 0.188%, and 0.144% respectively. Within the developing countries, a 1% increase in trade enhances renewable energy for a high-income, upper middle-income, lower middle-income and lowerincome by 0.152%, 0.137%, 0.129%, and 0.109% respectively. By considering the developed countries, this impact is more considerable in the most developed group (0.192) compared with others developed one (0.172). This finding is fitting in with the disclosures of Sebri and Ben-Salha [45]. At last, the capital stock affects positively the renewable energy at

significance level. A 1% increase in capital boosts more highly the renewable energy use in the developed countries (0.269) compared to developing countries (0.104) [15]. For the developing countries category, the lowest coefficient is registered in the case of lower-income developing countries (0.019). Concerning the panel of developed countries, the lowest impact is registered in the case of others developed countries (0.257). The estimated outcomes of Eq. (8) are given in Table 4. The results concern the whole, developed and developing countries. The results reveal uni-directional relationship moving from economic growth, trade and capital stock to renewable energy consumption. At first, the gross domestic product participates positively on the amelioration of renewable energy at 1% significance level in whole, developed and developing countries. This advantage is more wonderful in developed countries (0.873) compared with the developing ones (0.678).

Table 3 Results of Eq. (7).

gw(−1) rw tr k cte Sargan test (pvalue) AR2 test (pvalue)

Whole

Developed countries

Major developed countries

Others developed countries

Developing countries

High-income developing countries

Upper middleincome developing countries

Lower Middleincome developing countries

Lower-income developing countries

0.589* (0.000) 0.110* (0.000) 0.167* (0.000) 0.245* (0.000) 2.000* (0.000) 66.927 (0.223)

0.350* (0.000)

0.488* (0.001)

0.263* (0.007)

0.681* (0.000)

0.705* (0.004)

0.602* (0.007)

0.514** (0.037)

0.214* (0.005)

0.144* (0.000)

0.152* (0.007)

0.134* (0.003)

0.104* (0.000)

0.088* (0.000)

0.113* (0.007)

0.073* (0.001)

0.064* (0.000)

0.194* (0.000)

0.175* (0.003)

0.157* (0.000)

0.144* (0.000)

0.151* (0.000)

0.143* (0.037)

0.083* (0.009)

0.023* (0.002)

0.269* (0.000)

0.305* (0.000)

0.257* (0.003)

0.104* (0.000)

0.096* (0.008)

0.125* (0.007)

0.079* (0.004)

0.019* (0.000)

3.740* (0.000)

3.612* (0.000)

2.916* (0.000)

1.313* (0.000)

1.907* (0.000)

1.155* (0.000)

0.692* (0.000)

0.192* (0.000)

21.277 (1.000)

43.150 (0.338)

46.295 (0.228)

43.779 (0.930)

38.124 (0.930)

14.427 (0.999)

23.899 (0.979)

36.399 (0.633)

0.499 (0.617)

0.959 (0.390)

0.072 (0.942)

−1.524 (0.127)

0.881 (0.378)

0.997 (0.318)

−0.949 (0.342)

1.439 (0.150)

−1.465 (0.142)

P-values are in (). *** indicates significance level with 10%. Sargan test reflecting the Overidentifying restrictions test. AR (2) reflecting the Arellano–Bond test of the second-order autocorrelation in first differences. * indicates significance level with 1% ** indicates significance level with 5%

531

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

Table 4 Results of Eq. (8).

Rw(−1) gw tr k cte Sargan test (pvalue) AR2 test (pvalue)

Whole

Developed countries

Major developed countries

Others developed countries

Developing countries

High-income developing countries

Upper middleincome developing countries

Lower Middleincome developing countries

Lower-income developing countries

0.381* (0.000) 0.796* (0.000) 0.115* (0.195) 0.235* (0.000) −16.019* (0.000) 60.557 (0.419)

0.792* (0.000)

0.736** (0.0 37) 0.703* (0.003)

0.291* (0.000)

0.307*** (0.087)

0.237** (0.012)

0.222*** (0.066)

0.198* (0.005)

0.873* (0.000)

0.812*** (0.070) 0.896* (0.001)

0.678* (0.000)

0.693* (0.008)

0.653* (0.000)

0.591* (0.007)

0.463** (0.036)

0.188* (0.000)

0.192* (0.003)

0.172* (0.000)

0.144* (0.000)

0.152* (0.000)

0.137* (0.009)

0.129* (0.003)

0.109* (0.000)

0.116* (0.000)

0.168* (0.000)

0.109* (0.000)

0.287* (0.000)

0.292* (0.001)

0.268* (0.000)

0.239* (0.000)

0.209* (0.004)

−9.864* (0.000) 19.550 (1.000)

−6.205* (0.000) 39.031 (0.513)

−7.423* (0.000) 14.417 (0.999)

−16.490* (0.000) 44.019 (0.927)

−14.205* (0.000)

−15.333* (0.000)

−17.662* (0.000)

−17.162* (0.000)

29.278 (0.894)

13.813 (1.000)

9.555 (1.000)

16.830 (0.999)

0.412 (0.680)

−0.549 (0.582)

−0.990 (0.322)

1.270 (0.204)

0.617 (0.536)

0.634 (0.526)

−0.745 (0.455)

−0.843 (0.398)

1.851 (0.064)

P-values are in (). Sargan test reflecting the Overidentifying restrictions test. AR (2) reflecting the Arellano–Bond test of the second-order autocorrelation in first differences. * indicates significance level with 1% ** indicates significance level with 5% *** indicates significance level with 10%.

by 0.904%, 0.796%, 0.713%, and 0.507% respectively. This outcome is pertinent in with the results of Kyophilavong et al. [49]. Also, renewable energy has a positive and significant weight on trade at 1%. A 1% increase in renewable energy raises trade for a whole, developed and developing countries by 0.102%, 0.135%, and 0.107% respectively. This outcome is not coherent with the results of Ben Jebli and Ben Youssef [46]. There are different reasons for this difference. The first reason is related to the time period data used. Contrary to the authors of the mentioned paper who used the period between 1980 and 2009, we are focused on the period between 1990 and 2012. The second reason is associated with the countries characteristics. Contrary to the concerned study which focused on one country study (Tunisia), our work is interested in a group of countries. The third reason is attached to the trade measurement. Compared with Ben Jebli and Ben Youssef who used import or export as trade measurement, in our study,

1% significance level. A 1% increase in capital creates the better enhancement of renewable energy in the developing countries (0.287) compared to developed countries (0.116). The high-income developing countries group registered the most capital effects (0.292). In Table 5, we present the estimated outcomes of Eq. (9) by using the data of whole, developed and developing countries. The results engender uni-directional impact running from economic growth, renewable energy and capital stock to trade. In the beginning, economic growths affect positively the trade variable at 1%. This gain is clearer in developed countries (1.981) compared with the group of developing countries (0.811). Inside the developed countries, this impact is more considerable in major developed countries (1.992) compared with others developed countries (1.764). A 1% increase in economic growth boosts trade for a highincome, upper middle-income, lower middle-income and lower-income Table 5 Results of Eq. (9).

tr(−1) gw rw k cte Sargan test (pvalue) AR2 test (pvalue)

Whole

Developed countries

Major developed countries

Others developed countries

Developing countries

High-income developing countries

Upper middleincome developing countries

Lower Middleincome developing countries

Lower-income developing countries

0.494* (0.000) 0.843* (0.000) 0.102*** (0.051) 0.272* (0.000) −3.512 (0.000) 67.949 (0.198)

0.199* (0.000)

0.211* (0.008)

0.137** (0.016)

0.448* (0.000)

0.609*** (0.095)

0.416* (0.007)

0.398* (0.000)

0.306* (0.000)

1.981* (0.000)

1.992* (0.008)

1.764* (0.009)

0.811* (0.000)

0.904* (0.000)

0.796* (0.002)

0.713* (0.000)

0.507* (0.003)

0.135* (0.000)

0.166* (0.000)

0.123* (0.000)

0.107* (0.002)

0.136* (0.018)

0.101** (0.016)

0.096* (0.001)

0.065* (0.001)

0.266* (0.000)

0.285* (0.000)

0.261* (0.000)

0.188* (0.000)

0.191* (0.000)

0.177* (0.000)

0.161* (0.000)

0.121* (0.010)

−10.891 (0.000) 22.723 (1.000)

−9.404 (0.000)

−2.453 (0.000)

−2.133 (0.000)

−2.671 (0.000)

−2.994 (0.000)

−2.426 (0.000)

46.295 (0.228)

−10.717 (0.000) 12.138 (1.000)

45.744 (0.896)

16.385 (0.999)

15.662 (0.999)

27.166 (0.939)

3.306 (1.000)

0.425 (0.670)

0.072 (0.942)

−1.515 (0.129)

−2.401** (0.016)

0.663 (0.506)

−1.267 (0.204)

0.117 (0.906)

0.088 (0.929)

−2.753* (0.005)

P-values are in (). Sargan test reflecting the Overidentifying restrictions test. AR (2) reflecting the Arellano–Bond test of the second-order autocorrelation in first differences. * indicates significance level with 1% ** indicates significance level with 5% *** indicates significance level with 10%.

532

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

income developing, developed, major developed, and others developed countries. There is a feedback relationship across trade and renewable energy consumption in the case of whole, developing and developed countries. The primary suggestions emerging from our study can be introduced as follows. First, the emergence of feedback relationship among renewable energy category and income explains the strong association between them. Inside of the groups of countries analyzed, the mentioned association proposes that energy strategies intended to expand the utilization of renewable energy producing a positive effect on output. It involves that policy makers should encourage renewable energy projects to beat the limitations on non-renewable energy consumption and to ameliorate the economic growth. Second, we discover a bi-directional causality between trade and economic growth. It implicates that trade plays a vital role to enhance the growth of developing and developed countries. It involves that policy that accelerate economic growth will prompt the trade. Also, it implies that policymakers should prompt the trade to facilitate the creation of a competitive advantage, accelerate the transfer of technical Knowledge, and attract the foreign direct investment. Third, we show a feedback connection between trade and renewable energy consumption. This implicates that trade openness need to outlay more renewable energy consumption. In addition, it involves that renewable energy help the integration of the developing and developed countries in international trade. Openness is an important vehicle serving the transfer of the renewable energy technology. Policy makers ought to take distinctive measures to develop the renewable energy division, for example, giving fiscal incentives for renewable energy, promoting investing into renewable energy projects.

we proposed the total of exports and import as trade indicator. The fourth reason is in link to the econometric approach adopted. In comparison to Ben Jebli and Ben Youssef who used the times series data method, in the current study we exploited the panel data methods. In addition, the capital stock contributes positively on the amelioration of trade with a 1% significance level. A 1% increase in capital generates more openness in the developed countries (0.266) compared to the countries which have the low level of development (0.188). Respecting the developed countries, this coefficient is more remarkable in the major developed countries group (0.285) compared with others developed countries group (0.261). In developing countries panel, the high-income group has the most important trade impact compared to others developing groups (0.191). The relationship among GDP, renewable energy consumption, and trade are recapitulated in the following. In the first place, the findings reveal the presence of bi-directional linkage among renewable energy consumption and GDP in whole, developing, high-income developing, upper middle-income developing, lower middle-income developing, lower-income developing, developed, major developed, and others developed countries [1,6,7,9,45]. It implicates that GDP and renewable energy are interconnected. The renewable energy contributes to enhancing the economic growth and this prosperity needs more renewable energy. This implies that policy makers should develop the renewable energy sector to achieve an important level of economic growth. Also, this discovering suggests that high growth expands interest for renewable energy and consequently, the legislatures of these countries ought to seek after dynamic arrangements to advance renewable energy. In the second place, the results illustrate the persistence of a bidirectional relationship between trade and renewable energy consumption in whole, developing, high-income developing, upper middleincome developing, lower middle-income developing, lower-income developing, developed, major developed, and others developed countries [45], suggesting that a high degree of trade drives to high degree of economic growth and reciprocally. First, the opening of trade requires a more intensive use of renewable energy. On the other hand, the development of renewable energy stimulates the opening of the country to international exchange. Decision makers ought to incorporate the renewable energy in any liberalization strategy. In the third place, actual results prove the presence of a bidirectional linkage among trade and economic growth in whole, developing, high-income developing, upper middle-income developing, lower middle-income developing, lower-income developing, developed, major developed, and others developed countries [45,49]. It indicates that GDP and trade are interdependent. This implies that policymakers should facilitate the openness to get a better degree of growth. Likewise, this finding suggests that the strong growth widens the interest for trade and, as a result, the policymakers should promote the trade.

References [1] Omri A, BenMabrouk N, Sassi-Tmar A. Modeling the causal linkages between nuclear energy, renewable energy and economic growth in developed and developing countries. Renew Sust Energ Rev 2015;42:1012–22. [2] Sebri M. Use renewable to be cleaner: meta-analysis of the renewable energy consumption–economic growth nexus. Renew Sust Energ Rev 2015;42:657–65. [3] Payne JE. On the dynamics of energy consumption and output in the US. Appl Energy 2009;86:575–7. [4] Menegaki AN. Growth and renewable energy in Europe: a random effect model with evidence for neutrality hypothesis. Energy Econ 2011;33:257–63. [5] Yildirim E, Sarac S, Aslan A. Energy consumption and economic growth in the USA: Evidence from renewable energy. Renew Sust Energ Rev 2012;16:6770–4. [6] Al-mulali U, Fereidouni HG, Lee JYM. Electricity consumption from renewable and non-renewable sources and economic growth: evidence from Latin American countries. Renew Sust Energ Rev 2014;30:290–8. [7] Apergis N, Payne JE. Renewable energy consumption and growth in Eurasia. Energy Econ 2010;32:1392–7. [8] Apergis N, Payne JE. Renewable and non-renewable energy consumption-growth nexus: Evidence from a panel error correction model. Energy Econ 2012;34:733–8. [9] Lin B, Moubarak M. Renewable energy consumption – economic growth nexus for China. Renew Sustain Energy Rev 2014;40:111–7. [10] Pao HT, Fu HC. Renewable energy, non-renewable energy and economic growth in Brazil. Renew Sustain Energy Rev 2013;25:381–92. [11] Shahbaz M, Loganathan N, Zeshan M, Zaman K. Does renewable energy consumption add in economic growth? An application of auto-regressive distributed lag model in Pakistan. Renew Sustain Energy Rev 2015;44:576–85. [12] Ohler A, Fetters I. The causal relationship between renewable electricity generation and GDP growth: a study of energy sources. Energy Econ 2014;34:125–39. [13] Tugcu CT, Ozturk I, Aslan A. Renewable and non-renewable energy consumption and economic growth relationship revisited: evidence from G7 countries. Energy Econ 2012;34:1942–50. [14] Ocal O, Aslan A. Renewable energy consumption-economic growth nexus in Turkey. Renew Sustain Energy Rev 2013;28:494–9. [15] Salim RA, Hassan K, Shafiei S. Renewable and non-renewable energy consumption and economic activities: Further evidence from OECD countries. Energy Econ 2014;44:350–60. [16] Fang Y. Economic welfare impacts from renewable energy consumption: the China experience. Renew Sustain Energy Rev 2011;15:5120–8. [17] Pirlogea C, Cicea C. Econometric perspective of the energy consumption and economic growth relation in European Union. Renew Sustain Energy Rev 2012;16:5718–26. [18] Inglesi-Lotz R. The impact of renewable energy consumption to economic growth: a

4. Conclusion and policy implications The goal of this study is to investigate the nexus between renewable energy consumption, foreign trade, and economic growth by using dynamic panel data framework between in developing and developed countries over the period 1990–2012. Our strong motivation is to fill the gap related in the absence of the studies discussing this subject in developing and developed countries. The major findings can be summarized as follows: There is feedback relationship between economic growth and renewable energy consumption in the case of whole, developing and developed countries. There is feedback relationship amongst economic growth and trade in the case of whole, developing, high-income developing, upper middle-income developing, lower middle-income developing, lower533

Renewable and Sustainable Energy Reviews 69 (2017) 527–534

F. Amri

2014;42:27–34. [36] Dritsaki C, Stiakakis E. Foreign direct investments, exports, and economic growth in Croatia: a time series analysis. Proced Econ Financ 2014;14:181–90. [37] Menyah K, Nazlioglu S, Wolde-Rufael Y. Financial development, trade openness and economic growth in African countries: new insights from a panel causality approach. Econ Model 2014;37:386–94. [38] Trejos S, Barboza G. Dynamic estimation of the relationship between trade openness and output growth in Asia. J Asian Econ 2014. [39] Greenaway D, Morgan WYN, Wright W. Trade liberalization and growth in developing countries: some new evidence. World Dev 1997;25:1885–92. [40] Ben Aïssa MS, Ben Jebli M, Ben Youssef S. Output, renewable energy consumption and trade in Africa. Energy Policy 2014;66:11–8. [41] Bloch H, Rafiq S, Salim R. Economic growth with coal, oil and renewable energy consumption in China: prospects for fuel substitution. Econ Model 2015;44:104–15. [42] Amri F. The relationship amongst energy consumption, foreign direct investment and output in developed and developing countries. Renew Sustain Energy Rev 2016;64:694–702. [43] Rafindadi AA, Ozturk I. Effects of financial development, economic growth and trade on electricity consumption: evidence from post-Fukushima Japan. Renew Sustain Energy Rev 2016;54:1073–84. [44] Omri A, Kahouli B. Causal relationships between energy consumption, foreign direct investment and economic growth: fresh evidence from dynamic simultaneous-equations models. Energy Policy 2014;67:913–22. [45] Sebri M, Ben-Salha O. On the causal dynamics between economic growth, renewable energy consumption, CO2 emissions and trade openness: fresh evidence from BRICS countries. Renew Sustain Energy Rev 2014;39:14–23. [46] Ben Jebli M, Ben Youssef S. The environmental Kuznets curve, economic growth, renewable and non-renewable energy, and trade in Tunisia. Renew Sustain Energy Rev 2015;47:173–85. [47] Arellano M, Bond S. Some tests of specification for panel data: Monte Carlo Evidence and an application to employment equations. Rev Econ Stud 1991;58(2):277–97. [48] Im KS, Pesaran MH, Shin Y. Testing for unit roots in heterogeneous panels. J Econ 2003;115:53–74. [49] Kyophilavong P, Shahbaz M, Anwar S, Masood S. The energy-growth nexus in Thailand: does trade openness boost up energy consumption?. Renew Sust Energ Rev 2015;46:265–74.

panel data application. Energy Econ 2015. [19] Rivera-Batiz LA, Romer PM. International trade with endogenous technological change. Eur Econ Rev 1991;35:971–1001. [20] Grossman GM, Helpman E. Growth and welfare in a small open economy. In. In: Helpman E, Razin A, editors. International trade and trade policy. Cambridge: MIT Press; 1991. p. p141–p163. [21] Shahbaz M. Does trade openness affect long run growth? Cointegration, causality and forecast error variance decomposition tests for Pakistan. Econ Model 2012;29:2325–39. [22] Sadorsky P. Energy consumption, output and trade in South America. Energy Econ 2012;34:476–88. [23] Solow RM. Technical change and the aggregate production function. Rev Econ Stat 1957;39:312–20. [24] Jin JC. Can openness be an engine of sustained high growth rates and inflation? Evidence from Japan and Korea. Int Rev Econ Financ 2006;15:228–40. [25] Balassa B. Export incentives and export performance in developing countries: a comparative analysis. Weltwirtschaftliches Arch 1978;114:24–61. [26] Harrison A. Openness and growth: a time-series, cross-country analysis for developing countries. J Dev Econ 1996;48:419–47. [27] Onafowora OA, OWOYE O. Can trade liberalization stimulate economic growth in. Afr Worl Dev 1998;26:497–506. [28] Greenaway D, Morgan W, Wright P. Trade liberalisation and growth in developing countries. J Dev Econ 2002;67:229–44. [29] Yanikkaya H. Trade openness and economic growth: a cross-country empirical investigation. J Dev Econ 2003;72:57–89. [30] Chandran VGR. Munusamy. trade openness and manufacturing growth in Malaysia. J Policy Model 2009;3:637–47. [31] Falvey R, Foster N, Greenaway D. Trade liberalization, economic Crises, and growth. Worl Dev 2012;40:2177–93. [32] Eris MN, Ulasan B. Trade openness and economic growth: Bayesian model averaging estimate of cross-country growth regressions. Econ Model 2013;33:867–83. [33] Szkorupová Z. A causal relationship between foreign direct investment, economic growth and export for Slovakia. Proced Econ Financ 2014;15:123–8. [34] Chang R, Kaltani L, Loayza NV. Openness can be good for growth: the role of policy complementarities. J Dev Econ 2009;90:33–49. [35] Sahoo AK, Sahoo D, Sahu NC. Mining export, industrial production and economic growth: a cointegration and causality analysis for India. Resour Policy

534