GeoJournal DOI 10.1007/s10708-014-9558-6
Energy consumption, CO2 emissions and GDP in Nigeria Sulaiman Chindo • Abdulsamad Abdulrahim Salisu Ibrahim Waziri • Wong M. Huong • Abdulfatah Abubakar Ahmad
•
Ó Springer Science+Business Media Dordrecht 2014
Abstract This paper investigated the relationship between energy consumption, carbon dioxide (CO2) emissions and GDP in Nigeria using autoregressive distributed lag approach to cointegration. The empirical results revealed that there is a long run relationship energy consumption, CO2 emissions and GDP. Both in the long run and short run, CO2 emissions has been found to have a significant positive impact on GDP, meaning that an increase in CO2 emissions facilitates GDP growth. On the other hand, energy consumption shows significant negative impact on GDP in the short run. We therefore, suggested that renewable source of energy such as solar and wind could be explored and considered as an alternative source of energy since Nigeria is well endowed with solar energy. This will assist in reducing CO2 emissions and at the same time sustaining long run growth in GDP. Keywords Energy consumption CO2 emissions GDP ARDL bounds testing
Introduction Global warming and climate change have been the topical issues in climatology, environmental science, S. Chindo (&) A. Abdulrahim S. I. Waziri W. M. Huong A. A. Ahmad University Putra Malaysia, Serdang, Selangor, Malaysia e-mail:
[email protected]
environmental economics, in the recent years owing to the devastating effects of the former on the latter. Thus, greenhouse emissions have been constantly increasing due to human activities, energy consumption, and fossil fuel combustions, etc. The pursuance of growth targets has made so many countries to involve in industrial production that requires higher energy consumption. However, among the components of greenhouse emissions, carbon dioxide (CO2) emissions has been regarded as the major contributor with about more than 60 % of the total of green gases (Kaygusuz 2009). International organizations such as the United Nations have made efforts to cut down the hostile effects of global warming and climatic changes through governmental binding agreements, for example Kyoto Protocol (Halicioglu 2009). This is a protocol to the United Nations Framework Convention on Climate Change (i.e. UNFCCC), which seek to reduce global warming. The major goal of this protocol is to achieve stabilized greenhouse gas concentrations in the atmosphere. It was founded in 1997 in Kyoto, Japan; however, the agreement was revalidated with actions in 2005. As of June 2013, there were 192 parties to this protocol who have ratified it. Nigeria as a party to this agreement ratified it on December 10, 2004 as 130th member (UNFCCC 2003) Climate models suggest that climate in Africa will be more variable with a high degree of uncertainty about its projections in Sahel zone. The West Africa’s temperature especially, the Sahel zone has raised significantly
123
GeoJournal
than the global trend (PACJA 2009). In the case of Nigeria, it is predicted that there may be a rise in sea level up to 0.3 m by 2020 and also 1 m by 2050, and temperature may rise up to 3.2 °C by 2050 (DFID 2009). The predicted rise of 1 m may lead to the loss of 75 % of the Niger Delta region through flooding. PACJA (2009) further asserted that by 2020, if no measure is taken, about 2–11 % of Nigeria’s GDP may be potentially lost. Nigeria’s average growth has been around 6 % since the last decade. Despite this remarkable development, the supply of electricity which is supposed to be the main source of energy in Nigeria is epileptic. This has therefore necessitated a shift from electricity usage to other alternative sources of power that requires the burning of fossil fuels. This thereby leads to an increase in toxic emissions. The energy consumption index in Nigeria increased from 2.8 % in 2010 relative to the index in 2009 which was put at 1.9 %. Consequently, emissions of greenhouse gases have also increased. To sum it, the average change of pollutants emitted between 1990 and 2009 was put at 41.3 % which has growing negative effect on climatic condition and subsequently accelerates global warming. Therefore, climate change has now become a development issue rather than just an environmental issue, which threatens the sustainable development of Nigeria. It is against this background that this study aims to ascertain the relationship between energy consumption, CO2 emissions and economic growth in Nigeria using an autoregressive distributed lag (ARDL) approach to cointegration. The remainder of this study is organized as follows. ‘‘Literature review’’ section discusses the empirical literature on the relationship between economic growth, CO2 emissions and energy consumption. ‘‘Methodology and data’’ section presents the data and methodology used in the study. ‘‘Results and discussion’’ section discusses the empirical findings, and the conclusion and policy implications are included in ‘‘Conclusions and policy recommendation’’.
Literature review Several studies have been conducted in this area to ascertain the relationship and the causal relationship between energy consumption and economic growth or energy consumption, CO2 emissions and economic growth. For instance, (Kraft and Kraft 1978) is regarded as the pioneer author to investigate the
123
relationship between energy consumption and economic growth in United States over the period of 1947–1974. Thereafter, empirical studies followed with different methods of analysis. Ang (2007) investigated the relationship between CO2 emissions, energy consumption and output in France and reported that there exists significant long run relationship among these variables. Ozturk and Acaravci (2010) examined the causal relationship between carbon dioxide emissions, energy consumption and economic growth using ARDL bounds testing approach of cointegration and error-correction based Granger causality models for nineteen European countries for period 1960–2005. They found evidence of a long-run relationship between carbon emissions per capita, energy consumption per capita, real GDP per capita and the square of per capita real GDP only for Denmark, Germany, Greece, Iceland, Italy, Portugal and Switzerland. Despite that, there is a long-run unidirectional causal relationship in those countries. Alam et al. (2011) investigate the causality relationships among energy consumption, carbon dioxide emissions and income in India by adopting a dynamic modeling approach. By utilizing an innovation accounting method to investigate profiles of the macroeconomic variables persisting from an unanticipated shock in innovation, there is an evidence of the existence of bidirectional Granger causality between energy consumption and carbon dioxide emissions in the long run. But neither carbon dioxide emissions nor energy consumption causes movements in real income. There is no causality relationship between energy consumption and income in any direction in the long run. Arouri et al. (2012) investigate the relationship between carbon dioxide emissions, energy consumption, and real GDP for 12 Middle East and North African Countries over the period 1981–2005. Employing recent bootstrap panel unit root tests and cointegration techniques, they found that in the longrun energy consumption has a positive significant impact on carbon dioxide emissions and more interestingly real GDP exhibits a quadratic relationship with carbon dioxide emissions for the regions as a whole. Bloch et al. (2012) investigate the relationship between coal consumption and pollutant emission both in short-run and long-run in China by applying both supply side and demand side frameworks using
GeoJournal
data period from 1977 to 2008 and 1965 to 2008. Under a supply side analysis, there is a unidirectional causality running from coal consumption to output in both short run and long run. While there is also a unidirectional causality running from income to coal consumption in the short run and long run under the demand side analysis. The results also reveal that there is a bi-directional causality between coal consumption and pollutant emission both in the short and long run. Alkhathlan and Javid (2013) examine the relationship between economic growth, carbon emissions and energy consumption at the aggregate and disaggregate levels in Saudi Arabia over the 1980–2011 periods. The findings are long term income elasticity of carbon emission in three of four models are positive and higher than estimated short term income elasticity. The results suggest that carbon emission increase with the increase in per capita income which supports the belief that there is a monotonically increasing relationship between per capita carbon emissions and per capita income for the aggregate model and for the oil and electricity consumption models. Besides that, the long and short term income elasticity of carbon emissions is negative for the gas consumption models. Ozturk and Acaravci (2010) examined the long run and short run causal relationship between economic growth, carbon emissions, energy consumption and employment ratio in Turkey using ARDL bounds testing approach to cointegration over the period of 1968–2005. Their findings indicated an evidence of long run relationship between the variables at 5 % significance level and the estimated results for the existence and direction of causality revealed that neither carbon emissions per capita or energy consumption per capita cause real GDP per capita, but employment ratio causes real GDP per capita in the short run. They therefore concluded that energy conservation policies such as rationing energy consumption and controlling carbon dioxide emissions have no adverse effect on real output. Menyah and Wolde-Rufael (2010) investigated the long run and causal relationship between economic growth, pollutant emissions and energy consumption in South Africa for the period of 1995–2006. By employing bounds testing approach to cointegration, the results revealed long run relationship among the variables. The further applied modified Granger causality test which show unidirectional causality running from pollutant emission to economic growth; from energy consumption to growth and from energy consumption to CO2 emissions.
Lotfalipour et al. (2010) investigated the causal relation between economic growth, carbon emissions and fossil fuels using Toda–Yamamoto causality test method in the case of Iran over the period of 1967–2007. They reported that unidirectional Granger causality runs from GDP and energy consumption to carbon emissions. They further indicated that carbon emissions and energy consumption do not lead to growth. Khan et al. (2013) explored the causal relationship between greenhouse emission, growth and energy consumption using cointegration and Granger causality test in Pakistan during 1975–2011. Their findings reveal that energy consumption serves as an important driver of CO2 emissions and also indicated unidirectional causality running from energy consumption to CO2 emissions. Jahangir Alam et al. (2012) examined the possible dynamic causality between energy consumption, carbon emissions and economic growth in Bangladesh using cointegration test and granger causality test. Unidirectional causality was reported from energy consumption to economic growth, unidirectional causality from energy consumption to CO2 emissions and CO2 emissions granger cause economic growth.
Methodology and data Despite the fact that the interrelationships between between environmental pollution, capital accumulation and other growth variables are important in growth theory (Xepapadeas 2005), there few studies which have investigated the causal relationship between economic growth, energy consumption and pollutant emissions that include labor and capital. Most previous studies employed model of only these three variables and some with the addition of employment ratio as a control variable. However, recently there have been few studies that have highlighted the importance of energy consumption and emissions as additional variables to traditional growth theory model to examine their impact on economic growth (see Menyah and Wolde-Rufael 2010; Ang 2008, 2009; Soytas et al. 2007; Zhang and Cheng 2009). In this paper, following these authors, we employ ARDL approach to cointegration recently developed Pesaran et al. (2001) to assess the relationship between CO2 emissions, energy consumptions and labour in the case of Nigeria over the period of 1970–2010.
123
GeoJournal
Bounds testing approach to cointegration (ARDL) We have choosen to use ARDL approach to cointegration owing to its comparative advantages over other methods of testing cointegration. As stated by Emran et al. (2007) and (Menyah and Wolde-Rufael 2010), the Monte Carlo evidence shows that it has several important advantages over the other conventional methods which include correcting the possible endogeneity of explanatory variables, good properties for small sample estimation, does not formally requires unit root test as it is not affected by order of integration of the variables and lastly it allows both long run and short run model to be estimated simultaneously. To specify our model, we begin with original traditional production theory model as follows: Y ¼ f ðL; KÞ
ð1Þ
where output is function of labour (L) and capital (K). Introducing energy consumption and CO2 into the model, we have the following model: Y ¼ f ðL; K; ECÞ
ð2Þ
where output is function of labour (L), capital (K) and energy consumption (EC). Also, CO2 emissions which is a by-product can be modeled as follows: CO2 ¼ hðf ðL; K; ECÞÞ
D ln Yt ¼ a0 þ
vi D ln Yti þ
i¼1
þ
n X
ui D ln COti þ
i¼0
n X
The null hypothesis of no cointegration is tested against the alternative based on the F-statistics value obtained when compared with two sets of critical values tabulated by Narayan and Smyth (2005). The two sets of critical values are I(0), the lower critical bound and I(1), the upper bound. If the calculated Fstatistics exceeds upper bound, cointegration exists, therefore we reject the null. If calculated F-statistics falls below the lower bound, cointegration does not exist, therefore we fail to reject the null. While if the Fstatistics falls in between the upper and lower bound, inconclusive, inference cannot be made unless the order of integration of the variables are known. If we find cointegration, we estimate the following long run model: ln Yt ¼ a1 þ
n X
v1i ln Yti þ
i¼1
þ
n X
u1i ln COti þ
i¼0
d1i ln EMti c1i ln ECti þ m1t
Thereafter, we proceed to the estimation of error correction model below:
þ
ci D ln ECti
i¼0
i¼0 n X
ð5Þ
di D ln EMti
i¼0 n X
n X
i¼0
n X
v2i D ln Yti þ
i¼1
þ
n X i¼0 n X
n X
d2i D ln EMti
i¼0
u2i D ln COti c2i D ln ECti þ kECTt1 þ e2t
ð6Þ
i¼0
ð4Þ
where lnYt is the natural log of GDP per capita, lnEMt represents the natural log of Employment rate, lnCOt
123
Ha : p1 6¼ p2 6¼ p3 6¼ p4 6¼ 0 ðCointegration existsÞ
D ln Yt ¼ a2 þ
þ p1 ln Yt1 þ p2 ln EMt1 þ p3 ln COt1 þ p4 ln ECt1 þ #t
H0 : p1 ¼ p2 ¼ p3 ¼ p4 ¼ 0 ðNo cointegrationÞ
ð3Þ
To determine the relationship between CO2 emissions, energy consumption and economic growth, we harmonised the models into one as adopted from Ozturk and Acaravci (2010). At this point, ARDL approach to cointegration is estimated based on the following unrestricted error correction regression: n X
is the natural log of CO2 emissions and lnECt is the natural log of energy consumption (proxied by fossil fuel consumption). To estimate the long run relationship, we test the joint significance of the coefficients of lagged level variables using F test under the following hypotheses:
where k is the error correction term which shows the speed of adjustment of the variables to equilibrium in the long run annually and ECTt is defined as:
-5.29*** (0.000)
***,**, and * denote significant at 1, 5 and 10 % levels respectively
-5.24*** (0.000)
-0.98 (0.933) -1.45 (0.546) -2.00 (0.576) -3.09** (0.037) -1.91 (0.630) -1.97 (0.297)
-6.73*** (0.000)
lnEMt
-5.65*** (0.000)
-4.46*** (0.001) -5.25*** (0.000) -4.50*** (0.000) -3.49* (0.053) -4.87*** (0.000) -4.91*** (0.000)
-3.29* (0.081)
-4.88*** (0.001) -4.92*** (0.000)
-6.89*** (0.000) -6.68*** (0.000)
-4.91*** (0.001)
lnECt
Constant and trend Constant and trend
-6.77*** (0.000)
Constant Constant
-4.95*** (0.000)
ADF PP ADF
Constant
First difference Level Variables
Table 1 Unit root test using ADF and PP
Testing for cointegration using ARDL approach does not formally require pre-testing of variables for unit root. Yet, it requires that the order of integration of the series must not be greater than one. Otherwise, the presence I(2) variable(s) will render the computed statistics presented by Pesaran and Shin (1999) and Narayan and Smyth (2005) invalid as the statistics are only computed based on I(0) and I(1). Based on this, we employed Augmented Dickey Fuller (ADF) and Phillips Perron (PP) to test for the order of integration of the series. The results of these tests (Table 1) shows that energy consumption and labor were I(0) whereas economic growth, CO2 emissions is I(1). In view of this results of having mixture of order of integration among the series, ARDL approach to cointegration is more appropriate to be applied than any other methods of testing cointegration. At the same time Toda and Yamamoto procedure is also the most preferred approach for testing causality. To test for cointegration, in addition to determining the order of integration of the series, it is also paramount to determine the optimal lag order n of the specified equation (4). The lag is selected rightly such that the error terms in the equation are not serially correlated. Thus, the lag order ought to be high enough to lessen serial correlation problems and at the same time, it ought to be low enough such that the conditional error correction model is not subjected to over-parameterization problems (Narayan and Smyth 2005). The selection of our optimal lag which is lag 3 has been done based Schwarz Bayesian Criterion (SBC) after checking for serial correlation at each lag. The cointegration test results (Table 2) reveal that there exist long run relationship among the variables under study as the calculated F-statistics (7.133) is greater than the upper bound critical value (6.610) at 1 % level of significance tabulated by Narayan and Smyth (2005) for 40 sample observations. This result
Constant and trend
PP
Cointegration test results based on ARDL
Constant
Results and discussion
-2.53 (0.309)
ð7Þ
-1.39 (0.848)
c1i ln ECti
i¼0
-1.98 (0.289)
i¼0
n X
-1.21 (0.658)
u1i ln COti
-2.53 (0.309)
n X
-0.96 (0.937)
d1i ln EMti
i¼0
-1.98 (0.291)
i¼1
n X
-0.78 (0.811)
v1i ln Yti
lnCOt
n X
lnYt
ECTt ¼ ln Yt a1
Constant and trend
GeoJournal
123
GeoJournal Table 2 The bounds test results Model
F-stats.
F(lnYt|lnEMt,lnCOt,lnECt)
7.133
Lag
3
Level of Sig.
Bounds test critical values (Unrestricted intercept and no trend) I(0)
I(1)
1%
5.018
6.610
5%
3.548
4.803
10 %
2.933
4.020
Based on Narayan table case III The bold signifies that, at both 5 % and 1 % significance level of the critical bounds values, the F-ststistics is greater than the upper bound values which indicate presence of strong cointegration relation Table 3 The estimated long run coefficients based SBC Dependent Variable, lnYt Regressors CO2 (lnCOt)
Coefficients
T-ratio (p value)
Table 4 The estimated short run coefficients from the error correction model based on SBC Dependent Variable, DlnYt Regressors
Coefficients
T-ratio (p value)
1.708
1.951* (0.061)
Energy (lnECt)
-0.799
-0.863 (0.395)
DlnCOt
0.354
Labour (lnEMt)
37.157
1.367 (0.182)
DlnECt
0.262
0.648 (0.522)
-147.710
-1.395 (0.174)
-0.947
-2.620** (0.014)
Constant
***, ** and * are significant at 1, 5 and 10 % levels respectively
indicates that, we can safely reject the null hypothesis of no cointegration among economic growth, energy consumption, CO2 emissions, labour and capital. Having found a long run relationship between our series, we estimated our long run model (Eq. 5) to obtain the long run coefficients whose results are presented in Table 3. The results indicate that CO2 emissions’ coefficient is positive and statistically significant at 10 %. This means that CO2 emissions has significant positive impact on economic growth. That is to say, higher CO2 emissions accelerates economic growth in the long run. To be precised, a 1 % increase in CO2 emissions leads to 1.71 % increase in economic growth. This finding is consistent with Nnaji et al. (2013) and Akpan (2012) in the case of Nigeria, Menyah and Wolde-Rufael (2010) who conducted the same study in South Africa, Alkhathlan and Javid (2013) in the case of Saudi Arabia, where CO2 emissions has been found to have an increasing relation with GDP per capita. Energy consumption has been found to have insignificant impact on growth in the long run. While, employment rate is insignificant in this case, which imply that labour does not significantly determine economic growth in the long run in the case of Nigeria.
123
DlnECt-1 DlnEMt
1.794* (0.083)
7.712
1.895* (0.068)
Constant
-30.659
-1.938* (0.062)
ECM(-1)
-0.208
-2.370** (0.024)
***, **, and * indicate significant at 1, 5 and 10 % levels respectively
The estimation results of the short run model are presented in Table 4. As displayed in the table, CO2 emissions has significant positive impact on economic growth. This signifies that, CO2 emissions has an increasing relationship with GDP in the short run, i.e. an increase in CO2 emissions leads to an increase in economic growth. This result substantiates the findings of Alkhathlan and Javid (2013) who reported monotonically positive relationship between CO2 emissions and growth in the short in the case of Saudi Arabia. Energy consumption has significant negative impact on economic growth. This imply that, an increase in fossil fuel consumption may lead to decrease in economic growth in the short run. This finding may sound counter-intuitive but, it is still in line with some past literatures reported for Nigeria (Wolde-Rufael 2006). The coefficient of employment rate is also positive and significant which is consistent with the theory. The error correction term is negative, less than one in absolute value and significant. It confirms the earlier long run relationship among the series and also shows the speed of adjustment of the series towards long run equilibrium to be 20 % in the first year.
GeoJournal Table 5 The results of the ARDL Diagnostic tests Test statistics
LM Version
F-version
A: Serial correlation
CHSQ (1) = 0.369 [0.544]
F (1, 28) = 0.282 [0.600]
B: Functional form
CHSQ (1) = 15.889 [0.000]***
F(1, 28) = 21.075 [0.000]***
C: Normality
CHSQ (2) = 0.778 [0.678]
Not applicable
D: Heteroscedasticity
CHSQ (1) = 0.920 [0.338]
F (1, 35) = 0.892 [0.351]
and heteroskedasticity test as we could not reject their null hypotheses. On the other hand, the model suffers from functional form problem. But on the general note, the model is can produce efficient and reliable estimate having passed the major diagnostic tests. The two Cusum tests (Figs. 1 and 2) for stability of the model over the sample observations or period indicate significance at 5 % level as the blue line is within the critical bounds. That is the model is stable over the studied period.
A: Langrange multiplier test of residual serial correlation B: Ramsey’s RESET test using the square of the fitted values C: Based on skewness and kurtosis of residual D: Based on the regression of squared residuals on squared fitted values ***, **, and * are significant at 1, 5 and 10 % levels respectively
To check the efficiency and reliability of the model we have conducted diagnostic tests which are reported in Table 5. As we all know, serial correlation is the major time series’ problem. The results reveal that model has passed serial correlation test, normality test
Fig. 1 Plot of cumulative sum of recursive residuals
Conclusions and policy recommendation As a party to kyoto protocol to UNFCCC Nigeria is facing a challenge on how to use fossil fuel in a way that will not increase CO2 emissions. Therefore, it is confronted with the crucial decision of balancing of fossil fuel consumption and reducing green house gases emissions. This paper investigated the cointegration relation between energy consumption, CO2 emissions and economic growth in Nigeria over the period of 1971–2010 using ARDL approach to cointegration
20 15 10 5 0 -5 -10 -15 -20 1974
1979
1984
1989
1994
1999
2004
2009
2010
The straight lines represent critical bounds at 5% significance level
Fig. 2 Plot of cumulative sum of squares of recursive residuals
1.5 1.0 0.5 0.0 -0.5 1974
1979
1984
1989
1994
1999
2004
2009
2010
The straight lines represent critical bounds at 5% significance level
123
GeoJournal
developed by Pesaran et al. (2001). The empirical results revealed that there is a long run relationship energy consumption, CO2 emissions and GDP. Both in the long run and short run, CO2 emissions has been found to have a significant positive impact on GDP, meaning that an increase in CO2 emissions facilitates increase in GDP. On the other hand, energy consumption shows significant negative impact on GDP growth in the short run. However, it is insignificant in the long run. These empirical evidences suggest that an attempt to reduce CO2 emissions resulting from fossil fuel consumption will affect GDP growth. As such, alternative sources of energy with least CO2 emissions need to be explored in order to tackle emission issue and at the same time not sacrificing GDP growth. The policy recommendation that could be deduced from the empirical evidence is that Nigeria should as part of its ongoing transformation program, green its energy policies and also diversify into other alternative energy sources with less greenhouse gas emission. Renewable energy sources specifically solar and wind may be considered since Nigeria is well endowed with solar energy. This will assist in reducing CO2 emissions and at the same sustaining long run economic growth. References Akpan, G. E. (2012). Electricity consumption. Carbon Emissions and Economic Growth in Nigeria, 2(4), 292–306. Alam, M. J., Begum, I. A., Buysse, J., Rahman, S., & Van Huylenbroeck, G. (2011). Dynamic modeling of causal relationship between energy consumption, CO2 emissions and economic growth in India. Renewable and Sustainable Energy Reviews, 15(6), 3243–3251. Alkhathlan, K., & Javid, M. (2013). Energy consumption, carbon emissions and economic growth in Saudi Arabia: An aggregate and disaggregate analysis. Energy Policy, 62(2013), 1525–1532. Ang, J. B. (2007). Are saving and investment cointegrated? The case of Malaysia (1965–2003). Applied Economics, 39(17), 2167–2174. Ang, J. B. (2008). Economic development, pollutant emissions and energy consumption Malaysia. Journal of Policy Modeling, 30(2), 271–278. Ang, J. B. (2009). CO2 emissions, research and technology transfer in China. Ecological Economics, 68(10), 2658–2665. Arouri, M. E. H., Ben Youssef, A., M’henni, H., & Rault, C. (2012). Energy consumption, economic growth and CO2 emissions in Middle East and North African countries. Energy Policy, 45(2012), 342–349. Bloch, H., Rafiq, S., & Salim, R. (2012). Coal consumption, CO2 emission and economic growth in China: Empirical evidence and policy responses. Energy Economics, 34(2), 518–528.
123
DFID (Department for International Development). (2009). Impact of climate change on Nigeria’s economy. Emran, M. H., Shilip, F., & Alam, M. I. (2007). Economic liberalization and price response of aggregate private investment, time series evidence from India. Canadian Journal of Economics, 40(3), 914–934. Halicioglu, F. (2009). An econometric study of CO2 emissions, energy consumption, income and foreign trade in Turkey. Energy Policy, 37(3), 1156–1164. Jahangir Alam, M., Ara Begum, I., Buysse, J., & Van Huylenbroeck, G. (2012). Energy consumption, carbon emissions and economic growth nexus in Bangladesh: Cointegration and dynamic causality analysis. Energy Policy, 45(2012), 217–225. Kaygusuz, K. (2009). Energy and environmental issues relating to greenhouse gas emissions for sustainable development in Turkey. Renewable and Sustainable Energy Reviews, 13(1), 253–270. Khan, M. A., Khan, M. Z., Zaman, K., Khan, M. M., & Zahoor, H. (2013). Causal links between greenhouse gas emissions, economic growth and energy consumption in Pakistan: A fatal disorder of society. Renewable and Sustainable Energy Reviews, 25(2013), 166–176. Kraft, J., & Kraft, A. (1978). On the relationship between energy and GNP. Journal of Energy and Development, 3(2), 401–403. Lotfalipour, M. R., Falahi, M. A., & Ashena, M. (2010). Economic growth, CO2 emissions, and fossil fuels consumption in Iran. Energy, 35(12), 5115–5120. Menyah, K., & Wolde-Rufael, Y. (2010). Energy consumption, pollutant emissions and economic growth in South Africa. Energy Economics, 32(6), 1374–1382. Narayan, P., & Smyth, R. (2005). Trade liberalization and economic growth in Fiji. An empirical assessment using the ARDL approach. Journal of the Asia Pacific Economy, 10(1), 96–115. Ozturk, I., & Acaravci, A. (2010). CO2 emissions, energy consumption and economic growth in Turkey. Renewable and Sustainable Energy Reviews, 14(9), 3220–3225. PACJA (Pan African Climate Justice Alliance). (2009). The economic cost of climate change in Africa. Pesaran, H. M., & Shin, Y. (1999). Autoregressive distributed lag modelling approach to cointegration analysis. In S. Storm (Ed.), Econometrics and economic theory in the 20th century: The Ragnar Frisch centennial symposium, 1999. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289–326. Soytas, U., Sari, R., & Ewing, B. T. (2007). Energy consumption, income, and carbon Emissions in the United States. Ecological Economics, 62(3), 482–489. Wolde-Rufael, Y. (2006). Electricity consumption and economic growth: A time series experience for 17 African countries. Energy Policy, 34(10), 1106–1114. Xepapadeas, A. (2005). Economic growth and the environment. In K. G. Ma¨ler, J. R. Vincent (Eds.), Handbook of environmental economics (Vol. 3, pp. 1219–1271). Zhang, X.-P., & Cheng, X.-M. (2009). Energy consumption, carbon emissions, and economic growth in China. Ecological Economics, 68(10), 2706–2712.