Professional Documents
Culture Documents
Ecological Economics
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e c o l e c o n
Analysis
On the causal dynamics between emissions, nuclear energy, renewable energy, and
economic growth☆
Nicholas Apergis a, James E. Payne b,⁎, Kojo Menyah c, Yemane Wolde-Rufael d
a
Department of Banking and Financial Management, University of Piraeus, Karaoli and Dimitriou 80, Piraeus, ATTIKI 18534, Greece
b
Department of Economics, Illinois State University, Normal, IL 61790-4200, United States
c
London Metropolitan Business School, London Metropolitan University, 84 Moorgate, London, EC2M 6SQ, United Kingdom
d
135 Carnwath Road, London SW6 3HR, UK
a r t i c l e i n f o a b s t r a c t
Article history: This paper examines the causal relationship between CO2 emissions, nuclear energy consumption, renewable
Received 23 January 2010 energy consumption, and economic growth for a group of 19 developed and developing countries for the
Received in revised form 16 June 2010 period 1984–2007 using a panel error correction model. The long-run estimates indicate that there is a
Accepted 17 June 2010
statistically significant negative association between nuclear energy consumption and emissions, but a
Available online 6 July 2010
statistically significant positive relationship between emissions and renewable energy consumption. The
results from the panel Granger causality tests suggest that in the short-run nuclear energy consumption
Keywords:
Emissions
plays an important role in reducing CO2 emissions whereas renewable energy consumption does not
Nuclear energy contribute to reductions in emissions. This may be due to the lack of adequate storage technology to
Renewable energy overcome intermittent supply problems as a result electricity producers have to rely on emission generating
Economic growth energy sources to meet peak load demand.
© 2010 Elsevier B.V. All rights reserved.
0921-8009/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ecolecon.2010.06.014
2256 N. Apergis et al. / Ecological Economics 69 (2010) 2255–2260
nuclear and renewable energy as possible panacea for emission long-term mean global temperature rise is to be limited to between 2
reductions has been conspicuous by its absence. To date, empirical and 2.4 °C (IEA, 2009a). In contrast, nuclear energy is envisaged to
research has focused on investigating the causal relationship between contribute to only 6% of the global CO2 reduction target by 2050. This
energy consumption and economic growth, on the one hand, and is largely due to long lead times, high capital costs and public
between energy consumption and pollutant emissions on the other opposition to nuclear power in some countries. Nevertheless, both
(Coondoo and Dinda, 2002; Dinda, 2004; Soytas et al., 2007; nuclear and renewable energy can potentially play a pivotal role not
Chontanawat et al., 2008; Aslanidis and Iranzo, 2009; Apergis and only in energy supplies, but also in emission reductions.
Payne, 2009, 2010a); Sadorsky, 2009a; Sari and Soytas, 2009; Soytas
and Sari, 2009; Ozturk, 2010; Payne, 2009, 2010a,b). There is a dearth
of empirical research that looks into the causal relationship between 3. Data, Methodology, and Results
nuclear energy consumption, renewable energy consumption, and CO2
emissions using modern advances in time series econometrics Annual data from 1984 to 2007 was obtained from the World Bank
associated with causality testing. The purpose of this study is to fill Development Indicators, CD-ROM and the Energy Information Adminis-
this gap by investigating the causal link between nuclear energy tration for Argentina, Belgium, Brazil, Bulgaria, Canada, Finland, France,
consumption, renewable energy consumption, CO2 emissions, and Hungary, India, Japan, Netherlands, Pakistan, South Africa, South
economic growth for a group of 19 developed and developing Korea, Spain, Sweden, Switzerland, U.K., and the U.S. The multivariate
countries for the period 1984–2007 within a panel error correction framework for the analysis includes real GDP (Y) in billions of constant
model framework. The main reason for studying the relationship 2000 U.S. dollars, nuclear electricity net consumption (N) in millions of
between carbon emissions, nuclear energy and renewable energy is kilowatt hours, total renewable electricity net consumption (R) in
that both nuclear and renewable energy play an important role in the millions of kilowatt hours, and total carbon dioxide emissions (E) from
current debate on environmental protection and sustainable devel- energy consumption in millions of metric tons.
opment. To our knowledge, this is the first study to use a The analysis begins with the examination of the stationarity
heterogeneous panel error correction model to explore the causal properties of the variables by employing a battery of panel unit root
relationships between nuclear and renewable energy consumption, tests.1 Levin et al. (2002) propose a panel based ADF test with
CO2 emissions, and economic growth. This allows us to draw homogeneity in the dynamics of the autoregressive coefficients for all
conclusions from the analysis which may be applicable to a broad panel units with cross-sectional independence. The Im et al. (2003)
range of countries. This is important in light of the fact that there is a panel unit root test differs from Levin et al. (2002) by allowing for
global effort to find ways of reducing emissions and the role of various heterogeneity in the dynamics of the autoregressive coefficients for all
energy sources in helping to achieve such an objective. panel units. Finally, the nonparametric panel unit root tests of
In the light of the discussion above, the rest of the study is Maddala and Wu (1999) which combine the p-values from individual
organized as follows. Section 2 briefly discusses the potential role of unit root tests are estimated using the Fisher–ADF and Fisher–PP tests.
nuclear and renewable energy in mitigating emissions. Section 3 For each panel unit root test the null hypothesis is a unit root while
describes the data and methodology along with the empirical the alternative hypothesis is no unit root. Table 1 displays the results
evidence. Concluding remarks are presented in Section 4. of the panel unit root tests which indicate that each variable is
integrated of order one.
2. Potential Role of Nuclear and Renewable Energy in Mitigating To determine whether the variables are cointegrated, the Larsson
CO2 Emissions et al. (2001, hereafter LLL) procedure is employed. The LLL(2001)
procedure parallels the Johansen (1988) methodology within a panel
Nuclear energy plays an important role not only in meeting the error correction model framework and offers several advantages over
energy needs of many countries, but also in mitigating emissions. residual-based cointegration tests. First, the LLL (2001) procedure
According to Adamantiades and Kessides (2009) the worldwide allows for more than one cointegrating vector unlike, for instance,
operation of nuclear plants makes a significant contribution to the Pedroni (1999, 2004) which assumes there is only one cointegrating
mitigation of greenhouse gas emissions where nuclear plants vector. Second, using the LLL (2001) method, no choice has to be
currently save some 10% of CO2 emissions from world energy use. made with respect to the normalization of variables. Finally, while the
Nuclear power plants have played a major role in reducing the cointegrating relations are restricted to each cross-section, the rest of
amount of greenhouse gas emissions produced by the electricity the model is unrestricted which allows for short-run dependence
sector in OECD countries (Nuclear Energy Agency, 2002). Further- between the groups, the most appealing of which is that of cross-
more, it is claimed that without nuclear power, OECD power plant sectional dependence of the error terms.
carbon dioxide emissions would have been about one-third higher. To begin the analysis one must determine whether the Π matrix is
Moreover, the European Union (2006) argues that Europe would not of reduced rank and whether all N groups can each be characterized
have been able to make any significant impact on reducing CO2 by r cointegrating vectors. If cointegration is found, the next step is to
emissions without the use of nuclear energy. determine whether or not the cointegrating vectors are homoge-
Like nuclear energy, renewable energy offers significant opportu- neous. LLL (2001) consider testing for cointegration under the
nities for further growth that can facilitate the transition to a global assumption that Πi = αiβ′i = αβ′ = Π for i = 1, . . , N. The null
sustainable energy supply by the middle of this century (IEA, 2009a). hypothesis of the LLL test is H0:rk(Πi) = k for i = 1, . . . , N against
In 2004, renewable energy accounted for 13.1% of world total primary the alternative hypothesis that H1:rk(Πi) = m for a non-vanishing
energy supply. The use of non-hydro emerging renewable energy fraction of cross-section members. This test statistic is similar to that
sources (wind, solar, geothermal, tidal, wave, and bio-energy) of Im et al. (2003) and is given by a centered and scaled version of the
exhibits the fastest rate of increase, with most of the increase in cross-sectional average of the individual trace statistics. LRsi (k|m)
power generation. The share of non-hydro renewable energy in total denotes the trace statistic for the null hypothesis of a k-dimensional
power output is expected to rise from 2.5% in 2007 to 8.6% in 2030, cointegrating space for unit i where the superscript s indicates the
with wind power having the largest absolute increase (IEA, 2009a). specification of the deterministic components. Using the central limit
According to the most optimistic scenario created by the IEA, by 2050,
the share of renewable energy in the electricity generation mix could
increase from 18% in 2004 to 39% by 2050. Renewable energy also 1
To conserve space, the details of the panel unit root and stationarity tests have
serves a vital role in the global 50% CO2 reduction target by 2050 if the been omitted.
N. Apergis et al. / Ecological Economics 69 (2010) 2255–2260 2257
Table 1 Table 2
Panel unit root tests 1984–2007. Panel cointegration tests and long-run parameter estimates 1984-2007.
Variables Panel test Levels First differences Panel A: LLL panel cointegration tests
Table 3
Panel causality tests 1984–2007.
ΔE ΔN ΔR ΔY ECT
Notes: partial F-statistics reported with respect to short-run changes in the independent variables. The sum of the lagged coefficients for the respective short-run changes is denoted
in parentheses. ECT represents the coefficient of the error correction term. Probability values are in brackets and reported underneath the corresponding partial F-statistic and sum of
the lagged coefficients, respectively. Significance at the 1%, 5%, and 10% levels are denoted by “a”, “b”, and “c”.
determined by the statistical significance of the partial F-statistic ables (Sims, 2004), there was a substantial fall in R&D investment in
associated with the corresponding right hand side variables. Long-run renewable energy in some developed countries (Nemant and Kammen,
causality is determined by the statistical significance of the respective 2007) while for other countries, the development of the renewable
error correction terms using a t-test. energy industry has not grown significantly due to the restricted size of
The results of the short-run and long-run Granger causality tests the home market for the technology or the industry is in its early stage of
are reported in Table 3. Eq. (2a) shows that renewable energy development (Jäger-Waldau, 2007; Lewis and Wiser, 2007). At the same
consumption and economic growth each have a positive and time, in some countries, natural gas prices dropped dramatically,
statistically significant impact on emissions in the short-run while making renewable energy less attractive when compared to natural
nuclear energy consumption has a negative and statistically signifi- gas-fueled generation (Martinot et al., 2005). Furthermore, as the IEA
cant impact. In terms of Eq. (2b), emissions and economic growth (2009a) argues, strong financial incentives are needed to support
each have a positive and statistically significant impact on nuclear renewable energy development, which have been inadequate in most
energy consumption whereas renewable energy consumption has a countries. Technology standards for renewable energy technologies and
statistically insignificant impact. In the case of Eq. (2c), both emissions fuels are also lacking with significant uncertainties about the rate at
and nuclear energy consumption each have a significant negative which specific investment cost of renewables can be reduced (De Vries
impact on renewable energy consumption while economic growth et al., 2007). In addition, the current financial crisis has hit dispropor-
has a positive and statistically significant impact. In regard to Eq. (2d), tionately hard the renewable energy sector leading to reductions in
both emissions and renewable energy consumption each have a investments (IEA, 2009a). It is also possible that most of these countries
positive and statistically significant influence on economic growth may not have reached the threshold point where renewable energy
whereas nuclear energy consumption has a negative and statistically starts to mitigate CO2 emissions. According to Chiu and Chang (2009),
significant impact. renewable energy supply has to account for 8.39% of total energy supply
One of the most important results that emerge from the short-run before any impact on mitigating CO2 emissions is observed. Also, as
causality tests presented in Table 3 is that these tests show that nuclear pointed out by Heal (2009) and Forsberg (2009), renewable energy may
energy plays an important role in reducing CO2 emissions. Nuclear not reduce emissions because the intermittent nature of its output and
energy consumption has a negative and statistically significant impact the lack of adequate storage technology for renewable energy means
on emissions. Our results are similar to Iwata et al. (2009) for France and that for large peak electricity production, a back-up power source is
Menyah and Wolde-Rufael (2010) for the US. In contrast, renewable required which is usually provided by fossil fuel.
energy did not contribute to reductions in emissions. This is also in line The short-run causality tests further show that there is bidirectional
with the finding of Menyah and Wolde-Rufael (2010) for the US. Whilst causality between renewable energy consumption and economic growth.
it is beyond the scope of this paper to fully explore the possible causes for This is in line with the findings of Apergis and Payne (2010b) for a panel of
the inability of renewable energy consumption to reduce CO2 emissions, 20 OECD countries and in line with the long-run causality found by
it is possible to suggest a number of factors that may have contributed to Sadorsky (2009b) for a panel of 18 emerging economies. The presence of
this.6 In the first place, relative to nuclear energy, the share of renewable bidirectional causality between renewable energy and economic growth
energy consumption in total energy consumption has not increased lends support for the feedback hypothesis whereby renewable energy
much in these countries. Even though it is known that significant consumption and economic growth are interdependent. This interdepen-
learning investments are required to ensure widespread use of renew- dency suggests that energy policies aimed at increasing the production
and the consumption of renewable energy will have a positive impact on
economic growth. Moreover, the positive influence of the use of
renewable energy on economic growth further enhances the viability of
6
In light of the positive relationship between renewable energy and emissions, a the renewable energy sector which provides additional support for the
referee suggested this result may be due to collinearity. Upon the recommendation of
assertion that renewable energy can serve as an important energy source
the referee the following two models were estimated: (1) carbon emissions, nuclear
energy consumption, and growth and (2) carbon emissions, renewable energy for these countries.
consumption, and growth to determine if the signs change significantly with respect to Table 3 also reveals there is bidirectional causality between CO2
the variables in question. In each case nuclear and renewable energy consumption, emissions and economic growth. This is line with the findings of
respectively yielded a negative coefficient estimate in the long-run cointegrating Apergis and Payne (2009) for a panel of six Central American
equation as well as in the short-run causality results in regard to emissions. Also, the
inclusion of fossil fuel consumption to the four-variable model did not alter the
countries and Apergis and Payne (2010a) for a panel of 11 countries
positive coefficient on renewable energy consumption. The results are available upon within the Commonwealth of Independent States. This interdepen-
request. dency of CO2 emissions and economic growth suggest that though
N. Apergis et al. / Ecological Economics 69 (2010) 2255–2260 2259
Payne, J.E., Taylor, J.P., 2010. “Nuclear energy consumption and economic growth in the Sims, R.E.H., 2004. Renewable energy: a response to climate change. Solar Energy 76,
US: an empirical note”. Energy Sources, Part B: Economics, Planning, and Policy 5, 9–17.
301–307. Soytas, U., Sari, R., 2009. Energy consumption, economic growth, and carbon
Pedroni, P., 1999. Critical values for cointegration tests in heterogeneous panels with emissions: challenges faced by an EU candidate member. Ecological Economics
multiple regressors. Oxford Bulletin of Economics and Statistics 61, 653–670. 68, 1667–1675.
Pedroni, P., 2004. Panel cointegration: asymptotic and finite sample properties of Soytas, U., Sari, R., Ewing, B.T., 2007. Energy consumption, income, and carbon
pooled time series tests with an application to the PPP hypothesis: new results. emissions in the United States. Ecological Economics 62, 482–489.
Econometric Theory 20, 597–627. Stern, N., 2007. Stern review on the economics of climate change. http://www.hm-
Reddy, B.S., Assenza, G.B., 2009. The great climate debate. Energy Policy 37, 2997–3008. treasury.gov.uk/sternreview_index.htm2007.
Sadorsky, P., 2009a. Renewable energy consumption, CO2 emissions, and oil prices in Toth, F.L., Rogner, H.H., 2006. Oil and nuclear power: past, present, and future. Energy
the G7 countries. Energy Economics 31, 456–462. Economics 28, 1–25.
Sadorsky, P., 2009b. Renewable energy consumption and income in emerging Vaillancourt, K., Labriet, M., Loulou, R., Waaub, J., 2008. The role of nuclear energy in
economies. Energy Policy 37, 4021–4028. long-term climate scenarios: an analysis with the World-TIMES model. Energy
Sari, R., Soytas, U., 2009. Are global warming and economic growth compatible? Policy 36, 2086–2097.
Evidence from five OPEC countries. Applied Energy 86, 1887–1890.