speech

Energy Policy 55 (2013) 483–489

Contents lists available at SciVerse ScienceDirect

Energy Policy

0301-42

http://d

n Tel.:

E-m

journal homepage: www.elsevier.com/locate/enpol

The contribution of foreign direct investment to clean energy use, carbon emissions and economic growth

Jung Wan Lee n

Administrative Sciences Department, Boston University, 808 Commonwealth Avenue, Boston, MA 02215, United States

H I G H L I G H T S

c FDI inflows strongly lead to economic growth in the G20. c FDI inflows lead to an increase in energy use in the G20. c FDI inflows are in no relation to CO2 emissions in the G20. c FDI inflows are in no relation to clean energy use in the G20. c Economic growth is in negative relation to CO2 emissions in the G20.

a r t i c l e i n f o

Article history:

Received 10 April 2012

Accepted 11 December 2012 Available online 9 January 2013

Keywords:

Clean energy use

Foreign direct investment

Carbon emissions

15/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.enpol.2012.12.039

þ1 617 358 5627; fax: þ1 617 353 6840.

ail addresses: [email protected], jwlee1119@g

a b s t r a c t

The paper investigates the contributions of foreign direct investment (FDI) net inflows to clean energy

use, carbon emissions, and economic growth. The paper employs cointegration tests to examine a long-

run equilibrium relationship among the variables and fixed effects models to examine the magnitude of

FDI contributions to the other variables. The paper analyzes panel data of 19 nations of the G20 from

1971 to 2009. The test results indicate that FDI has played an important role in economic growth for the

G20 whereas it limits its impact on an increase in CO2 emissions in the economies. The research finds

no compelling evidence of FDI link with clean energy use. Given the results, the paper discusses FDI’s

potential role in achieving green growth goals.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Within policy circles, there is a widespread belief that foreign direct investment (FDI) enhances the productivity of host coun- tries and promotes economic growth. The notion supports FDI may not only provide direct capital financing but may also create positive externalities via the adoption of foreign technology and know-how. Batten and Vo (2009) have shown that FDI stimulates economic growth through technology transfer, spillover effects, productivity gains, and the introduction of new processes and managerial skills. Fernandes and Paunov (2012) have recently shown that FDI has positive effects on innovation activities and manufacturing productivity. Hermes and Lensink (2003) reported that FDI plays an important role in modernizing the economy and promoting economic growth.

The historical data released by the World Bank indicate that FDI may have played an important role in addressing the growth challenges, in particular, in the group of twenty (G20) countries.

ll rights reserved.

mail.com

The G20 is a group of heads of government or state from 20 leading economies, 19 countries plus the European Union, includ- ing Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, South Korea, Mexico, Russia, Saudi Arabia, South Africa, Turkey, the United Kingdom, and the United States. Collectively, the G20 economies account for more than 80 percent of the gross world product, 80 percent of the world trade, and 62 percent of the world population, according to data from the World Growth Indicators. Most of the G20 economies are growing rapidly and as economic growth increases so too does the demand for energy. According to the International Energy Agency (2007), between 2005 and 2030 the world energy demand is expected to grow at an average annual rate of 1.8%. The G20 economies will contribute to 84% of the increase in the world energy demand.

Table 1 displays the summary statistics of 19 countries of the G20 during 1971–2010. There is a great deal of variation in mean per capita income with the highest mean per capita income levels 50,746 US dollars in Australia and the lowest 1375 US dollars in India with exhibiting an average of 23,078 US dollars in the G20, which is 252 percent higher than that of the world, an average of 9157 US dollars per capita in 2010. FDI net inflows per capita indicate a great deal

Table 1 Summary statistics, the G20 countries.

Country Name GDP (1) CO2 (2) Energy (3) Clean energy (4) FDI (5)

Argentina 9123 4.84 1853 6.78 3339

Australia 50,746 18.56 5642 1.31 15,032

Brazil 10,992 2.05 1242 15.61 2170

Canada 46,212 16.32 7481 21.26 17,896

China 4433 5.30 1695 3.65 1057

France 39,170 5.85 4060 44.72 13,821

Germany 39,851 9.58 4053 13.09 8917

India 1375 1.46 559 2.34 163

Indonesia 2952 1.72 851 8.39 281

Italy 33,788 7.43 2813 5.95 4238

Japan 43,063 9.46 3883 17.16 934

Korea 20,540 10.40 4989 15.93 1524

Mexico 9133 4.30 1497 6.26 3258

Russia 10,481 12.03 4558 9.01 2138

Saudi Arabia 16,423 16.56 5888 0 6589

South Africa 7272 8.93 2920 2.44 988

Turkey 10,049 4.00 1440 6.51 1515

United Kingdom 36,186 8.51 3282 8.53 24,678

United States 46,701 17.95 7225 11.59 10,862

Consolidated

Mean 23,078 8.70 3470 10.55 6284

St. Deviation. 17,437 5.48 2107 10.12 7045

Minimum 1375 1.46 559 0 163

Maximum 50,746 18.56 7481 44.72 24,678

World average 9157 4.76 1790 9.22 2853

(1) GDP per capita in 2010 (current US$).

(2) CO2 emissions per capita in 2009 (metric tons).

(3) Energy use per capita in 2010 (kilogram of oil equivalent).

(4) Clean energy use (% of energy use).

(5) FDI net inflows per capita during 1971–2010 (current US$).

J.W. Lee / Energy Policy 55 (2013) 483–489484

of variation with the highest mean per capita FDI levels 24,678 US dollars in the United Kingdom and the lowest 163 US dollars in India, with exhibiting an average of 6284 US dollars in the G20. It is 220 percent higher than that of the world, an average of 2853 US dollars per capita during the period from 1971 to 2010. The mean energy use per capita ranges from 7481 kg of oil equivalent in Canada to 559 kg in India with exhibiting an average of 3470 kg per capita in the G20, which is 194 percent larger than that of the world, an average of 1790 kg per capita in 2010.

According to a report by the International Energy Agency (2011), the CO2 emission levels are relatively high in most of the G20 countries. Table 1 displays that the mean CO2 emission per capita ranges from 1.46 metric tons in India to 18.56 metric tons in Australia with exhibiting an average of 8.70 metric tons in the G20, which is 183 percent larger than that of the world, an average of 4.76 metric tons per capita in 2009. Along with the rapid economic growth in the G20 countries, the emission levels have been growing fast. Although a wide gap exists among the G20 countries, FDI and the growth of specific business sectors such as manufacturing and infrastructure may be putting tremendous pressure on energy resources and the environment in the countries. However, environ- mental problems know no economic boundaries since they are complex in nature and transcend national boundaries.

In the 2008 G20 Summit, the G20 discussed clean energy, economic growth, and the fiscal elements of growth. Since the meeting, understanding the determinants of energy demand and the use of clean energy is essential for making better energy policies in the future. A better understanding of how to manage global emissions of greenhouse gases is critical because energy related emissions make up mainly the bulk of CO2 emissions. In this regard, the challenge facing the G20 is how to develop policy responses to counter the effects of the current environmental problems and climate change and lay the foundations for sustain- able growth that achieves economic growth and at the same time

reduces the CO2 emissions from the results of their economic growth. In addition, the increased economic importance of FDI raises new questions for the governments regarding the best policy frameworks to encourage continued economic growth, the reduction of CO2 emissions, the efficient use of energy resources, and the increased use of clean energy resources.

In sum, this paper assumes that FDI contributes to economic growth whereas it may also lead to an increase in energy consump- tion, and thus result in high CO2 emissions. Following the assump- tion above, FDI leads to an increase in CO2 emissions while it may also lead to the increased use of clean energy.

2. Literature review and hypotheses

2.1. FDI and economic growth

Hypothesis 1. FDI leads to economic growth.

The role of capital investment in economic growth has been considered one of the basic principles in economics. Many research- ers conclude that the rate of capital formation determines the rate of economic growth (Blomstrom et al., 1996; Ekanayake and Vogel, 2003; Tsang and Yip, 2007). For example, De Long et al. (1992) found a strong causal relationship between equipment investment and economic growth. Blomstrom et al. (1996) also reported that the growth rate is more closely related to the capital formation rates in succeeding periods than to the contemporary or preceding rates. Alfaro et al. (2010) have shown that FDI leads to higher additional growth in developed economies. Lee and Chang (2009) reported that FDI has a large direct effect on economic growth and extends the potential gains associated with FDI.

FDI is of special interest due to its supposed positive effects on growth. There is a widely accepted view that FDI promotes growth not only directly by augmenting capital formation in the recipient economy, but also indirectly by inducing human capital growth, helping technology transfers, and strengthening competi- tion (Aitken et al., 1997; Kneller and Pisu, 2007). Thanks to these potential merits of FDI, both developing and developed countries have become more receptive to FDI inflows, and the global FDI flows have continued to increase except for short declines during 1982–1983, 1991–1992, 2001–2003, and 2008–2010.

Aitken et al. (1997) have shown evidence of beneficial spillovers from multinational enterprises to the host economy, whereas Hsiao and Shen (2003) reported that economic growth is one of the important factors in attracting FDI, in particular in developing countries. Some studies indicate that the direction of causality between economic growth and FDI is subject to country-specific factors (Zhang, 2001). Kim and Seo (2003) reported that FDI has a positive but insignificant effect on GDP growth, while GDP growth has a significant and highly persistent effect on the future level of FDI in South Korea. Qi (2007) reported that the countries that are heavily dependent on petroleum exports have more difficulties than other countries in benefiting from FDI, and the role of total investment in impelling growth is also weakened in oil-exporting countries. The findings in the literature indicate that a country’s capacity to take advantage of FDI externalities might be limited by local conditions.

2.2. Economic growth, energy consumption and CO2 emissions

Hypothesis 2. Economic growth is in positive relation to CO2 emissions.

A fairly large amount of literature finds a causal relationship between energy consumption and economic growth, especially in

J.W. Lee / Energy Policy 55 (2013) 483–489 485

OECD countries (Lee et al., 2008), in the G7 countries (Narayan and Smyth, 2008), in OPEC member countries (Squalli, 2007), in African countries (Akinlo, 2008; Wolde-Rufael, 2009), in Central America (Apergis and Payne, 2009), in South America (Yoo and Kwak, 2010), in the Middle East (Al-Iriani, 2006; Narayan and Smyth, 2009), in Asian countries (Chen et al., 2007; Lee and Chang, 2008), in the Common- wealth of Independent States (Apergis and Payne, 2010), in European countries (Ciarreta and Zarraga, 2010), in developing countries (Lee, 2005; Sari and Soytas, 2007), and in developed and developing countries (Chontanawat et al., 2008; Mahadevan and Asafu-Adjaye, 2007; Sharma, 2010). They find that economic growth exerts a Granger causal influence on energy consumption in the long-run, and energy consumption points to output growth in the short run.

Though the general consensus of these studies is that there is a positive correlation between economic growth and energy con- sumption, some results contradict. For example, Huang et al. (2008) and Costantini and Martini (2009) argued that the causal relationship between economic growth and energy consumption is mixed depending on the functional form adopted and the sample of countries analyzed. More recently, some researchers have examined the time series dynamics between income and CO2 emissions to infer the direction of causality, for example, for the Commonwealth of Independent States (Apergis and Payne, 2010), for a panel of 109 countries (Lee and Lee, 2009), for a panel of 43 developing countries (Narayan and Narayan, 2010), and in developed countries (Coondoo and Dinda, 2002). However, the empirical results of the relationship between economic growth and CO2 emissions are mixed.

In this regard, many researchers employ a combined approach by examining the dynamic relationships between economic growth, energy consumption, and CO2 emissions together, especially in the EU (Keppler and Mansanet-Bataller, 2010), in Asian-Pacific countries (Niu et al., 2011), in the BRIC countries (Pao and Tsai, 2010), in France (Ang, 2007), in India (Ghosh, 2010), and in China (Zhang and Cheng, 2009). They find that economic growth is in positive relation to CO2 emissions. The results of other studies show that there are different causal links between economic growth, energy consump- tion, and CO2 emissions at different stages of economic growth (Dinda and Coondoo, 2006; Soytas and Sari, 2009). In sum, this study assumes that higher economic growth may require greater energy consumption and thus result in high CO2 emissions.

2.3. FDI, energy consumption, clean energy use and CO2 emissions

Hypothesis 3. FDI is in positive relation to CO2 emissions.

Hypothesis 4. FDI is in positive relation to clean energy use.

While there is a fairly large amount of literature investigating the link between economic growth and energy demand, the impact of FDI on the demand for energy is a topic that has received little attention. Mielnik and Goldemberg (2002) found a positive relationship between FDI and energy intensity in a sample of 20 developing countries. Sadorsky (2010) also found a positive and statistically significant relationship between FDI and energy consumption in a sample of 22 developing economies. FDI allows businesses cheaper and/or easier access to financial capital, which can be used to expand their existing operations or construct new plants and factories, all of which increase the demand for energy. Consistent with this view that FDI leads to greater economic growth is the likelihood that energy demand should be positively affected by increases in FDI. If FDI has an impact on the demand for energy, then this relationship can affect energy policy and carbon emissions strategies.

In addition, the G20 economies undergoing a process of rapid industrialization and urbanization will grow continuously, which may inevitably cause a large increase in CO2 emissions. Besides,

the CO2 emissions in a country do not necessarily depend on its income level alone; FDI may be another source. Where govern- ments attract FDI to accelerate their economic growth this may instigate an increase in CO2 emissions. There is much evidence of the significant influence of FDI on CO2 emissions. Zhang (2011) has shown that FDI plays a pivotal role in the increase of CO2 emissions in China. Xing and Kolstad (2002) reported that a positive relationship exists between FDI and pollutant emissions in the host countries.

However, List and Co (2000) reported that the inflow of FDI helps promote the energy efficiency of the host countries and cut CO2 emissions. Tamazian et al. (2009) reported that FDI helps enterprises promote technology innovation and adopt new technologies and thus increase energy efficiency and advance low carbon economic growth. Sadorsky (2009) reported that renewable energy consump- tion is related to income. In sum, it seems that the nexus of FDI, energy consumption, clean energy use and CO2 emissions has appeared unclear till now, and further empirical study is necessary.

3. Data and methodology

3.1. Data

This section describes the data and outlines the methodology used in the growth or selection of the indicators and the normal- ization of the data.

3.1.1. GDP

The real GDP is used to measure economic growth. Real GDP figures allow us to calculate a GDP growth rate, which tells us how much a country’s production has increased or decreased in comparison with the previous year.

3.1.2. CO2 CO2 emissions are used to measure the level of environmental

degradation. Data is in thousand metric tons. The CO2 Information Analysis Center and the International Energy Agency collect data and publish these statistics.

3.1.3. FDI

FDI net inflows are used to measure FDI. FDI is the sum of equity capital, reinvestment of earnings, other long-term capital and short-term capital as shown in the balance of payments. Data is in current US dollars.

3.1.4. Energy

Energy represents energy use in thousand metric tons of oil equivalent. Energy use refers to the use of primary energy that refers to the energy forms before transformation to other end-use fuels, which is equal to indigenous production plus imports and stock changes, minus exports and fuels supplied to ships and aircraft engaged in international transport. Primary energy dis- tinguishes from energy carriers, which correspond to the concept of secondary energy in energy statistics. The International Energy Agency collects data and publishes these statistics.

3.1.5. Clean energy

Clean energy is non-carbohydrate energy that does not pro- duce carbon dioxide when generated. It includes hydropower, nuclear, geothermal, and solar power, among others. Clean energy use represents alternative and nuclear energy use as a percentage of the total energy use. The International Energy Agency collects data and publishes these statistics.

Table 4 Results of Fisher-type Johansen panel cointegration test.

J.W. Lee / Energy Policy 55 (2013) 483–489486

The sample is restricted to the period for which annual data are available: from 1971 to 2009 (39 observations for each country). The above time series data are collected and retrieved from the World Growth Indicator database published by the World Bank. Since some indicators are expressed as thousand metric tons while others are expressed as US dollars, the normalization of the data is necessary before any aggregation can be made. Therefore, trans- formation into a natural log is carrying out to mitigate the possible distortions of the dynamic properties of the series. Log transforma- tion is a preferred method since each resulting coefficient in regression equation represents elasticity which is the ratio of the incremental change of the logarithm of a function with respect to an incremental change of the logarithm of the argument.

Table 2 displays the results of Pearson correlation analysis among the panel series. Correlations between FDI and GDP, energy use, CO2 emissions, and clean energy use are all highly significant. FDI is positively related to GDP and positively related to energy use and clean energy use. All in pairs exhibit positively and highly significant correlation. The pairwise relationship may change when all variables are included in panel based multivariate regression models.

3.2. Panel unit root test

Many economic variables are characterized by stochastic trends that might result in spurious inference. A time series is said to be stationary if the autocovariances of the series do not depend on time. Any series that is not stationary has a unit root. The formal method of testing stationarity is the unit root test. The recent literature suggests that panel based unit root tests have a higher power than unit root tests based on individual time series (Breitung, 2000; Im et al., 2003; Levin et al., 2002; Maddala and Wu, 1999).

Table 3 displays the results of the panel unit root tests. All of the test equations that include individual intercepts as regressors were tested by the least squares method. The Breitung unit root test equation includes individual fixed effects and individual linear trends as regressors. The probabilities for Fisher-type tests are computed using an asymptotic chi-square distribution

Table 2 Pearson correlations.

GDP FDI Energy CO2 Clean energy

GDP 1.000

FDI 0.959nnn 1.000

Energy 0.929nnn 0.886nnn 1.000

CO2 0.855 nnn 0.800nnn 0.917nnn 1.000

Clean energy 0.933nnn 0.908nnn 0.814nnn 0.729nnn 1.000

nnn p-valueo0.01, Correlation is significant at the 0.01 level (2-tailed).

Table 3 Results of panel unit root test.

Methods/variables GDP CO2

Levin et al. Level

1st dif.

�1.566

�8.698nnn �1.393

�11.209nnn

Breitung Level

1st dif.

4.929

�3.155nnn �0.398

�7.052nnn

Im et al. Level

1st dif.

�0.452

�9.622nnn �0.636

�13.014nnn

Fisher-ADF Level

1st dif.

38.186

172.395nnn 43.723

238.944nnn

Fisher-PP Level

1st dif.

41.155

233.360nnn 46.542

373.389nnn

Note: Probability values for rejection of the null hypothesis of a unit root are employe nnn p-valueo0.01. nn p-valueo0.05 based on MacKinnon (1996) one-sided p-values.

(Maddala and Wu, 1999). The numeric values show that the null hypothesis of a unit root cannot be rejected at the level of the series, implying that each time series is panel non-stationary. However, when applying the panel unit root tests to the first difference of the series, the null hypothesis for each of the series can be rejected at the 1% significance level. All the series with the first difference are stationary, which confirms that they are integrated of first order.

3.3. Panel cointegration test

Engle and Granger (1987) reported that a linear combination of two or more non-stationary time series may be stationary. Accord- ing to Granger (1988), cointegration means that the two or more non-stationary variables are integrated in the same order with the stationary of residuals. If the variables are cointegrated, there exists a force that converges into a long-run equilibrium. The cointegrating equation may be interpreted as a long-run equilibrium relationship among variables. In testing cointegration equations, Maddala and Wu (1999) proposed that the Fisher-type panel cointegration test using the Johansen (1991) test methodology is more efficient than using the Engle–Granger test method because the maximum like- lihood procedure has significantly large and finite sample properties. The Johansen procedure uses two ratio tests, a trace test and a maximum eigenvalue test, to test the number of cointegration relationships. Both can be used to determine the number of cointegrating vectors present although they do not always indicate the same number of cointegrating vectors. If trace statistics and maximum eigenvalue statistics yield different results, the result of the maximum eigenvalue test is preferred because of the benefit of carrying out separate tests on each eigenvalue.

Table 4 reports the results of the Johansen panel cointegration test. The test equations were tested by the panel least squares

Energy Clean energy FDI

�1.385

�6.424nnn �1.313

�9.768nnn �1.188

�9.016nnn

5.855

�2.167nn 4.633

�8.014nnn 1.463

�2.881nnn

�0.308

�10.726nnn �0.397

�8.402nnn �0.584

�11.472nnn

31.300

193.669nnn 34.679

192.742nnn 40.324

207.493nnn

32.914

321.153nnn 40.840

329.609nnn 43.266

320.464nnn

d at the 0.05 level.

Number of cointegrating equations Trace statistic Maximum-eigen statistic

None 460.2nnn 323.7nnn

At most 1 200.1nnn 131.6nnn

At most 2 98.4nnn 59.1nn

At most 3 49.7 28.2

At most 4 12.5 12.5

Note: The test equation is as follow: GDP¼f(FDI, Energy, CO2, Clean energy).

Probability values for rejection of the null hypothesis of no cointegration are

employed at the 0.05 level. nnn p-valueo0.01. nn p-valueo0.05 based on the Mackinnon et al. (1999) p-values.

Table 5 Results of hypothesis test using fixed effects models.

Variables/models Regression modela Regression modelb Regression modelc

GDP �0.053nnn 0.332nnn

FDI 0.201nnn 0.000 �0.023

CO2 �0.608 nnn

�1.131nnn

Energy 1.735nnn 1.171nnn 1.742nnn

Clean energy 0.176nnn �0.053nnn

Constant 8.974 0.319 �13.165

R-squared 0.930 0.990 0.804

Adj. R-squared 0.928 0.990 0.797

F-statistic 382.303 2974.273 116.557

Note: Probability values for rejection of the null hypothesis of zero (0) coefficient

are employed at the 0.01 level. nnn p-valueo0.01. a Regression model 1: GDP¼f(FDI, CO2, Energy, Clean energy). b Regression model 2: CO2¼f(GDP, FDI, Energy, Clean energy). c Regression model 3: Clean energy¼f(FDI, GDP, CO2, Energy).

J.W. Lee / Energy Policy 55 (2013) 483–489 487

method. For the Johansen panel cointegration test, the assump- tions of cointegration tests allow for individual effects but no individual linear trends in vector autoregression. The null hypoth- esis of no cointegration is rejected at the 0.01 level. Moreover, the trace statistic indicates that at least three cointegrating vectors exist at the 0.01 level and the maximum eigenvalue statistic also indicates that at least three cointegrating vectors exist at the 0.05 level. The cointegrating equations imply the presence of a long- run relationship between the variables.

Granger (1988) suggested that if two time series variables are not cointegrated, then there may be unidirectional or bidirectional Granger causality in the short-run. The Granger causality is tested by the joint significance of the coefficient of the differenced explanatory variable by using an F-test or Wald test. The pairwise Granger causality test provides F-statistic of coefficients on the lagged endogenous variable, which interprets the statistical sig- nificance of the coefficient of the regressor. In this way, the F-statistic can be used to find the Granger causal effect on the dependent variable. The hypothesis in this test is that the lagged endogenous variable does not cause the Granger dependent vari- able. However, the results of the panel cointegration test indicate that the time series variables are cointegrated. Therefore, this study will not employ Granger causality tests.

3.4. Testing the hypotheses using fixed effects models

Engle and Granger (1987) and Granger (1988) reported that if non-stationary variables are cointegrated, a corresponding error correction representation in the short-run dynamics of the vari- ables can be influenced by the deviation from equilibrium. Given the results of a long-run equilibrium relationship from the panel cointegration test, a panel based error correction model is used to account for a long-run relationship using the two-step procedure. Accordingly, panel based error correction models can be con- structed as follows:

DInY it ¼aiY þ Xn�1

j ¼ 1

bijDInXit�jþ Xn�1

j ¼ 1

o1jDInY it�jþpiECT t�jþeit ð1Þ

where Y it is the observation of the dependent variable for country i at time t. t represents 1, 2, 3, y, n observations. D is the difference operator. a is the deterministic component (constant). b, o, and p are the parameters of regressors. ECT t�1 is the error correction term obtained from the cointegrating vectors. eit is a stationary random error with a zero mean. j is the lag length.

In testing panel data analytic models, there are several types of test techniques, including constant coefficient models, fixed effects models, and random effects models. The quandary of random effects or fixed effects models can be solved using the Hausman specification test. The Hausman test is the classical test of whether the fixed or random effects models should be used. The cross-section random effects test equations are tested using the method of panel two stage least squares. The results of the Hausman test indicate that the fixed effects model is more robust for each case of using GDP (Chi-square statistic¼23.142), CO2 emissions (Chi-square statistic¼11.679), and Clean energy use (Chi-square statistic¼24.385) as the dependent variable. Prob- ability values for rejection of the null hypothesis of no correlation are employed at the 0.05 level. The results of the Hausman test support that we impose time independent effects for each entity that are possibly correlated with the regressors.

4. Results

Table 5 reports the results of panel based error correction models using fixed effects models. In Table 5, the numeric values

in the cells are coefficients of the regressors in the models, which correspond to the short-run elasticity. In testing Hypothesis 1, which states that FDI leads to economic growth, Table 5 shows that FDI has a positive direct effect on economic growth and is statistically significant at the 0.01 level. When we run a panel regression in a bivariate setting, examining the direct effects of FDI on economic growth, we find a positive relationship between FDI and economic growth for the G20 (the coefficient of FDI¼0.390). It implies that as the volume of FDI increases, so does economic growth. When we run a regression in a multivariate setting, we still find a positive relationship between the two variables. The results of the two regression models suggest that the association between FDI net inflows and economic growth is robust. It suggests that a 1% increase in FDI net inflows increases economic growth by 0.201% for the G20.

In testing Hypothesis 2, which states that economic growth is in positive relation to CO2 emissions, Table 5 shows that economic growth has a negative impact on CO2 emissions in the G20 and is statistically significant at the 0.01 level. When we run a panel regression in a bivariate setting, examining the direct effects of economic growth on CO2 emissions, we find a positive relationship between economic growth and CO2 emissions (the coefficient of economic growth¼0.379). It implies that as economic growth increases, so does CO2 emission. However, when we run a regression in a multivariate setting, we find a negative relationship between the two variables. The conflict results between the two regression models indicate that the association between economic growth and CO2 emissions may be spurious. If we had failed to accompany other variables, we would have mistakenly found that economic growth has a positive relation to CO2 emissions. Instead, we find a 1% increase in economic growth is accompanied by 0.053% reduction of CO2 emissions in the G20. The results of the Hypothesis 2 test conflict with the findings of many of previous studies.

In testing Hypothesis 3, which states that FDI is in positive relation to CO2 emissions, Table 5 shows that FDI exhibits in no relation to CO2 emissions. When we run a panel regression in a bivariate setting, examining the direct effects of FDI on CO2 emissions, we find a positive relationship between FDI and CO2 emissions (the coefficient of FDI¼0.128). It implies that as the volume of FDI increases, so does CO2 emission. However, when we run a regression in a multivariate setting, we find no relation- ship between the two variables. The conflict results between the two regression models suggest that the association between FDI and CO2 emissions may be spurious. If we had failed to accom- pany other variables, we would have mistakenly found that FDI has a positive relation to CO2 emissions. In other words, as FDI

J.W. Lee / Energy Policy 55 (2013) 483–489488

increases, so does economic growth and energy use, resulting in a positive correlation with CO2 emissions. Instead, we find that an increase in FDI net inflows is not necessarily accompanied by an increase in CO2 emissions for the G20. This further implies that FDI may have played a lot of externalities in economic growth and extended spillover effects by improving energy efficiency and promoting clean energy development, and thus resulted in redu- cing CO2 emissions.

In testing Hypothesis 4, which states that FDI is in positive to clean energy use, Table 5 shows that FDI exhibits in no relation to clean energy use. When we run a panel regression in a bivariate setting, examining the direct effects of FDI on clean energy use, we find a positive relationship between FDI and clean energy use for the G20 (the coefficient of FDI¼0.198). It implies that as the volume of FDI increases, so does clean energy use. However, when we run a regression in a multivariate setting, we find no relationship between the two variables. The conflict results between the two regression models suggest that the association between FDI and clean energy use may be spurious. If we had failed to accompany other variables, we would have mistakenly found that FDI has a positive relation to clean energy use. In other words, as FDI increases, so does economic growth and clean energy use. Instead, we find that an increase in FDI net inflows is not necessarily accompanied by an increase in clean energy use in the G20. This further implies that, although FDI may have played a lot of externalities in economic growth, it is difficult to find compelling evidence of spillover effects that promote clean energy use.

5. Discussion and policy implications

The results of this study suggest that FDI strongly leads to economic growth while it is in no direct positive relation to an increase in CO2 emissions for the G20. Although it is generally agreed that FDI has some influence on carbon emissions, the results of this research indicate that FDI in the G20 may have been utilized in a relatively innovative way, which is not in line with the linearly increasing trend of carbon emissions. The findings indicate that FDI plays a critical role in boosting economic growth while it marginally contributes to an increase in CO2 emissions for the G20. As a consequence, FDI has been increasingly used as an important input in their economies.

In addition to their direct contribution to output growth, FDI may have generated spillovers or augmented benefits that might exceed the direct returns to economic growth. The technological progress accompanied by FDI might have led to a rapid improvement in the efficient use of energy resources and thus resulted in a reduction of CO2 emissions. For example, there has been significant technological advancement and innovation in energy use and the production of energy carriers and thus have led to the production of environment friendly goods and services in the G20. FDI lends itself to new models of green growth and innovation leading to new technologies for increasing energy efficiency and developing clean energy sources. FDI enhances countries’ ability to tackle environmental problems and leads to the creation of green products and technol- ogies that benefit company, society, and government. In their efforts to the green use of FDI coupled with progressive policy making, FDI will play a critical role in making a commitment to greenhouse gas emission reduction for the G20.

The results of this study find that clean energy use strongly leads to economic growth while it is in negative relation to an increase in CO2 emissions. The finding implies that clean energy use has played a critical role in boosting economic growth while it has reduced a large portion of CO2 emissions. The finding also implies that clean energy use may have been accentuated because technological advancement accompanied by FDI may have led to a rapid improvement in the use

of clean energy and the development of clean energy resources, and thus resulted in reducing CO2 emissions.

For example, in contrast to most of developed countries where growth in the nuclear power generating capacity leveled out or held down for many years, a number of countries in the G20, especially China, India, Indonesia, South Korea, and Turkey are planning and building new nuclear power plants to meet their increasing demand for energy. The nuclear share of these countries to 2020 is expected to be considerable. In addition, most of developed countries in the G20 have already established an energy system for non-fossil energy sources and the share of renewable energy such as hydraulic, biomass, wind, and solar will account for a significant amount of the total energy use. In doing so, the mandatory clean energy use targets, which are to supply 20% of the energy demand with renew- able energy by 2020 in the EU and Australia and up to 15% of the total energy demand with non-fossil energy sources by 2020 in a number of developing countries in the G20 like China, must be realized.

In addition to its direct contribution to greenhouse gas emission reduction, clean energy use may have generated spillovers or augmented benefits that might exceed the direct returns to envir- onmental sustainability. Given the evidence, in their role as an accelerator of green growth, clean energy use has contributed increasingly to output growth and at the same time to environ- mental sustainability. Therefore, investments in increasing energy efficiency must be accentuated, and new low carbon technologies should be introduced, thus lowering the overall costs of the deployment of many energy efficient technologies. In their efforts, technology support policies need to evolve as technology matures from a research stage to full commercialization. In addition, govern- ments need to help remove any stubborn barriers and improve rules and regulations to promulgate at all levels of government. In summary, there is a growing role for FDI and clean energy use in the pursuit of green growth goals. Governments must design strategies that will allow their nation to move from their current growth process onto the sustainable growth path. This will require policy changes in many countries, with respect to their own growth and to their impacts on other nations’ growth possibilities.

6. Conclusions

This paper integrates FDI and economic growth, CO2 emissions, and clean energy use in a multivariate format for cointegration tests with the panel data of 19 nations of the G20. The current research discovers and enhances an understanding of a long-run equilibrium relationship among these variables. More importantly, as an explora- tory study, the findings indicate that FDI directly affects economic growth and exhibits no direct effect on CO2 emissions in the G20. The applicability of the findings implies that FDI plays a critical role in the continuing economic growth for the region and in achieving emission reductions through policy and practice changes. Increasing energy efficiency, developing renewable energy resources, and introducing new technologies for low carbon energy will require widespread deployment. There is evidence that when policy makers make strong efforts to attract FDI through policy campaigns, both economy and environment benefit.

7. Suggestions for future research

The relationship between FDI and CO2 emissions can be different for a developing versus a developed country. In addition, the relation- ship between FDI and CO2 emissions can be different for the energy structure of non-renewable energy sources versus renewable energy sources. Setting different hypotheses for the energy structure to compare results across developing versus developed economies

J.W. Lee / Energy Policy 55 (2013) 483–489 489

may be of interest to see if there is a difference. A final suggestion is in order regarding the prospect of global CO2 emissions underlines the need to transform carbon-intensive economies rapidly into green economies. In order to implement such an approach, further empiri- cal analyses are needed to estimate the future associated emissions and the future mitigation potential of growth. This provides an initial guide for ranking sustainable growth actions. Policy makers can then weigh the emissions reduction potential against other sustainability aspects of the action in choosing the appropriate policy to implement.

References

Aitken, B., Hanson, G.H., Harrison, A.E., 1997. Spillovers, foreign investment, and export behavior. Journal of International Economics 43, 103–132.

Akinlo, A.E., 2008. Energy consumption and economic growth: evidence from 11 African countries. Energy Economics 30, 2391–2400.

Alfaro, L., Chanda, A., Kalemli-Ozcan, S., Sayek, S., 2010. Does foreign direct investment promote growth? Exploring the role of financial markets on linkages. Journal of Growth Economics 91, 242–256.

Al-Iriani, M., 2006. Energy–GDP relationship revisited: an example from GCC countries using panel causality. Energy Policy 34, 3342–3350.

Ang, J.B., 2007. CO2 emissions, energy consumption, and output in France. Energy Policy 35, 4772–4778.

Apergis, N., Payne, J.E., 2009. Energy consumption and economic growth in Central America: evidence from a panel cointegration and error correction model. Energy Economics 31, 211–216.

Apergis, N., Payne, J.E., 2010. The emissions, energy consumption, and growth nexus: evidence from the commonwealth of independent states. Energy Policy 38, 650–655.

Batten, J.A., Vo, X.V., 2009. An analysis of the relationship between foreign direct investment and economic growth. Applied Economics 41, 1621–1641.

Blomstrom, M., Lipsey, R.E., Zejan, M., 1996. Is fixed investment the key to economic growth quest? Quarterly Journal of Economics 111, 269–276.

Breitung, J., 2000. The local power of some unit root tests for panel data. In: Baltagi, B.H. (Ed.), Advances in Econometrics: Nonstationary Panels, Panel Cointegration and Dynamic Panels. Elsevier, Amsterdam, The Netherlands, pp. 161–178.

Chen, S.-T., Kuo, H.-I., Chen, C.-C., 2007. The relationship between GDP and electricity consumption in 10 Asian countries. Energy Policy 35, 2611–2621.

Chontanawat, J., Hunt, L.C., Pierce, R., 2008. Does energy consumption cause economic growth: evidence from systematic study of over 100 countries. Journal of Policy Modeling 30, 209–220.

Ciarreta, A., Zarraga, A., 2010. Economic growth-electricity consumption causality in 12 European countries: a dynamic panel data approach. Energy Policy 38, 3790–3796.

Coondoo, D., Dinda, S., 2002. Causality between income and emission: a country group specific econometric analysis. Ecological Economics 40, 351–367.

Costantini, V., Martini, C., 2009. The previous term causality next term between energy consumption and economic growth: a multi-sectoral analysis using non-stationary cointegrated panel data. Energy Economics 32, 591–603.

De Long, J.B., Summers, L.H., Abel, A.B., 1992. Equipment investment and economic growth: how strong is the nexus quest. Brookings Papers on Economic Activity 2, 157–211.

Dinda, S., Coondoo, D., 2006. Income and emission: a panel data-based cointegra- tion analysis. Ecological Economics 57, 167–181.

Ekanayake, E.M., Vogel, R., 2003. Openness and economic growth: empirical evidence on the relationship between output, FDI, and trade. Journal of Business Strategies 20 (1), 59–72.

Engle, R.F., Granger, C.W.J., 1987. Cointegration and error correction representa- tion, estimation and testing. Econometrica: Journal of the Econometric Society 55, 251–276.

Fernandes, A.M., Paunov, C., 2012. Foreign direct investment in services and manufacturing productivity: evidence for Chile. Journal of Growth Economics 97, 305–321.

Ghosh, S., 2010. Examining carbon emissions economic growth nexus for India: a multivariate cointegration approach. Energy Policy 38, 3008–3014.

Granger, C.W.J., 1988. Some recent growths in a concept of causality. Journal of Econometrics 39, 199–211.

Hermes, N., Lensink, R., 2003. Foreign direct investment, financial growth and economic growth. Journal of Growth Studies 40, 142–163.

Hsiao, C., Shen, Y., 2003. Foreign direct investment and economic growth: the importance of institutions and urbanization. Economic Growth and Cultural Change 51, 883–896.

Huang, B.-N., Hwang, M.-J., Yang, C.-W., 2008. Causal relationship between energy consumption and GDP growth revisited: a dynamic panel data approach. Ecological Economics 67, 41–54.

Im, K.S., Pesaran, M.H., Shin, Y., 2003. Testing for unit roots in heterogeneous panels. Journal of Econometrics 115, 53–74.

International Energy Agency, 2007. World Energy Outlook. Paris, France. International Energy Agency, 2011. CO2 Emissions from Fuel Combustion. Paris,

France.

Johansen, S., 1991. Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econometrica: Journal of the Econo- metric Society 59, 1551–1580.

Keppler, J.H., Mansanet-Bataller, M., 2010. Causalities between CO2, electricity, and other energy variables during phase I and phase II of the EU ETS. Energy Policy 38, 3329–3341.

Kim, D.-D., Seo, J.-S., 2003. Does FDI inflow crowd out DI in Korea quest. Journal of Economic Studies 30, 605–622.

Kneller, R., Pisu, M., 2007. Industrial linkages and export spillovers from FDI. The World Economy 30, 105–134.

Lee, C.-C., 2005. Energy consumption and GDP in developing countries: a coin- tegrated panel analysis. Energy Economics 27, 415–427.

Lee, C.-C., Chang, C.-P., 2008. Energy consumption and economic growth in Asian economies: a more comprehensive analysis using panel data. Resources and Energy Economics 30, 50–65.

Lee, C.-C., Chang, C.-P., 2009. FDI, financial growth, and economic growth: international evidence. Journal of Applied Economics 12, 249–271.

Lee, C.-C., Lee, J.-D., 2009. Income and CO2 emissions: evidence from panel unit root and cointegration tests. Energy Policy 37, 413–423.

Lee, C.-C., Chang, C.-P., Chen, P.-F., 2008. Energy-income causality in OECD countries revisited: the key role of capital stock. Energy Economics 30, 2359–2373.

Levin, A., Lin, C.F., Chu, C.J., 2002. Unit root tests in panel data: asymptotic and finite-sample properties. Journal of Econometrics 108, 1–24.

List, J.A., Co, C.Y., 2000. The effect of environmental regulation on foreign direct investment. Journal of Environmental Economics and Management 40, 1–20.

MacKinnon, J.G., 1996. Numerical distribution functions for unit root and coin- tegration tests. Journal of Applied Econometrics 11, 601–618.

Mackinnon, J.G., Haug, A.A., Michelis, L., 1999. Numerical distribution functions of likelihood ratio tests for co-integration. Journal of Applied Econometrics 14, 563–577.

Maddala, G.S., Wu, S., 1999. A comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and Statistics 61, 631–652.

Mahadevan, R., Asafu-Adjaye, J., 2007. Energy consumption, economic growth and prices: a reassessment using panel VECM for developed and developing countries. Energy Policy 35, 2481–2490.

Mielnik, O., Goldemberg, J., 2002. Foreign direct investment and decoupling between energy and gross domestic product in developing countries. Energy Policy 30, 87–89.

Narayan, P.K., Narayan, S., 2010. Carbon dioxide emissions and economic growth: panel data evidence from developing countries. Energy Policy 38, 661–666.

Narayan, P.K., Smyth, R., 2008. Energy consumption and GDP in G7 countries: new evidence from panel cointegration with structural breaks. Energy Economics 30, 2331–2341.

Narayan, P.K., Smyth, R., 2009. Multivariate granger causality between electricity consumption, exports and GDP: evidence from a panel of Middle Eastern countries. Energy Policy 37, 229–236.

Niu, S., Ding, Y., Niu, Y., Li, Y., Luo, G., 2011. Economic growth, energy conservation and emissions reduction: a comparative analysis based on panel data for 8 Asian-Pacific countries. Energy Policy 39, 2121–2131.

Pao, H.-T., Tsai, C.-M., 2010. CO2 emissions, energy consumption and economic growth in BRIC countries. Energy Policy 38, 7850–7860.

Qi, L., 2007. The relationship between growth, total investment and inward FDI: Evidence from time series data. International Review of Applied Economics 21, 119–133.

Sadorsky, P., 2009. Renewable energy consumption and income in developing economies. Energy Policy 37, 4021–4028.

Sadorsky, P., 2010. The impact of financial growth on energy consumption in developing economies. Energy Policy 38, 2528–2535.

Sari, R., Soytas, U., 2007. The growth of income and energy consumption in six developing countries. Energy Policy 35, 889–898.

Sharma, S.S., 2010. The relationship between energy and economic growth: empirical evidence from 66 countries. Applied Energy 87, 3565–3574.

Soytas, U., Sari, R., 2009. Energy consumption, economic growth, and carbon emissions: challenges faced by an EU candidate member. Ecological Economics 68, 1667–1675.

Squalli, J., 2007. Electricity consumption and economic growth: bounds and causality analyses for OPEC members. Energy Economics 29, 1192–1205.

Tamazian, A., Chousa, J.P., Vadlamannati, C., 2009. Does higher economic and financial growth lead to environmental degradation: evidence from the BRIC countries. Energy Policy 37, 246–253.

Tsang, E.W.K., Yip, P.S.L., 2007. Economic distance and the survival of foreign direct investments. Academy of Management Journal 50, 1156–1168.

Wolde-Rufael, Y., 2009. Energy consumption and economic growth: the experi- ence of African countries revisited. Energy Economics 31, 217–224.

Xing, Y., Kolstad, C.D., 2002. Do lax environmental regulations attract foreign investment? Environmental and Resource Economics 21, 1–22.

Yoo, S.-H., Kwak, S.-Y., 2010. Electricity consumption and economic growth in seven South American countries. Energy Policy 38, 181–188.

Zhang, K.-H., 2001. Does foreign direct investment promote economic growth quest: evidence from East Asia and Latin America. Contemporary Economic Policy 19, 175–185.

Zhang, X.-P., Cheng, X.-M., 2009. Energy consumption, carbon emissions, and economic growth in China. Ecological Economics 68, 2706–2712.

Zhang, Y.-J., 2011. The impact of financial growth on carbon emissions: an empirical analysis in China. Energy Policy 39, 2197–2203.