#### DMCA

## Energy Consumption and Economic Growth: New Insights into The Cointegration Relationship (2011)

Venue: | Energy Economics |

Citations: | 16 - 1 self |

### Citations

4053 |
Some tests of Specification for panel data: Monte Carlo evidence and an application to employment equations
- Arellano, Bond
- 1991
(Show Context)
Citation Context ... differenced form takes care of the OLS estimation problem, which is due to correlation between country-specific effects and explanatory variables. But differencing introduces the problem of simultaneity because the lagged dependent variables are correlated with the differenced error term. Additionally, heteroscedasticity in the errors across the cross-section members is expected to occur. We have to apply an instrumental variable estimator to cope with these problems. A widely used estimator for the system in Eqs. (5)–(7) is the panel generalised method of moments (GMM) estimator proposed by Arellano and Bond (1991). Predetermined lags of the system variables are used as instruments to obtain consistent results. According to our empirical investigations, a lag length of k=2 proves to be necessary to remove serial correlation in the error term. Hence, we employ variables lagged three and four periods as instruments for the lagged dependent variables. 787A. Belke et al. / Energy Economics 33 (2011) 782–789The direction of causality can be determined by testing for the significance of the coefficients of each dependent variable in Eqs. (5) to (7). First, to check for short-run causality we test H0:θ12ik=0, ... |

2673 |
Cointegration and error correction: representation, estimation and testing
- Engle, Granger
- 1987
(Show Context)
Citation Context ...R)model in the tradition of Sims (1972). For example, the seminal work of Kraft and Kraft (1978), using a VARmodel, found evidence in favour of causality running from income to energy consumption in the United States for the period 1947–1974. Further, studies of the first generation examined the direction of causality assuming stationarity of the underlying variables. By contrast, second generation studies accounted for non-stationarity in the data and performed cointegration analysis to investigate the long-run relationship between energy consumption and growth. This literature, based on the Engle and Granger (1987) two-step procedure, studied pairs of variables to check for cointegration relationships and used estimated errorcorrection models to test for Granger causality. Third generation studies used multivariate estimators in the style of Johansen (1991). Johansen's multivariate approach also allows for more than two variables in the cointegration relationship. Finally, fourth generation studies employ recently developed panel-econometric methods to test for unit roots and cointegration relations. This literature estimates panel-based error-correction models to perform Granger causality tests.2 Accor... |

1659 |
Estimation and hypothesis testing of cointegrating vectors in Gaussian vector autoregressive models
- Johansen
- 1991
(Show Context)
Citation Context ...ies of the first generation examined the direction of causality assuming stationarity of the underlying variables. By contrast, second generation studies accounted for non-stationarity in the data and performed cointegration analysis to investigate the long-run relationship between energy consumption and growth. This literature, based on the Engle and Granger (1987) two-step procedure, studied pairs of variables to check for cointegration relationships and used estimated errorcorrection models to test for Granger causality. Third generation studies used multivariate estimators in the style of Johansen (1991). Johansen's multivariate approach also allows for more than two variables in the cointegration relationship. Finally, fourth generation studies employ recently developed panel-econometric methods to test for unit roots and cointegration relations. This literature estimates panel-based error-correction models to perform Granger causality tests.2 According to our analysis of 25 OECD countries Table 1 summarises preferably all studies of the last five years on developed countries and their empirical results.3 The ambiguous evidence of the empirical literature on the causal relationship between e... |

1313 |
Likelihood-Based Inference in Cointegrated Vector Autoregressive Models
- Johansen
- 1995
(Show Context)
Citation Context ...s are lag levels based on the Schwarz Information Criterion. Numbers in brackets represents the automatic NeweyWest bandwidth selection using the Bartlett kernel. We include a constant and a linear 785A. Belke et al. / Energy Economics 33 (2011) 782–7893.3. Cointegration analysis As integration of order one is established for the common components of the variables under investigation, the next step is to determine whether a long-run relationship exists.9 Cointegration between the common components can be investigated using standard time series tests such as the Johansen reduced rank approach (Johansen, 1995). As aforementioned, small sample size can induce biased Johansen test statistics. Hence, we apply the small sample modification proposed by Reinsel and Ahn (1992) and Reimers (1992), who suggest the multiplication of the Johansen statistics with the scale factor (T−pk)/T, where T is the number of observations, p the number of variables and k the lag order of the VAR. This approach corrects for small sample bias such that a proper inference can bemade. The empirical realisations of themodified Johansen trace statistic as well as those of the modified Johansen maximum eigenvalue statistic sugge... |

1287 |
Testing the null hypothesis of stationarity against the alternative of a unit root
- Kwiatkowski, Phillips, et al.
- 1992
(Show Context)
Citation Context ...n-stationarity in the panel will be driven entirely by a reduced number of international stochastic trends. In that case, cointegration between the series requires that the common factors of the variables cointegrate (Dreger and Reimers, 2009). 3.2. Unit root tests In the analysis of the common components of energy consumption per capita, real GDP per capita and energy prices, standard time series unit root tests can be applied. To ensure robustness we use several unit root tests, including the augmentedDickey and Fuller (1979) (ADF) test, the Phillips andPerron (1988) (PP) test, aswell as theKwiatkowski et al. (1992) (KPPS) test. The latter tests the null of stationarity whereas the former two investigate the null of a unit root.We do not further discuss the details of these well-known time series unit root tests but instead call attention to Maddala and Kim (1998) for their excellent treatment of ADF, PP and KPSS.7 According to our results, displayed in Table 2, the common components of energy consumption per capita, real GDP per capita and energy prices all turn out to be integrated of order one, I(1). Since the defactored series are independent by construction, stochastic trends in the idiosyncratic co... |

1255 |
Testing for unit roots in heterogeneous panels
- Im, Pesaran, et al.
- 2003
(Show Context)
Citation Context ...all attention to Maddala and Kim (1998) for their excellent treatment of ADF, PP and KPSS.7 According to our results, displayed in Table 2, the common components of energy consumption per capita, real GDP per capita and energy prices all turn out to be integrated of order one, I(1). Since the defactored series are independent by construction, stochastic trends in the idiosyncratic components are efficiently explored by first generation panel unit root tests to exploit the additional information provided by the cross-sectional data. We apply the tests suggested by Levin et al. (2002) (LLC) and Im et al. (2003) (IPS). The LLC test restrictively assumes that all cross-sections have the same first order autoregressive parameter. By contrast, the IPS test relaxes this assumption by allowing heterogeneity in this coefficient for all cross-section units.8 Although the IPS tests is preferable according to that, this study also reports the results of the LLC test to provide an additional check for robustness. In contrast to the time series unit root evidence for the common components, the LLC and IPS panel unit root tests propose that the idiosyncratic components are widely stationary (see Table 3). Hence,... |

1137 | Testing for a unit root in time series regression - B, Perron - 1988 |

939 |
Unit root tests in panel data: Asymptotic and finite-sample properties
- Levin, Lin, et al.
- 2002
(Show Context)
Citation Context ... unit root tests but instead call attention to Maddala and Kim (1998) for their excellent treatment of ADF, PP and KPSS.7 According to our results, displayed in Table 2, the common components of energy consumption per capita, real GDP per capita and energy prices all turn out to be integrated of order one, I(1). Since the defactored series are independent by construction, stochastic trends in the idiosyncratic components are efficiently explored by first generation panel unit root tests to exploit the additional information provided by the cross-sectional data. We apply the tests suggested by Levin et al. (2002) (LLC) and Im et al. (2003) (IPS). The LLC test restrictively assumes that all cross-sections have the same first order autoregressive parameter. By contrast, the IPS test relaxes this assumption by allowing heterogeneity in this coefficient for all cross-section units.8 Although the IPS tests is preferable according to that, this study also reports the results of the LLC test to provide an additional check for robustness. In contrast to the time series unit root evidence for the common components, the LLC and IPS panel unit root tests propose that the idiosyncratic components are widely stati... |

606 |
Estimating vector autoregressions with panel data
- Holtz-Eakin, Newey, et al.
- 1988
(Show Context)
Citation Context ...han previous oil crises which we exclude from our sample. 3.4. Dynamic panel causality Having established a cointegration relationship, we estimate a panel-based error-correctionmodel to test for Granger causality among energy consumption per capita, real GDPper capita and energy prices. A two-step procedure is applied. First, the long-run equations specified in Eq. (4) are used to obtain the deviations from the long-run equilibrium (υ,ε and η). Then the error-correctionmodel is estimated with the oneperiod lagged residuals from the first step as dynamic error-correction terms, as proposed in Holtz-Eakin et al. (1988): ΔEi;t = α1i + ∑ h k=1 θ11i;kΔEi;t−k + ∑ h k=1 θ12i;kΔYi;t−k + ∑ h k=1 θ13i;kΔPi;t−k + λ1iυi;t−1 + u1i;t ð5Þ ΔYi;t = α2i + ∑ h k=1 θ21i;kΔEi;t−k + ∑ h k=1 θ22i;kΔYi;t−k + ∑ h k=1 θ23i;kΔPi;t−k + λ2iεi;t−1 + u2i;t ð6Þ ΔPi;t = α3i + ∑ h k=1 θ31i;kΔEi;t−k + ∑ h k=1 θ32ikΔYi;t−k + ∑ h k=1 θ33i;kΔPi;t−k + λ3iηi;t−1 + u3i;t ; ð7Þ where Δ is the first-difference operator, k is the lag length, λi is the speed of adjustment and ui, t is the serially uncorrelated error term withmean zero. The differenced form takes care of the OLS estimation problem, which is due to correlation between country-specific... |

336 |
Pitfalls and Opportunities: What Macroeconomists Should Know about Unit Roots,”
- Campbell, Perron
- 1991
(Show Context)
Citation Context ...ita in constant 2000 U.S. dollars using purchasing power parities (PPPs) are used as a proxy of economic growth (Y).5 Furthermore, time series data for the final energy consumption in kilograms of oil equivalent per capita (E) and for the energy price index (P) in U.S. dollars (PPP) have been collected. All variables are in natural logarithms and have been obtained from the International Energy Agency's (IEA) online database.6 It is widely known that standard unit root and cointegration tests based on individual time series have low statistical power, especially when the time series is short (Campbell and Perron, 1991). Panelbased tests represent an improvement in this respect by exploiting additional information that results from the inclusion of the crosssectional dimension. However, first generation panel unit root and cointegration tests often assume that the cross-section members are independent. This condition is often likely to be violated, for example, because of common oil price shocks. Inappropriately assuming crosssectional independence can distort the panel results (see Banerjee et al. (2004), Urbain and Westerlund (2006)). Therefore, our study controls for cross-section dependencies by taking i... |

303 |
Money, income and causality
- Sims
- 1972
(Show Context)
Citation Context ...lusions and policy implications.2. Literature review The empirical literature provides mixed and conflicting evidence with respect to the energy consumption-growth nexus. This discrepancy in results is due largely to the use of different econometric methods and time periods, besides country-specific heterogeneity in climate conditions, economic development and energy consumption patterns, among other things. From a methodological perspective, four generations of contributions can be identified. First generation studies applied a traditional vector autoregression (VAR)model in the tradition of Sims (1972). For example, the seminal work of Kraft and Kraft (1978), using a VARmodel, found evidence in favour of causality running from income to energy consumption in the United States for the period 1947–1974. Further, studies of the first generation examined the direction of causality assuming stationarity of the underlying variables. By contrast, second generation studies accounted for non-stationarity in the data and performed cointegration analysis to investigate the long-run relationship between energy consumption and growth. This literature, based on the Engle and Granger (1987) two-step proce... |

299 | Numerical Distribution Functions of Likelihood Ratio Tests for Cointegration.
- MacKinnon, Haug, et al.
- 1999
(Show Context)
Citation Context ...k for possible parameterTable 4 Results of Johansen's tests for cointegration among common components. H0 Trace Statistic Critical Value Max. Eigenvalue Statistic Critical Value None 45.78* 42.92 26.93* 25.82 At most 1 19.85 25.87 13.93 19.39 At most 2 5.93 12.52 5.93 12.52 Notes: Potential small sample bias is corrected by multiplying the Johansen statistics with the scale factor (T−pk)/T, where T is the number of observations, p is the number of variables and k is the lag order of the underlying VARmodel in levels, see Reinsel and Ahn (1992) and Reimers (1992). Critical values are taken fromMacKinnon et al. (1999), and are also valid for the small sample correction. * indicates the rejection of the null hypothesis of no cointegration at least at the 5% level of significance.instability by applying Wald F-tests to test whether the coefficients are different. The DOLS estimates for the two sets of subsamples with 1990 and 2000 as break dates show that there are no significant variations in the long-run elasticities. On the basis of these results we infer that there are no substantial structural breaks in the long-run relationships between energy consumption per capita, real GDP per capita and energy pric... |

151 |
Comparisons of tests for multivariate cointegration.
- Reimers
- 1992
(Show Context)
Citation Context ...t and a linear 785A. Belke et al. / Energy Economics 33 (2011) 782–7893.3. Cointegration analysis As integration of order one is established for the common components of the variables under investigation, the next step is to determine whether a long-run relationship exists.9 Cointegration between the common components can be investigated using standard time series tests such as the Johansen reduced rank approach (Johansen, 1995). As aforementioned, small sample size can induce biased Johansen test statistics. Hence, we apply the small sample modification proposed by Reinsel and Ahn (1992) and Reimers (1992), who suggest the multiplication of the Johansen statistics with the scale factor (T−pk)/T, where T is the number of observations, p the number of variables and k the lag order of the VAR. This approach corrects for small sample bias such that a proper inference can bemade. The empirical realisations of themodified Johansen trace statistic as well as those of the modified Johansen maximum eigenvalue statistic suggest evidence in favour of a long-run relationship between the common factors of energy consumption per capita, real GDP per capita and energy prices (see Table 4). As a next step, we ... |

142 | A panic attack on unit roots and cointegration.
- Bai, Ng
- 2004
(Show Context)
Citation Context ...ountries (annual), while data for energy prices are drawn from IEA statistics on energy prices and taxes (quarterly). Additionally, we use data for population of the IMF's International Financial Statistics to obtain energy consumption in terms of per capita.geographically concentrated effects. The common component of a variable is ‘common’ in the sense that it depends on a small number of common shocks, which affect the respective variable of all the countries. The decomposition is based on differenced data because of potential non-stationarity of the levels of the variables, as suggested by Bai and Ng (2004). After estimating the common factors they are recumulated to match the integration properties of the original variables. We obtain the idiosyncratic components from a regression of the original series on their common factors. For all three variables we use two common components, which is enough to capture more than 50% of the overall variance. Any further component would add only a small proportion and the evidence shows that results do not qualitatively change. As a second step, the common factors and idiosyncratic components are tested separately for unit roots and cointegration relationshi... |

135 |
Applied Time Series Modelling and Forecasting.
- Hharris, Sollis
- 2003
(Show Context)
Citation Context ...ding to that, this study also reports the results of the LLC test to provide an additional check for robustness. In contrast to the time series unit root evidence for the common components, the LLC and IPS panel unit root tests propose that the idiosyncratic components are widely stationary (see Table 3). Hence, the results suggest that random walks in the data are driven mainly by international developments. As a consequence, cointegration i.e. a long-run relationship may exist between the common rather than the idiosyncratic components.7 Further details of the ADF tests can also be found in Harris and Sollis (2003). 8 For a more detailed description of the LLC and IPS test we recommend Baltagi (2008). Table 1 Overview of studies on developed countries of the last few years. Study Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~... |

108 |
On the relationship between energy and GNP.
- Kraft, Kraft
- 1978
(Show Context)
Citation Context ...review The empirical literature provides mixed and conflicting evidence with respect to the energy consumption-growth nexus. This discrepancy in results is due largely to the use of different econometric methods and time periods, besides country-specific heterogeneity in climate conditions, economic development and energy consumption patterns, among other things. From a methodological perspective, four generations of contributions can be identified. First generation studies applied a traditional vector autoregression (VAR)model in the tradition of Sims (1972). For example, the seminal work of Kraft and Kraft (1978), using a VARmodel, found evidence in favour of causality running from income to energy consumption in the United States for the period 1947–1974. Further, studies of the first generation examined the direction of causality assuming stationarity of the underlying variables. By contrast, second generation studies accounted for non-stationarity in the data and performed cointegration analysis to investigate the long-run relationship between energy consumption and growth. This literature, based on the Engle and Granger (1987) two-step procedure, studied pairs of variables to check for cointegrati... |

96 |
Unit Roots, Cointegration and Structural Change
- Maddala, Kim
- 1998
(Show Context)
Citation Context ...oot tests In the analysis of the common components of energy consumption per capita, real GDP per capita and energy prices, standard time series unit root tests can be applied. To ensure robustness we use several unit root tests, including the augmentedDickey and Fuller (1979) (ADF) test, the Phillips andPerron (1988) (PP) test, aswell as theKwiatkowski et al. (1992) (KPPS) test. The latter tests the null of stationarity whereas the former two investigate the null of a unit root.We do not further discuss the details of these well-known time series unit root tests but instead call attention to Maddala and Kim (1998) for their excellent treatment of ADF, PP and KPSS.7 According to our results, displayed in Table 2, the common components of energy consumption per capita, real GDP per capita and energy prices all turn out to be integrated of order one, I(1). Since the defactored series are independent by construction, stochastic trends in the idiosyncratic components are efficiently explored by first generation panel unit root tests to exploit the additional information provided by the cross-sectional data. We apply the tests suggested by Levin et al. (2002) (LLC) and Im et al. (2003) (IPS). The LLC test re... |

92 | Cointegration vector estimation by panel DOLS and long-run money demand
- Mark, Sul
- 2003
(Show Context)
Citation Context ...the number of observations, p the number of variables and k the lag order of the VAR. This approach corrects for small sample bias such that a proper inference can bemade. The empirical realisations of themodified Johansen trace statistic as well as those of the modified Johansen maximum eigenvalue statistic suggest evidence in favour of a long-run relationship between the common factors of energy consumption per capita, real GDP per capita and energy prices (see Table 4). As a next step, we estimate the long-run relationships using the dynamic ordinary least squares (DOLS) estimatorproposedbyMark and Sul (2003). The DOLS estimator corrects standard OLS for bias induced by endogeneity and serial correlation. First, the endogenous variable in each equation is regressed on the leads and lags of the first-differenced regressors from all equations to control for potential endogeneities. Then the OLS method is applied using the residuals from the first step regression. The DOLS estimator is preferred to the non-parametric9 Since the panel unit root tests of the idiosyncratic components suggest stationarity, we do not test for cointegration between the idiosyncratic components of energy consumption, GDP an... |

88 |
Vector Autoregressive Models with Unit Roots and Reduced Rank Structure: Estimation, Likelihood Ratio Test, and Forecasting
- Reinsel, Ahn
- 1992
(Show Context)
Citation Context ...ernel. We include a constant and a linear 785A. Belke et al. / Energy Economics 33 (2011) 782–7893.3. Cointegration analysis As integration of order one is established for the common components of the variables under investigation, the next step is to determine whether a long-run relationship exists.9 Cointegration between the common components can be investigated using standard time series tests such as the Johansen reduced rank approach (Johansen, 1995). As aforementioned, small sample size can induce biased Johansen test statistics. Hence, we apply the small sample modification proposed by Reinsel and Ahn (1992) and Reimers (1992), who suggest the multiplication of the Johansen statistics with the scale factor (T−pk)/T, where T is the number of observations, p the number of variables and k the lag order of the VAR. This approach corrects for small sample bias such that a proper inference can bemade. The empirical realisations of themodified Johansen trace statistic as well as those of the modified Johansen maximum eigenvalue statistic suggest evidence in favour of a long-run relationship between the common factors of energy consumption per capita, real GDP per capita and energy prices (see Table 4). ... |

85 |
The Relationship between Energy Consumption, Energy Prices and Economic Growth: Time Series Evidence from Asian Developing Countries”,
- Asafu-Adjaye
- 2000
(Show Context)
Citation Context ...e economic growth without a considerable increase in energy consumption. Therefore, the empirical results of causality analyses should be interpreted with caution. Our analysis of the relationship between energy consumption and economic activity is based on a sample of 25 OECD countries from 1981 to 2007 and uses recently developed panel-econometric methods. We explore an additional channel of causality by introducing energy prices. As energy prices have been neglected in many previous studies, the long-run parameters and the evidence of causality may be biased, see Masih and Masih (1997) and Asafu-Adjaye (2000). But in contrast to these two studies, we examine the original energy price index rather than the consumer price index (CPI) as a proxy. Income and price elasticities provide policy makers a suggestion of the extent to which prices need to increase, in the form of energy taxes, in order to reduce energy consumption and the potential for the market to conserve energy (Lee and Lee, 2010). Additionally, energy companies need this information to design their demand management policies. But only a few papers have estimated income and price elasticities for energy consumption in a panel framework. ... |

80 |
The causality relationship between energy consumption and GDP
- Lee
- 2006
(Show Context)
Citation Context ...on i.e. a long-run relationship may exist between the common rather than the idiosyncratic components.7 Further details of the ADF tests can also be found in Harris and Sollis (2003). 8 For a more detailed description of the LLC and IPS test we recommend Baltagi (2008). Table 1 Overview of studies on developed countries of the last few years. Study Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OE... |

56 |
A literature survey on energy–growth nexus.
- Ozturk
- 2010
(Show Context)
Citation Context ...ic growth or vice versa. Hence, energy conservation policies would not have any impact on real GDP. In consideration of such a pure statistical causality analysis with a few variables it should be noted that the policy implications of causality between energy consumption and real GDP are not straightforward. Energy conservation policies cannot sensibly be constituted without considering economic or environmental factors such as energy supply 2 For a detailed overview of the empirical literature on the causal relationship between energy consumption and economic growth see the recent surveys by Ozturk (2010) and Payne (2010). 3 Although many of the listed studies also report results for developing countries, we only show their results with respect to developed countries to save space. 783A. Belke et al. / Energy Economics 33 (2011) 782–789infrastructure, energy efficiency considerations or institutional constraints (Mahadevan and Asafu-Adjaye, 2007). Hence, the formulation of an efficient energy policy is more complicated than empirical results might suggest. For instance, energy conservation policies accomplishing a reduction in energy consumption due to an improved energy efficiency may raise t... |

52 | Energy consumption and GDP in developing countries: a cointegrated panel analysis.
- Lee
- 2005
(Show Context)
Citation Context ...ECT ΔP, ECT ΔE – 13.58*** 8.22*** 13.92*** – 9.31*** 13.55*** ΔY 3.01** – 17.87*** 14.04*** 8.92*** – 14.26*** ΔP 2.53* 1.98 – 8.93*** 4.18*** 3.08** – Notes: We report empirical realisations of the Wald F-test statistics. Potential heteroscedasticity of the error terms is corrected by using White robust standard errors. ECT represents the coefficient of the error-correction terms υ, ε and η, respectively. ***, ** and * indicate that the null hypothesis of no causation is rejected at the 1%, 5% and 10% levels.growth to energy consumption, and, on the other, those of the panel data analyses by Lee (2005) (18developing countries), Sari and Soytas (2007) (6 developing countries) Lee and Chang (2008) (16 Asian countries), Narayan and Smyth (2008) (G-7 countries) and Apergis and Payne (2009b) (6 Central American countries), who inferred that causality runs from energy consumption to economic growth. Further, our empirical results also refute the neutrality hypothesis such as all other panel data studies on the energy consumption-growth nexus. 4. Conclusions and policy implications In our contribution, we study the causal relationship between energy consumption and economic growth for 25 OECD coun... |

49 |
The relationship between energy consumption and Chinese economic growth based on panel data.
- Yu, Meng
- 2008
(Show Context)
Citation Context ...nd UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decomposition Canada, Italy and UK Energy→Growth France and Japan Growth→Energy Germany and USA E... |

46 |
Some cautions on the use of panel methods for integrated series of macro-economic data
- Banerjee, Marcellino, et al.
- 2004
(Show Context)
Citation Context ...ased on individual time series have low statistical power, especially when the time series is short (Campbell and Perron, 1991). Panelbased tests represent an improvement in this respect by exploiting additional information that results from the inclusion of the crosssectional dimension. However, first generation panel unit root and cointegration tests often assume that the cross-section members are independent. This condition is often likely to be violated, for example, because of common oil price shocks. Inappropriately assuming crosssectional independence can distort the panel results (see Banerjee et al. (2004), Urbain and Westerlund (2006)). Therefore, our study controls for cross-section dependencies by taking into account a common factor structure Yi;t = ξ1iF1t + E1i;t ; and ð1Þ Xi;t = ξ2iF2t + E2i;t ; ð2Þ where i=1,…,N represents the cross-section member and t=1,…,T refers to the time period, F denotes the common factors and E the idiosyncratic components. The common and idiosyncratic components can be either integrated of order one, I(1), or stationary, I(0), and we therefore have to test both components separately for unit roots and cointegration relationships, following the sequential testing... |

39 |
Energy Consumption and Economic Growth in Asian Economies: A More Comprehensive Analysis Using Panel Data. Resource and energy
- Lee, Chang
- 2008
(Show Context)
Citation Context ....04*** 8.92*** – 14.26*** ΔP 2.53* 1.98 – 8.93*** 4.18*** 3.08** – Notes: We report empirical realisations of the Wald F-test statistics. Potential heteroscedasticity of the error terms is corrected by using White robust standard errors. ECT represents the coefficient of the error-correction terms υ, ε and η, respectively. ***, ** and * indicate that the null hypothesis of no causation is rejected at the 1%, 5% and 10% levels.growth to energy consumption, and, on the other, those of the panel data analyses by Lee (2005) (18developing countries), Sari and Soytas (2007) (6 developing countries) Lee and Chang (2008) (16 Asian countries), Narayan and Smyth (2008) (G-7 countries) and Apergis and Payne (2009b) (6 Central American countries), who inferred that causality runs from energy consumption to economic growth. Further, our empirical results also refute the neutrality hypothesis such as all other panel data studies on the energy consumption-growth nexus. 4. Conclusions and policy implications In our contribution, we study the causal relationship between energy consumption and economic growth for 25 OECD countries from 1981 to 2007, explicitly taking into account the role of energy prices. We provide n... |

32 |
On the temporal causal relationship between energy consumption, real income, and prices: some new evidence from Asian-energy dependent NICs Based on a multivariate cointegration/vector error-correction approach.
- Masih, Masih
- 1997
(Show Context)
Citation Context ...on side can explain possible economic growth without a considerable increase in energy consumption. Therefore, the empirical results of causality analyses should be interpreted with caution. Our analysis of the relationship between energy consumption and economic activity is based on a sample of 25 OECD countries from 1981 to 2007 and uses recently developed panel-econometric methods. We explore an additional channel of causality by introducing energy prices. As energy prices have been neglected in many previous studies, the long-run parameters and the evidence of causality may be biased, see Masih and Masih (1997) and Asafu-Adjaye (2000). But in contrast to these two studies, we examine the original energy price index rather than the consumer price index (CPI) as a proxy. Income and price elasticities provide policy makers a suggestion of the extent to which prices need to increase, in the form of energy taxes, in order to reduce energy consumption and the potential for the market to conserve energy (Lee and Lee, 2010). Additionally, energy companies need this information to design their demand management policies. But only a few papers have estimated income and price elasticities for energy consumptio... |

30 |
Energy consumption and real GDP in G7 countries: New evidence from panel cointegration with structural breaks.
- Narayan, Smyth
- 2008
(Show Context)
Citation Context ...g (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decomposition Canada, Italy and UK Energy→Growth France and Japan Growth→Energy Germany and USA Energy~Growth Lee and Lee (2010) Trivariate Panel VECM 25 OECD countries Energy ↔ Growth Notes: X→Y means variable X Granger-causes variable Y. X~Y means that there... |

30 |
Energy consumption, income, and carbon emissions in the United States.
- Soytas, Sari, et al.
- 2007
(Show Context)
Citation Context ...ium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-... |

27 | Energy Consumption, Economic Growth and Prices: A Reassessment Using Panel VECM for Developed and Developing Countries. Energy Policy,
- Mahadevan, Asafu-Adjaye
- 2007
(Show Context)
Citation Context ...n policies cannot sensibly be constituted without considering economic or environmental factors such as energy supply 2 For a detailed overview of the empirical literature on the causal relationship between energy consumption and economic growth see the recent surveys by Ozturk (2010) and Payne (2010). 3 Although many of the listed studies also report results for developing countries, we only show their results with respect to developed countries to save space. 783A. Belke et al. / Energy Economics 33 (2011) 782–789infrastructure, energy efficiency considerations or institutional constraints (Mahadevan and Asafu-Adjaye, 2007). Hence, the formulation of an efficient energy policy is more complicated than empirical results might suggest. For instance, energy conservation policies accomplishing a reduction in energy consumption due to an improved energy efficiency may raise the productivity of energy consumption, which in turn may stimulate economic growth. Thus, a shift from less efficient energy sources tomore efficient and less pollutingoptionsmayestablish a stimulus rather than an obstacle to economic growth (Costantini and Martini, 2010). The other way around, bad energy supply infrastructure or other supply sid... |

26 |
Energy-GDP relationship revisited: an example from GCC countries using panel causality.
- Al-Iriani
- 2006
(Show Context)
Citation Context ...f energy prices. With respect to the widely studied energy consumption–growth nexus, a bi-directional causal relationship between energy consumption and economic growth is also reported by all recent panel data analyses of OECD countries (Lee and Chang (2007), Mahadevan and Asafu-Adjaye (2007), Lee et al. (2008), Costantini and Martini (2010) and Lee and Lee (2010)) and by Apergis and Payne (2009a) for a panel of 11 countries of the Commonwealth of Independent States. However, compared with other previous studies our findings contradict, on the one hand, those of Kraft and Kraft (1978) (USA), Al-Iriani (2006) (6 Gulf Cooperation Council countries) and Huang et al. (2008) (26 high income countries), who found a uni-directional causal relationship running from economicTable 6 Panel causality test results for energy consumption, GDP and energy prices. Dependent variable Sources of causation (independent variables) Short-run Long-run Strong causality ΔE ΔY ΔP ECT ΔE, ECT ΔY, ECT ΔP, ECT ΔE – 13.58*** 8.22*** 13.92*** – 9.31*** 13.55*** ΔY 3.01** – 17.87*** 14.04*** 8.92*** – 14.26*** ΔP 2.53* 1.98 – 8.93*** 4.18*** 3.08** – Notes: We report empirical realisations of the Wald F-test statistics. Potenti... |

24 |
The performance of panel cointegration methods: results from a large scale simulation study.
- Wagner, Hlouskova
- 2010
(Show Context)
Citation Context ...erial correlation. First, the endogenous variable in each equation is regressed on the leads and lags of the first-differenced regressors from all equations to control for potential endogeneities. Then the OLS method is applied using the residuals from the first step regression. The DOLS estimator is preferred to the non-parametric9 Since the panel unit root tests of the idiosyncratic components suggest stationarity, we do not test for cointegration between the idiosyncratic components of energy consumption, GDP and energy prices.FMOLS estimator because of its better performance. According to Wagner and Hlouskova (2010), the DOLS estimator outperforms all other studied estimators, both single equation estimators and system estimators, even for large samples. Furthermore, Harris and Sollis (2003) suggest that non-parametric approaches such as FMOLS are less robust if the data have significant outliers and also have problems in cases where the residuals have large negative moving average components, which is a fairly common occurrence in macro time series data. The estimated models are: Ei;t = α1i + δ1it + β1iYi;t + γ1iPi;t + υi;t Yi;t = α2i + δ2it + β2iEi;t + γ2iPi;t + εi;t Pi;t = α3i + δ3it + β3iEi;t + γ3iYi... |

23 |
Energy Consumption and GDP revisited: a panel analysis of and Technical Economics 25,17-31. (Chinese Journal) Soytas,
- Lee, Chang
- 2007
(Show Context)
Citation Context ..., Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan ... |

22 |
Energy Consumption and Economic Growth in Central America: Evidence from a Panel Cointegration and Error Correction Model,
- Apergis, Payne
- 2009
(Show Context)
Citation Context ... c Institute for the Study of Labour (IZA), Schaumburg-Lippe-Strasse 5–9, 53113 Bonn, Germany d Ruhr Graduate School in Economics (RGS Econ), Germany e Department of Economics, University of Frankfurt (Oder), post office box 1786, 15207 Frankfurt (Oder), Germany⁎ Corresponding author at: Department of Economics Universitätsstrasse 12, 45117 Essen, Germany. Tel.: +4 183 4181. E-mail address: Frauke.Dobnik@uni-due.de (F. Dobn 1 The different directions of causality between energ growth have been described in many previous studie hypothesis’ has often been used. The denotations of th proposed by Apergis and Payne (2009a,b). 0140-9883/$ – see front matter © 2011 Elsevier B.V. A doi:10.1016/j.eneco.2011.02.005a b s t r a c ta r t i c l e i n f oArticle history: Received 8 June 2010 Received in revised form 10 December 2010 Accepted 6 February 2011 Available online 21 February 2011 JEL classification: C33 O13 Q43 Keywords: Energy consumption Panel unit roots Panel cointegration Vector error-correction models Granger causalityThis paper examines the long-run relationship between energy consumption and real GDP, including energy prices, for 25 OECD countries from 1981 to 2007. The distinction between common fact... |

21 |
Cointegration testing in panels with common factors
- Gengenbach, Palm, et al.
- 2006
(Show Context)
Citation Context ...d (2006)). Therefore, our study controls for cross-section dependencies by taking into account a common factor structure Yi;t = ξ1iF1t + E1i;t ; and ð1Þ Xi;t = ξ2iF2t + E2i;t ; ð2Þ where i=1,…,N represents the cross-section member and t=1,…,T refers to the time period, F denotes the common factors and E the idiosyncratic components. The common and idiosyncratic components can be either integrated of order one, I(1), or stationary, I(0), and we therefore have to test both components separately for unit roots and cointegration relationships, following the sequential testing strategy proposed by Gengenbach et al. (2006). Cointegration implies that both the common and idiosyncratic parts of the error term are stationary. 3.1. Variable decomposition The first and novel step of this paper regarding the energy consumption–growth nexus is to decompose each variable into two uncorrelated components, i.e. a common and an idiosyncratic component, by principal component analysis. The idiosyncratic component is a residual, which captures the impact of shocks affecting the respective variable of one specific country. These country-specific shocks, such as changes in national energy taxes, may have large but4 As expecte... |

20 |
Survey of the international evidence on the causal relationship between energy consumption and growth.
- Payne
- 2010
(Show Context)
Citation Context ...versa. Hence, energy conservation policies would not have any impact on real GDP. In consideration of such a pure statistical causality analysis with a few variables it should be noted that the policy implications of causality between energy consumption and real GDP are not straightforward. Energy conservation policies cannot sensibly be constituted without considering economic or environmental factors such as energy supply 2 For a detailed overview of the empirical literature on the causal relationship between energy consumption and economic growth see the recent surveys by Ozturk (2010) and Payne (2010). 3 Although many of the listed studies also report results for developing countries, we only show their results with respect to developed countries to save space. 783A. Belke et al. / Energy Economics 33 (2011) 782–789infrastructure, energy efficiency considerations or institutional constraints (Mahadevan and Asafu-Adjaye, 2007). Hence, the formulation of an efficient energy policy is more complicated than empirical results might suggest. For instance, energy conservation policies accomplishing a reduction in energy consumption due to an improved energy efficiency may raise the productivity o... |

19 |
CO2 emissions, energy consumption, and output in France.
- Ang
- 2007
(Show Context)
Citation Context ...Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Pane... |

18 |
Energy consumption and economic growth in Korea: testing the causality relation.
- Oh, Lee
- 2004
(Show Context)
Citation Context ...results.3 The ambiguous evidence of the empirical literature on the causal relationship between energy consumption and economic growth can already be seen from this recent studies on developed countries. Panel data analyses of OECD countries, however, all indicate bi-directional causality. Most of the studies dealing with the energy consumption-growth nexus focus on production side models, which often include capital stock and labour in addition to energy consumption and GDP. If one concentrates on energy demand, trivariatemodelswith energy prices as an additional variable should be used (see Oh and Lee (2004b)). The studies by Masih and Masih (1998), Asafu-Adjaye (2000), Fatai et al. (2004) as well as Mahadevan and Asafu-Adjaye (2007) take the consumer price index (CPI) as a proxy of the energy price. However, as the CPI is known not to capture the energy price very well, we employ the real energy price index, such as Lee and Lee (2010) and Costantini and Martini (2010). Masih and Masih (1998) and Asafu-Adjaye (2000) previously used the vector error-correction model (VECM); Fatai et al. (2004) applied the autoregressive distributed lag (ARDL) approach; and Mahadevan and Asafu-Adjaye (2007), Lee a... |

16 |
Modelling the causal relationship between energy consumption and GDP in
- Fatai, Oxley, et al.
- 2004
(Show Context)
Citation Context ...onship between energy consumption and economic growth can already be seen from this recent studies on developed countries. Panel data analyses of OECD countries, however, all indicate bi-directional causality. Most of the studies dealing with the energy consumption-growth nexus focus on production side models, which often include capital stock and labour in addition to energy consumption and GDP. If one concentrates on energy demand, trivariatemodelswith energy prices as an additional variable should be used (see Oh and Lee (2004b)). The studies by Masih and Masih (1998), Asafu-Adjaye (2000), Fatai et al. (2004) as well as Mahadevan and Asafu-Adjaye (2007) take the consumer price index (CPI) as a proxy of the energy price. However, as the CPI is known not to capture the energy price very well, we employ the real energy price index, such as Lee and Lee (2010) and Costantini and Martini (2010). Masih and Masih (1998) and Asafu-Adjaye (2000) previously used the vector error-correction model (VECM); Fatai et al. (2004) applied the autoregressive distributed lag (ARDL) approach; and Mahadevan and Asafu-Adjaye (2007), Lee and Lee, (2010) as well as Costantini and Martini (2010) used a panel vector error-co... |

16 |
Trends and breaks in per-capita carbon dioxide emissions,
- Lanne, Liski
- 2004
(Show Context)
Citation Context ...cratic component, by principal component analysis. The idiosyncratic component is a residual, which captures the impact of shocks affecting the respective variable of one specific country. These country-specific shocks, such as changes in national energy taxes, may have large but4 As expected, the time pattern of the energy prices exhibits considerable peaks at the period of the oil crisis. We originally collected data from 1978 to 2007. 5 We use per capita data because they are less sensitive to territorial changes and provide the variables in the same units for countries of different sizes (Lanne and Liski, 2004). 6 More precisely, data for real GDP per capita and final energy consumption are taken from IEA publications on energy balances of OECD countries (annual), while data for energy prices are drawn from IEA statistics on energy prices and taxes (quarterly). Additionally, we use data for population of the IMF's International Financial Statistics to obtain energy consumption in terms of per capita.geographically concentrated effects. The common component of a variable is ‘common’ in the sense that it depends on a small number of common shocks, which affect the respective variable of all the countr... |

16 |
A multivariate cointegrated modeling approach in testing temporal causality between energy consumption, real income and prices with an application to two Asian LDCs.
- Masih, Masih
- 1998
(Show Context)
Citation Context ...the empirical literature on the causal relationship between energy consumption and economic growth can already be seen from this recent studies on developed countries. Panel data analyses of OECD countries, however, all indicate bi-directional causality. Most of the studies dealing with the energy consumption-growth nexus focus on production side models, which often include capital stock and labour in addition to energy consumption and GDP. If one concentrates on energy demand, trivariatemodelswith energy prices as an additional variable should be used (see Oh and Lee (2004b)). The studies by Masih and Masih (1998), Asafu-Adjaye (2000), Fatai et al. (2004) as well as Mahadevan and Asafu-Adjaye (2007) take the consumer price index (CPI) as a proxy of the energy price. However, as the CPI is known not to capture the energy price very well, we employ the real energy price index, such as Lee and Lee (2010) and Costantini and Martini (2010). Masih and Masih (1998) and Asafu-Adjaye (2000) previously used the vector error-correction model (VECM); Fatai et al. (2004) applied the autoregressive distributed lag (ARDL) approach; and Mahadevan and Asafu-Adjaye (2007), Lee and Lee, (2010) as well as Costantini and M... |

16 |
Energy consumption and income in G-7 countries.
- Soytas, Sari
- 2006
(Show Context)
Citation Context ... a more detailed description of the LLC and IPS test we recommend Baltagi (2008). Table 1 Overview of studies on developed countries of the last few years. Study Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy... |

15 |
Energy consumption and economic growth: evidence from the Commonwealth of Independent States.
- Apergis, Payne
- 2009
(Show Context)
Citation Context ... c Institute for the Study of Labour (IZA), Schaumburg-Lippe-Strasse 5–9, 53113 Bonn, Germany d Ruhr Graduate School in Economics (RGS Econ), Germany e Department of Economics, University of Frankfurt (Oder), post office box 1786, 15207 Frankfurt (Oder), Germany⁎ Corresponding author at: Department of Economics Universitätsstrasse 12, 45117 Essen, Germany. Tel.: +4 183 4181. E-mail address: Frauke.Dobnik@uni-due.de (F. Dobn 1 The different directions of causality between energ growth have been described in many previous studie hypothesis’ has often been used. The denotations of th proposed by Apergis and Payne (2009a,b). 0140-9883/$ – see front matter © 2011 Elsevier B.V. A doi:10.1016/j.eneco.2011.02.005a b s t r a c ta r t i c l e i n f oArticle history: Received 8 June 2010 Received in revised form 10 December 2010 Accepted 6 February 2011 Available online 21 February 2011 JEL classification: C33 O13 Q43 Keywords: Energy consumption Panel unit roots Panel cointegration Vector error-correction models Granger causalityThis paper examines the long-run relationship between energy consumption and real GDP, including energy prices, for 25 OECD countries from 1981 to 2007. The distinction between common fact... |

15 |
The causal relationship between U.S. energy consumption and real output: a disaggregated analysis.
- Bowden, Payne
- 2009
(Show Context)
Citation Context ...d Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decomposition Canada, Italy and UK Energy→Growth France and Japan Growth→Energy Germany and USA Energy~Growth Lee and Lee (2010) Trivariate Panel VECM 25 OECD countries Energy ↔ Growth Notes: X→Y means variable X Granger-causes variable Y. X~Y means that there exists no Granger-causality. Table 2 Time series unit root tests for comm... |

15 |
Causality between energy consumption and GDP: evidence from 30 OECD and 78 Non-OECD countries. Surrey Energy Economics Discussion Paper Series.
- Chontanawat, Hunt, et al.
- 2006
(Show Context)
Citation Context ...root tests propose that the idiosyncratic components are widely stationary (see Table 3). Hence, the results suggest that random walks in the data are driven mainly by international developments. As a consequence, cointegration i.e. a long-run relationship may exist between the common rather than the idiosyncratic components.7 Further details of the ADF tests can also be found in Harris and Sollis (2003). 8 For a more detailed description of the LLC and IPS test we recommend Baltagi (2008). Table 1 Overview of studies on developed countries of the last few years. Study Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Ener... |

15 |
The causality between energy consumption and economic growth: a multi-sectoral analysis using non-stationary cointegrated panel data”,
- Costantini, Martini
- 2010
(Show Context)
Citation Context ...ructure, energy efficiency considerations or institutional constraints (Mahadevan and Asafu-Adjaye, 2007). Hence, the formulation of an efficient energy policy is more complicated than empirical results might suggest. For instance, energy conservation policies accomplishing a reduction in energy consumption due to an improved energy efficiency may raise the productivity of energy consumption, which in turn may stimulate economic growth. Thus, a shift from less efficient energy sources tomore efficient and less pollutingoptionsmayestablish a stimulus rather than an obstacle to economic growth (Costantini and Martini, 2010). The other way around, bad energy supply infrastructure or other supply side disruptions decreasing energy consumption could indeed induce an adverse impact on economic growth. Further, high substitutability between energy and other input factors on the production side can explain possible economic growth without a considerable increase in energy consumption. Therefore, the empirical results of causality analyses should be interpreted with caution. Our analysis of the relationship between energy consumption and economic activity is based on a sample of 25 OECD countries from 1981 to 2007 and ... |

14 |
Energy-income causality in OECD countries revisited: the key role of capital stock.
- Lee, Chang, et al.
- 2008
(Show Context)
Citation Context ...ment and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decomposition Canada, Italy and UK Energy→Growth France and Japan Growth→Energy Germany and USA Energy~Growth Lee and Lee (2010) Trivariate Panel VECM 25 OECD countries Energy ↔ G... |

14 |
On the dynamics of energy consumption and output in the US,
- Payne
- 2009
(Show Context)
Citation Context ... for the Study of Labour (IZA), Schaumburg-Lippe-Strasse 5–9, 53113 Bonn, Germany d Ruhr Graduate School in Economics (RGS Econ), Germany e Department of Economics, University of Frankfurt (Oder), post office box 1786, 15207 Frankfurt (Oder), Germany⁎ Corresponding author at: Department of Economics Universitätsstrasse 12, 45117 Essen, Germany. Tel.: +4 183 4181. E-mail address: Frauke.Dobnik@uni-due.de (F. Dobn 1 The different directions of causality between energ growth have been described in many previous studie hypothesis’ has often been used. The denotations of th proposed by Apergis and Payne (2009a,b). 0140-9883/$ – see front matter © 2011 Elsevier B.V. A doi:10.1016/j.eneco.2011.02.005a b s t r a c ta r t i c l e i n f oArticle history: Received 8 June 2010 Received in revised form 10 December 2010 Accepted 6 February 2011 Available online 21 February 2011 JEL classification: C33 O13 Q43 Keywords: Energy consumption Panel unit roots Panel cointegration Vector error-correction models Granger causalityThis paper examines the long-run relationship between energy consumption and real GDP, including energy prices, for 25 OECD countries from 1981 to 2007. The distinction between common fact... |

13 |
Exploring the relationship between energy use and economic growth with bivariate models: New evidence from G-7 countries,
- Zachariadis
- 2007
(Show Context)
Citation Context ...ly and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and... |

12 |
Economic growth and energy consumption revisited — evidence from linear and non-linear Granger causality.
- Chiou-Wei, Chen, et al.
- 2008
(Show Context)
Citation Context ...alised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decompositi... |

8 |
The role of asset markets for private consumption: Evidence from paneleconometric models.
- Dreger, Reimers
- 2009
(Show Context)
Citation Context ...owever, have a different relevance for individual countries. Taking this decomposition as a starting point, cointegration between the common components means that the common components of energy consumption, real GDP and energy prices move together in the long run and do not deviate permanently from one another. Hence, cointegration between the common components suggests that the relationship between these variables depends to a great extent on international developments. Instead, cointegration between idiosyncratic components refers to developments relevant exclusively on the national level (Dreger and Reimers, 2009). Depending on the results of the cointegration tests, this distinction has important implications for policy makers. If the common components cointegrate, national energy policies may not have a large impact on economic growth. Indeed, this paper delivers empirical evidence that energy consumption, real GDP and energy prices are cointegrated in their common factors, but not in their idiosyncratic components. The remainder of this paper is organised as follows. Section 2 briefly reviews the literature related to the causal relationship between energy consumption and economic growth. Section 3 ... |

7 |
A panel data analysis of the demand for total energy and electricity in OECD countries.
- Lee, Lee
- 2010
(Show Context)
Citation Context ...nel of causality by introducing energy prices. As energy prices have been neglected in many previous studies, the long-run parameters and the evidence of causality may be biased, see Masih and Masih (1997) and Asafu-Adjaye (2000). But in contrast to these two studies, we examine the original energy price index rather than the consumer price index (CPI) as a proxy. Income and price elasticities provide policy makers a suggestion of the extent to which prices need to increase, in the form of energy taxes, in order to reduce energy consumption and the potential for the market to conserve energy (Lee and Lee, 2010). Additionally, energy companies need this information to design their demand management policies. But only a few papers have estimated income and price elasticities for energy consumption in a panel framework. Furthermore, the long-run equilibrium relationship is studied in both directions, i.e. with either energy consumption or real GDP as a dependent variable (Costantini and Martini, 2010). The innovative contribution of our paper is to determine the longrun relationship between energy consumption, real GDP and energy prices in more detail. In contrast to other studies concerning the energy... |

5 |
Decoupling factors on the energy-output linkage: The Spanish case. Energy Policy
- Climent, Pardo
- 2007
(Show Context)
Citation Context ... Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition France and USA Energy→Growth Canada, Germany, Italy, Japan and UK Energy ↔ Growth Ang (2007) Multivariate VECM and ARDL France Energy→Growth Climent and Pardo (2007) Multivariate VECM Spain Energy ↔ Growth Lee and Chang (2007) Bivariate Panel VAR 22 OECD countries Energy ↔ Growth Mahadevan and Asafu-Adjaye (2007) Trivariate Panel VECM 6 OECD countries Energy ↔ Growth Soytas et al. (2007) Multivariate Toda-Yamamoto USA Energy~Growth Zachariadis (2007) Bivariate VECM, ARDL & Toda-Yamamoto Canada and UK Growth→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Tri... |

5 |
Dynamic modelling of energy consumption, capital stock, and real income in G-7 countries.
- Lee, Chien
- 2010
(Show Context)
Citation Context ...h→Energy Italy and Japan Energy ↔ Growth USA Energy~Growth Chiou-Wei et al. (2008) Bivariate non-linear Baek-Brock USA and South Korea Energy~Growth Huang et al. (2008) Dynamic Panel Estimation 26 high income countries Growth→Energy Lee et al. (2008) Trivariate Panel VECM 22 OECD countries Energy ↔ Growth Narayan and Smyth (2008) Trivariate Panel VECM G-7 countries Energy→Growth Bowden and Payne (2009) Multivariate Toda-Yamamoto USA Energy→Growth Payne (2009) Bivariate Toda-Yamamoto USA Energy~Growth Costantini and Martini (2010) Bi- and trivariate Panel VECM 26 OECD countries Energy ↔ Growth Lee and Chien (2010) Trivariate Toda-Yamamoto and generalised variance decomposition Canada, Italy and UK Energy→Growth France and Japan Growth→Energy Germany and USA Energy~Growth Lee and Lee (2010) Trivariate Panel VECM 25 OECD countries Energy ↔ Growth Notes: X→Y means variable X Granger-causes variable Y. X~Y means that there exists no Granger-causality. Table 2 Time series unit root tests for common components of energy consumption, GDP and energy prices. Variable ADF PP KPSS E −2.19(0) −2.71[2] 0.17[3]** ΔE −3.06(0)*** −3.07[3]*** 0.15[2] Y −2.61(1) −1.72[2] 0.14[3]* ΔY −2.47(0)** −2.47[0]*** 0.23[2] P −0.5... |

1 |
auflage edition. Econometric Analysis of Panel Data,
- Baltagi
- 2008
(Show Context)
Citation Context ...for robustness. In contrast to the time series unit root evidence for the common components, the LLC and IPS panel unit root tests propose that the idiosyncratic components are widely stationary (see Table 3). Hence, the results suggest that random walks in the data are driven mainly by international developments. As a consequence, cointegration i.e. a long-run relationship may exist between the common rather than the idiosyncratic components.7 Further details of the ADF tests can also be found in Harris and Sollis (2003). 8 For a more detailed description of the LLC and IPS test we recommend Baltagi (2008). Table 1 Overview of studies on developed countries of the last few years. Study Method Countries Result Chontanawat et al. (2006) Bivariate Hsiao Bg, Cz, Dk, Ir, Ko, Mx, Nl, Oe, Po and Sw Energy→Growth Au, Cn, Fn, Es and Sd Growth→Energy Fr, Bd, Gr, Hu, Ic, It Jp, Nz, Nw, Pt, and Sx Energy ↔ Growth Lx, UK and USA Energy~Growth Lee (2006) Bivariate Toda-Yamamoto Belgium, Netherlands, Canada and Switzerland Energy→Growth France, Italy and Japan Growth→Energy USA Energy ↔ Growth Germany, Sweden and UK Energy~Growth Soytas and Sari (2006) Multivariate VECM and generalised variance decomposition ... |

1 | The growth of income and energy consumption in six developing countries. - Belke - 2011 |