ABSTRACT
The expected gains from RES deployment to the reduction of carbon dioxide emissions (CO2) and the cut-off of external dependence of electricity sources could be important. However, it is crucial to understand the determinants of RES growth to help policymakers drawing effective energy polices, involving a commitment of both citizens and governments. In this paper, we use novel panel econometric tools (taking into account structural breaks and cross-section dependence) and find evidence of nonstationary issues and cointegration issues between renewable energy production and its drivers (CO2 emissions, GDP per capita, energy use and dependency). The results thus reveal that non-stationary issues should be attended, otherwise they could be biased. Using suitable estimators (DOLS, FMOLS) with two different data sets and different proxies and taking common factors into account by MG estimates, we find that there is no environmental concerns effect explaining the growth of renewables in European countries. However, national revenues, energy consumption (demand effect) and energy dependency have a positive impact on renewables deployment. Considering these results, economic assistance (subsidies) might be a mean to increase further the renewables deployment in EU countries and education about renewables deployment is needed.
Acknowledgments
Support from Region Lorraine, FEDER and ARIANE grants are gratefully acknowledged. This paper was presented at the IIES International Conference at Prague (2016), at the Dijon seminar of the University of Burgundy (2017), LEM at Lille (2018), at the SCSE conference in Ottawa (2017), at the EBES conference in Madrid (2017) and at the Poznan mathematical economics seminar (2017).
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Note however that some time series studies outlined non-stationary issues like Rafiq, Bloch, and Salim (Citation2014) concerning India and China cases using VECM estimations.
2 See also Cadoret and Padovano (Citation2016) but they focus on the role of political factors in the deployment of renewable energy.
3 Abbreviations for estimators hereafter: FE: Fixed-Effects; FMOLS: Fully Modified Ordinary Least Squares; DOLS: Dynamic Ordinary Least Squares (Kao and Chiang Citation2000); MG: Mean Group estimator (Pesaran and Smith, 1995, https://www.sciencedirect.com/science/article/pii/030440769401644F); CCEMG: Common Correlated Effects Mean Group estimator (Pesaran Citation2006); AMG: Augmented Mean Group estimator (Bond and Eberhardt Citation2009; Eberhardt and Teal Citationn.d.).
4 Moscone and Tosetti (Citation2009) evaluate other tests to assess cross-sectional dependence but none perform better than the Pesaran (Citation2004) one.
5 In the same spirit, see also the extension of the Pesaran test to the case of several common stationary factors (Pesaran Smith and Yamagata, 2013 developing CIPSm and CSBm statistics).