ABSTRACT
This paper studies the nonlinear impact of renewable energy on CO2 emissions with a panel of 30 Chinese provinces over the period 1990–2016. The Pooled Mean Group (PMG) estimator is applied for estimation, allowing for both long-run effects and short-run heterogeneity across provinces. The results reveal an inverted U-shaped relationship between renewable energy and CO2 emissions.
KEYWORDS:
Acknowledgments
We thank the financial support provided by the Contemporary China Marxist Political Economics Innovation Think Tank Project of Chinese Academy of Social Sciences and by the China Postdoctoral Science Foundation under Grant [No.2019M650945].
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Compared with static panel approaches of fixed effects or random effects models, the panel ARDL model can not only capture the dynamic nature of the data in the carbon-renewable-energy nexus but also allows heterogenous slop coefficients across the cross-section units. The panel ARDL model can produce a more reliable and consistent estimator than the GMM estimator in the case of small number of provinces (N) and long time period (T).
2 Tibet and Taiwan are not considered due to data unavailability.
3 The details for this calculation method is not presented here. See more in Dong et al. (Citation2017).
4 Before estimations, the tests of cross-section dependence and panel unit root tests are conducted and confirm the cross-section dependence and stationarity of all series. The results can be obtained upon request from the authors.
5 This result is also robust when non-fossil energy is estimated on CO2 emissions by adding nuclear energy to renewable energy.
6 The homogeneity tests indicate the existence of two-regimes switching relationship between renewable energy and CO2 emissions and the misspecification tests confirm the adequacy of the two-regime PSTR model.