Exploring the effect of renewable energy on low-carbon sustainable development in the Belt and Road Initiative countries: evidence from the spatial-temporal perspective

The Belt and Road Initiative (BRI) has promoted the deployment of renewable energy to achieve sustainability. It is essential to reveal the influence of renewable energy on low-carbon economic development. The share of renewable energy consumption (SREC) is taken as the core explanatory variable in this paper, and its impacts on carbon emission intensity (CEI) and economic growth are investigated from the spatial-temporal perspective. First, the panel Granger causality test is applied for revealing the causal links among SREC, CEI, and economic growth during 1999–2017. Then, this paper investigates the impacts of SREC on economic growth and CEI through rigorous econometric techniques. Based on the regression results, Shapley value decomposition is utilized to account for the cross-country inequalities of economic growth and CEI. The main findings are as follows: (1) There exist bidirectional Granger causalities between SREC, economic growth, and CEI, which shows there is a systematic link between the three variables. (2) All models demonstrate SREC negatively influences economic growth, indicating renewable energy deployment costs are urgent to be decreased with SREG increasing. Besides, capital investment and openness positively affect economic growth, but energy intensity has an opposite impact. (3) From the spatial heterogeneity perspective, the cross-country inequality in economic growth is primarily due to the regional inequality of capital investment, followed by energy intensity and SREC. By contrast, the impacts of labor and openness are negligible. (4) SREC has a negative effect on CEI. In addition, an inverted U-shaped nexus between economic growth and CEI is observed. Energy intensity positively affects CEI, while the impacts of urbanization and openness are insignificant. (5) From the spatial heterogeneity perspective, the cross-country CEI inequality is mostly caused by the inequality of energy intensity, followed by SREC, urbanization, and economic growth, while the contribution of the openness gap is little. This article provides important implications for low-carbon development in the BRI countries.


37
Since the industrial revolution, economic growth has been accompanied by a large  the relationships between EG, CEs, non-renewable and renewable electricity generation. 168 They find no causality between renewable electricity and EG and CEs. Many studies 169 emphasize revealing the impacts of REC on CEs or EG, while CEs and EG in turn have 170 important impacts on REC (Sadorsky 2009 Where − refers to the k-order lagged term of , and represents its 193 coefficient; − indicates the k-order lagged term of , and is its coefficient.

204
This article checks the causal dynamics between SREC, EG, and CEI through the 205 following VAR model: Douglas production function, to construct an economic growth model as:     Table 5 465 Regression results of the SREC-EG model.    Table 5.  Table 5). Model (1) and (2) present the estimation results obtained 474 from the 2SLS method. Models (3)-(6) report the results obtained from GMM methods.

475
On basis of the panel Granger causality test from Table 4, a bidirectional causality is 476 found between EG and SREC. Therefore, lnSREC is regarded as endogenous variables 477 in the econometric models. Specifically, 2SLS, difference-GMM and system-GMM 478 approaches are adopted for addressing the endogeneity problems. As shown in Table 5,   Table 6.  Table 4, EG and CEI, and SREC and CEI all display bidirectional causalities. 558 Therefore, SREC and economic growth are treated as endogenous variables in the 559 SREC-CEI model. This paper adopts three estimation strategies: 2SLS, difference-560 GMM and system-GMM methods, thereby solving the endogeneity problems. As 561 presented in Table 6, the R 2 reported in Model (7) and Model (8) is at least 92%, 562 indicating that the models fit well. 563 Table 6 564 Regression results of the SREC-CEI model.  Note: (1) * represents p<0.10, ** represents p<0.05, *** represents p<0.01.

614
In this paper, the inequalities in EG and CEI are evaluated by the Gini coefficient.

615
The changes in the inequalities indexes during 1999-2017 are shown in Fig. 3. As 616 shown in Fig. 3, over the period 199-2017, the Gini coefficient of CEI is significantly 617 larger than the Gini coefficient of EG, indicating the regional inequality in CEI is 618 significantly larger than the regional inequality in EG. It is clear that the two inequality In order to reveal the mechanism behind the regional inequality of EG, this study  Gini coefficient of EG Gini coefficient of CEI Furthermore, this paper compares the contribution rates of different factors (see Fig. 5).

642
As shown in Fig. 4 and Fig. 5, the contributions of all factors are positive, indicating 643 that the spatial distribution of individual factors contributes to the spatial heterogeneity 644 in economic growth. On the whole, the inequality in EG is primarily due to the regional 645 differences in capital investment, followed by energy intensity and renewable energy 646 consumption. By contrast, the impacts of labor and openness are negligible.

647
It is found that, in every single year, capital is the most critical factor influencing 648 the regional disparities of EG, with an average contribution rate of 81.54% during 1999-

686
As shown in Fig. 7