Scholarlink

<p>This study investigates the impact of renewable energy (RE) and artificial intelligence (AI) on green growth in 18 G20 countries from 2000 to 2023, employing Cross-Sectional Pooled Mean Group ARDL (CS-PMG-ARDL) and Nonlinear ARDL (CS-PMG-NARDL) models to capture symmetric and asymmetric dynamics. The bounds test confirms cointegration (F = 28.27, <i>p</i> &lt; 0.001), and the error correction term indicates stable long-run adjustment (ECT = −0.145, <i>p</i> &lt; 0.001 in ARDL; ECT = −0.115, <i>p</i> = 0.024 in NARDL). Results reveal that renewable energy exerts a positive and significant long-run effect on green growth (0.101, <i>p</i> &lt; 0.001), with asymmetric responses confirmed by Wald tests (short-run χ² = 4.102, <i>p</i> = 0.043; long-run χ² = 5.42, <i>p</i> = 0.020): positive RE shocks yield stronger benefits (0.012, <i>p</i> = 0.011) than the adverse effects of negative shocks, which are statistically weaker in the short run ( − 0.020, <i>p</i> = 0.111) and significant but smaller in the long run ( − 0.012, <i>p</i> = 0.015). AI shows a significant short-run impact (0.018, <i>p</i> &lt; 0.001) but becomes insignificant in the long run (0.001, <i>p</i> = 0.806), suggesting its effects are conditional on institutional and technological contexts. Critically, GMM estimations highlight significant synergistic interaction effects: the RE×AI term is positive (0.007, <i>p</i> = 0.004), indicating that AI amplifies the marginal contribution of renewable energy to green growth, while the RE×CO<sub>2</sub> interaction is negative ( − 0.041, <i>p</i> &lt; 0.001), underscoring that high emissions undermine renewable energy benefits. The models exhibit strong explanatory power (adjusted R² = 0.909 for ARDL; 0.999 for NARDL) and pass all diagnostic tests for instrument validity and absence of second-order autocorrelation. The study provides tailored policy recommendations for G20 nations, emphasizing integrated strategies that combine green AI applications, renewable energy expansion, and institutional reforms to foster resilient, low-carbon economic growth.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Asymmetric effects of renewable energy and artificial intelligence on green growth: evidence from G20 countries

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月10日

查看全文
  • Olfa ZARRAD,
  • Mariem BOUATTOUR,
  • Sourour GUIDARA,
  • Kamel HELALI

摘要

This study investigates the impact of renewable energy (RE) and artificial intelligence (AI) on green growth in 18 G20 countries from 2000 to 2023, employing Cross-Sectional Pooled Mean Group ARDL (CS-PMG-ARDL) and Nonlinear ARDL (CS-PMG-NARDL) models to capture symmetric and asymmetric dynamics. The bounds test confirms cointegration (F = 28.27, p < 0.001), and the error correction term indicates stable long-run adjustment (ECT = −0.145, p < 0.001 in ARDL; ECT = −0.115, p = 0.024 in NARDL). Results reveal that renewable energy exerts a positive and significant long-run effect on green growth (0.101, p < 0.001), with asymmetric responses confirmed by Wald tests (short-run χ² = 4.102, p = 0.043; long-run χ² = 5.42, p = 0.020): positive RE shocks yield stronger benefits (0.012, p = 0.011) than the adverse effects of negative shocks, which are statistically weaker in the short run ( − 0.020, p = 0.111) and significant but smaller in the long run ( − 0.012, p = 0.015). AI shows a significant short-run impact (0.018, p < 0.001) but becomes insignificant in the long run (0.001, p = 0.806), suggesting its effects are conditional on institutional and technological contexts. Critically, GMM estimations highlight significant synergistic interaction effects: the RE×AI term is positive (0.007, p = 0.004), indicating that AI amplifies the marginal contribution of renewable energy to green growth, while the RE×CO2 interaction is negative ( − 0.041, p < 0.001), underscoring that high emissions undermine renewable energy benefits. The models exhibit strong explanatory power (adjusted R² = 0.909 for ARDL; 0.999 for NARDL) and pass all diagnostic tests for instrument validity and absence of second-order autocorrelation. The study provides tailored policy recommendations for G20 nations, emphasizing integrated strategies that combine green AI applications, renewable energy expansion, and institutional reforms to foster resilient, low-carbon economic growth.