In Table 2, the descriptive statistics for Natural Resource Rent (NRR) show a minimum value of –2.343. How can NRR, measured as a percentage of GDP, be negative? Does this indicate data errors, transformation artifacts, or misreported values, and how might this affect your regression results?
Your correlation matrix (Table 2) shows a very high negative correlation (–0.848) between Economic Growth (EGC) and Renewable Energy Use (REU), and a high positive correlation (0.809) between EGC and ICT. Given these strong correlations, how did you ensure that multicollinearity did not bias your CS-ARDL estimates, especially since the mean VIF is reported but individual VIFs for EGC and ICT are not shown?
Your study uses only five South Asian countries over 28 years (N=5, T=28). How do you justify the use of second-generation panel techniques (CS-ARDL, CIPS, CADF) with such a small cross-sectional dimension, and what are the risks of overfitting or limited external validity?
Your model excludes potentially critical variables such as urbanization, industrialization, trade openness, and institutional quality. How might their omission affect the estimated coefficients, particularly for green finance and ICT, and how do you address the possibility of omitted variable bias?
