Driving sustainable economic growth through the energy Indicators: Insights from the ARDL and machine learning approaches

Achieving affordable and clean energy targets is essential for sustainable development and high-quality economic growth, in alignment with the United Nations Agenda 2030. Previous studies have examined the effects of clean and non-clean energy on economic growth, yet limited attention has been given to assessing the role of Sustainable Development Goal 7 (SDG—7) and its interaction with economic growth. This study employs pre- and post-estimation tests, selecting the Autoregressive Distributed Lag (ARDL) and Machine Learning (ML) approaches to evaluate the interconnectedness among SDG—7 indicators and economic growth using annual time series data from 1990 to 2023. The findings reveal that the ARDL model offers significant understanding into both short- and long-term relationships between the underlined variables. Notably, energy indicators, including access to electricity, energy intensity, installed renewable energy capacity, and international financial inflows, were found to substantially contribute to long-term economic growth. Additionally, the Granger causality test identified both bidirectional and unidirectional causal relationships, with most variables exhibiting unidirectional causality. To validate the ARDL results, this study applies ML techniques, including Random Forests (RF) and Generalized Additive Models (GAMs), as robustness checks. Using four ML evaluation metrics (Mean Squared Error, Root Mean Squared Error, R 2 , and Out-of-Bag Error), the model demonstrated 93.4 % accuracy. These novel findings for small, open economies highlight critical policy implications, emphasizing the strategic prioritization of SDG—7 to foster environmental sustainability and sustainable economic growth.