DO FOREIGN DIRECT INVESTMENT, RENEWABLE ENERGY CONSUMPTION, ECONOMIC GROWTH, AND TRADE OPENNESS MATTER FOR CO₂ EMISSIONS IN PAKISTAN? EVIDENCE FROM ARDL ANALYSIS

Abstract

Climate change has emerged as a critical global challenge, with carbon dioxide (CO₂) emissions being a primary driver of environmental degradation. For developing economies such as Pakistan, achieving sustainable economic growth while maintaining environmental quality remains a significant policy concern. This study examines the impact of foreign direct investment (FDI), renewable energy consumption (REC), economic growth (GDP), and trade openness (TO) on CO₂ emissions in Pakistan over the period 1990–2023. The study employs the Autoregressive Distributed Lag (ARDL) approach to estimate both short-run and long-run relationships among the variables. Unit root results confirm a mixed order of integration, justifying the application of the ARDL framework, while the bounds testing approach provides evidence of a long-run cointegrating relationship. The empirical findings reveal that renewable energy consumption significantly reduces CO₂ emissions, highlighting its crucial role in environmental sustainability. In contrast, foreign direct investment and trade openness exert positive and significant effects on emissions in the short run but contribute to emission reduction in the long run, supporting the pollution halo hypothesis through technology transfer and efficiency improvements. Economic growth exhibits a positive but statistically insignificant relationship with emissions in the short run. Additionally, the persistence of CO₂ emissions suggests a structural dependence on carbon-intensive activities. The study offers important policy implications, emphasizing the need to promote renewable energy adoption, attract environmentally sustainable FDI, and implement green trade policies to achieve long-term low-carbon development in Pakistan. These findings contribute to the environmental economics literature by providing updated country-specific evidence using a dynamic ARDL framework