Green growth or pollution haven? Investigating the nexus between carbon emissions, FDI, and trade openness based on artificial neural network

Abstract

This paper employs a neural network algorithm to analyze the effects of trade openness and foreign direct investment (FDI) on carbon emissions. The neural network technique effectively captures nonlinear interactions among variables, as seen by the model's high prediction accuracy (MSE = 0.004 and R² = 0.93). The analysis reveals a strong positive relationship between carbon emissions, trade openness, and foreign direct investment (FDI). Significant results show that FDI (0.518573) has a greater impact on carbon emissions than trade openness (0.44). Foreign investment, particularly in industrial and energy-intensive sectors, can significantly increase carbon dioxide emissions through increased energy consumption. Similarly, increased trade openness increases emissions through increased production and transportation, as well as countries' acceleration of production to meet global demand for their products. The study shows that both FDI and trade openness contribute to higher carbon emissions, but FDI is the more significant factor.  Although these economic factors promote growth and globalization, they also pose significant environmental risks if they are not regulated and internationally controlled. The findings underscore the importance of cooperation among countries to establish a regulatory framework and controls that govern foreign direct investment (FDI) and trade activities. To achieve a balance between economic growth and reducing carbon emissions, policymakers should encourage foreign investment in environmentally friendly activities, use energy-efficient technologies, and establish trade regulations that support environmental sustainability.