Determinants of Foreign Direct Investments in Selected MENA Countries: A Country-Specific Approach using Machine Learning Techniques

This study investigates the country-specific determinants of foreign direct investment (FDI) inflows in seven MENA countries using a machine learning framework. Unlike traditional cross-country studies that impose the same set of variables and coefficients for all countries, our approach uses CatBoost and SHAP to identify the most influential predictors for each country individually from 66 economic, demographic, institutional and trade indicators. The results reveal substantial heterogeneity in the drivers of FDI across countries, highlighting that variables positively affecting FDI in one country may have negative or negligible effects in another. These insights provide policymakers with targeted guidance for attracting FDI by identifying the country-specific factors that matter most.