Foreign Direct Investment, Education, and Gender Inequality in Middle-Income Nations: An Empirical Analysis Using a Panel Autoregressive Distributed Lag Approach

Background and Objectives: Gender equality is a cornerstone of the United Nations Sustainable Development Goals (SDG 5). Despite rapid economic transformation, persistent disparities continue to characterize many middle-income nations. Over the past two decades, these economies have become primary recipients of global foreign direct investment (FDI) while achieving significant progress in educational attainment, particularly in female primary enrollment. However, these advancements have not translated proportionately into reduced gender inequality, suggesting deeper structural constraints. This study examines the dynamic and potentially non-linear relationships between FDI, education, and gender inequality in middle-income countries. It specifically investigates whether FDI and education exert diverging short-run and long-run effects and explores the mechanisms through which these factors interact during the transition toward skill-intensive economic structures. Methodology: The analysis utilizes a balanced panel dataset of 18 middle-income countries from 2000 to 2023, totaling 432 annual observations. Gender inequality is measured via the Gender Inequality Index, capturing disparities in reproductive health, empowerment, and labor market participation. Key explanatory variables include FDI inflows (as a percentage of GDP) and female primary school enrollment as a proxy for education, alongside controls for female labor force participation, life expectancy, and political rights. To capture both short-run dynamics and long-run equilibrium, the study employs the Pooled Mean Group (PMG) estimator within a panel Autoregressive Distributed Lag (ARDL) framework. This approach accommodates mixed orders of integration and allows for heterogeneous short-run adjustments while imposing long-run homogeneity. The robustness of the specification is confirmed through unit root and panel cointegration tests, followed by panel Granger causality tests to determine the directionality of the relationships. Key Findings: The results reveal a clear divergence between short-run and long-run effects. In the short run, neither FDI nor education has a statistically significant impact on gender inequality, indicating that structural changes require a longer gestation period. The short-run coefficient for education is positive but insignificant, consistent with an 'opportunity cost' mechanism where increased school enrollment temporarily reduces female labor force participation. Similarly, FDI inflows do not yield immediate effects, reflecting time lags in labor market transmission. Conversely, long-run estimates show that both FDI and education significantly reduce gender inequality. FDI facilitates improved outcomes through formal employment creation and the diffusion of inclusive labor practices, while education enhances human capital and decision-making capacity. Female life expectancy also emerges as a pivotal factor in reducing disparities. Sub-sample analysis indicates that the equalizing impact of FDI is more pronounced in upper-middle-income countries, highlighting the role of institutional quality. Finally, bidirectional Granger causality confirms a mutually reinforcing relationship between gender equality, FDI, and education. Policy Implications: The findings suggest that FDI and education are not immediate policy levers but long-term structural drivers requiring supportive institutional frameworks. First, policymakers should implement targeted interventions to mitigate the short-run opportunity costs of education, such as scholarships and conditional cash transfers for female students. Second, governments should adopt a 'quality-oriented' FDI strategy, prioritizing sectors that promote formal employment and gender-inclusive practices rather than focusing solely on capital volume. Third, sustained investment in women’s health is essential for long-term empowerment and labor market participation. Finally, given the bidirectional