Foreign Direct Investment and Economic Growth in Morocco: Revisiting the Evidence with 2SLS

Background and Motivation: Foreign direct investment (FDI) has been widely examined as a potential driver of economic growth, yet empirical evidence for Morocco remains inconclusive due to methodological limitations and endogeneity concerns. This study re-examines the FDI–growth relationship in Morocco over the period 1977–2022 using a five-equation simultaneous system estimated by Two-Stage Least Squares (2SLS). The framework jointly models GDP growth, FDI, exports, human capital, and domestic investment in order to account for bidirectional relationships among the main variables. Methods: Unit root and Johansen cointegration tests support the existence of long-run equilibrium relationships among the series, while a Chow test identifies a significant structural break in 2003 (F = 11.43, p = 0.003). Diagnostic tests confirm instrument relevance (Cragg–Donald F > 10) and fail to reject over-identification validity (Hansen J p > 0.10). Results: The results indicate a positive but statistically fragile association between FDI and economic growth—positive in sign across all specifications but sensitive to sample size and instrument choice (β = 2.179, t = 1.728, p = 0.092; 90% CI: −0.341, 4.699 in constant 2015 USD billions). FDI is associated with growth primarily through indirect channels—particularly export expansion and human capital accumulation—rather than through direct capital deepening alone, consistent with an absorptive-capacity interpretation. The estimated structural break in 2003 reflects a broader package of concurrent institutional and macroeconomic reforms; the model cannot isolate the independent contribution of FDI within this composite effect. The results should therefore be interpreted as evidence of long-run reduced-form associations rather than definitive causal effects. Conclusions: Overall, the study contributes to the Morocco-specific literature by integrating simultaneous equations, indirect transmission channels, and structural break analysis within a unified long-run framework.