Abstract

This study examines whether the contribution of agricultural productive capacities to economic growth in Equatorial Guinea changed as the economy moved from an agrarian base to hydrocarbon dependence and then into the current phase of declining oil production. Using 85 annual observations for 1941–2025, the empirical strategy combines Yeo–Johnson and logit transformations, unit-root tests with structural breaks, multiple-break analysis, a Markov-switching model with regime-dependent coefficients and variances, recursive time variation, and a complementary ARDL specification. The preferred converged two-state capacity model contains agricultural credit, the agricultural budget, logistics infrastructure, and aggregate investment (BIC=435.54; mean maximum smoothed probability=0.961). Investment is positively and strongly associated with growth in both regimes, while the remaining productive-capacity variables display marked state dependence and substantial collinearity. Reviewer-requested robustness work additionally estimates a two-state Student-t specification; that sensitivity fit reaches the iteration limit and is therefore reported as non-converged rather than used for inference. Residual non-normality and the inability to compute exact retransformed marginal effects without the archived Yeo–Johnson lambda parameters remain explicit limitations. The findings support a post-oil diversification strategy based on productivity, value-chain development, financing quality, and productive use of accumulated capital, while remaining associative rather than causal.

This work is licensed under a Creative Commons Attribution 4.0 License.