Climate extremes, soil moisture, and agricultural production in Somalia: evidence from national annual data

Using 50 national annual observations for 1974–2023, this study examines short-run associations between agricultural production, drought severity, excess-rainfall intensity, soil moisture, agricultural land, and rural-population growth in Somalia. Production is measured by the FAOSTAT Agriculture Gross Production Index; climate and soil-water indicators are derived from monthly TerraClimate data. The analysis combines ARDL bounds testing, a mixed-order dynamic regression, robustness checks, bootstrap cointegration testing, and stability diagnostics. Inconclusive bounds evidence precludes long-run and error-correction interpretation. The principal model detects no statistically reliable associations at the 5% level, while nonlinear effects and soil-moisture interactions are unsupported. Confidence intervals and approximate detection thresholds indicate limited precision. Stability tests do not detect parameter instability, and bootstrap testing does not establish cointegration. Supplementary augmented VARs fail the assessed residual-whiteness requirements, so Toda–Yamamoto predictive-direction inference is omitted. The findings provide a cautionary empirical benchmark for this national annual specification. They establish neither the absence of agricultural responses nor aggregation as the cause of insignificant estimates. Seasonal, district-level, and crop-level comparisons remain priorities for future research. This study examines how drought severity, excess rainfall, and soil moisture relate to agricultural production in Somalia using national annual data for 1974–2023. It provides a transparent empirical benchmark by considering climate extremes alongside soil-water conditions and productive capacity. The analysis finds no statistically reliable short-run associations and does not establish a long-run relationship; these findings do not demonstrate that climate risks are unimportant. Its significance lies in clarifying the uncertainty surrounding policy conclusions drawn from these estimates and identifying priorities for stronger agricultural monitoring. Seasonal, district-level, and crop-specific evidence could help inform climate adaptation and food-security planning.