Abstract
Global warming poses a significant socio-economic threat, with uneven effects across regions and sectors. This study examines whether climate variables have a long-run relationship with agricultural output, evaluates whether their impacts change over time, and whether their impacts vary by crop type in East Africa, using the fixed effects, the Dynamic Fixed Effects (DFE) and Pooled Mean Group (PMG) models. The study revealed that the included climate variables exhibit long-term effects on agricultural output. However, there are notable differences in the impacts of climate variables on agricultural production, influenced by temporal and methodological contexts. Aggregate analysis generally shows limited short-run significance, with temperature having negative and significant effects under the DFE model during 1980–2021. In contrast, the long-run findings reveal a positive effect of rainfall under both DFE and PMG models. A 1% increase in annual rainfall increases agricultural output by 0.76% and 0.27%, respectively, although this effect turns negative in the latter case over the more recent period. The effect of temperature is insignificant except under the FE model. Crop-specific results highlight rainfall as a generally positive factor, while temperature impacts vary, with the short run negatively affecting most crops with elastic effects. The long-term relationship between climate variables and agricultural output, together with the significant effect of irrigation, underscores the need for sustained investment in irrigation infrastructure and broader climate-smart agricultural practices that consider crop-specific responses. Overall, the study emphasizes the importance of climate-resilient agricultural investments and water management policies.