Agricultural Investment, Climatic Conditions and Food Production In Africa: A Panel Data Analysis of Nonlinear Climate Effects

Abstract

This study addresses the lack of research jointly examining agricultural investment and climate nonlinearities in Africa. It analyzes the relationship between agricultural investment, climatic conditions, and food production using a balanced panel of eleven African countries over the period 1996–2022. The sample comprises Cameroon, Egypt, Ethiopia, Kenya, Mali, Morocco, Nigeria, South Africa, Tanzania, Tunisia, and Uganda, selected according to the continuous availability of data. Agricultural investment is measured by agricultural gross fixed capital formation, while mean temperature and annual precipitation represent climatic conditions. Arable land, agricultural employment, and fertilizer use are included as control variables in the empirical specifications. The empirical analysis primarily relies on a fixed-effects model with Driscoll–Kraay standard errors robust to heteroskedasticity, serial correlation, and cross-sectional dependence. The results show that agricultural investment is positively and significantly associated with food production. In the main specification, a 1% increase in agricultural gross fixed capital formation is associated with an approximately 0.140% increase in the food production index. No robust nonlinear effect is identified for temperature. By contrast, the Lind–Mehlum (2010) test confirms, in the main specification, an inverted U-shaped relationship between precipitation and food production: the relationship is positive up to an estimated threshold of approximately 733 mm per year and becomes negative beyond this level. However, this precipitation result should be interpreted as conditional on the estimator used, because it is not confirmed by the PCSE and FGLS estimates. The findings highlight the importance of productive agricultural investment and water-management policies adapted to national climatic conditions.

JEL Classification : C23, O13, Q18, Q25, Q54.

Paper type : Empirical Research.