Warrantage and maize productivity in northern Benin: Evidence from a Seemingly Unrelated Regression (SUR) model

Abstract

Abstract

This study examines the relationship between warrantage adoption and maize productivity in northern Benin, with particular attention to access to agricultural inputs as a potential transmission pathway. The study addresses a gap in the literature by jointly examining adoption, input access and yield rather than treating these outcomes separately. A two-stage stratified sample of 250 maize producers was surveyed in Sinendé and N’Dali. Data were collected through structured interviews and focus group discussions and analyzed using a Seemingly Unrelated Regression (SUR) system comprising equations for warrantage adoption, access to agricultural inputs and maize yield. The results show that warrantage adoption is positively associated with access to agricultural advisory services (β = 0.206; p < 0.001) and warrantage-related credit (β = 0.158; p = 0.006), whereas conventional agricultural credit (β = −0.205; p < 0.001) and rainfed farming (β = −0.144; p = 0.045) are negatively associated with adoption. Warrantage adoption is significantly associated with improved access to agricultural inputs (δ = 0.218; p < 0.001), but its direct association with maize yield is not statistically significant (α = 60.714; p = 0.773), and input access is also not significant in the yield equation (α = 177.369; p = 0.437). The findings therefore provide evidence of an input-access effect but do not establish a statistically significant mediation effect on yield. They suggest that the productive contribution of warrantage depends on complementary factors, including the intensity of credit mobilization, education, farm size, intercropping and extension support.

The results show that warrantage adoption is positively associated with gender (β = 0.121; p = 0.029), access to agricultural advisory services (β = 0.206; p < 0.001), and warrantage-related credit (β = 0.158; p = 0.006), while conventional agricultural credit (β = −0.205; p < 0.001) and rainfed farming (β = −0.144; p = 0.045) are negatively associated with adoption. Warrantage adoption is significantly associated with access to agricultural inputs (δ = 0.218; p < 0.001), but neither warrantage adoption nor input access is statistically significant in the maize-yield equation. The findings highlight the importance of integrated agricultural-finance and extension interventions rather than treating warrantage as a stand-alone yield-enhancing instrument.

Keywords: Warrantage; Inventory credit; Maize productivity; Agricultural inputs; Financial inclusion; Seemingly Unrelated Regression; Northern Benin.