This article aims to assess the gender-differentiated effect of farmer training on cocoa farm yields in Greater South Cameroon, using both bivariate probit and propensity score matching (PSM) approaches.
Using data from the 2019 survey conducted by the National Cocoa and Coffee Board (NCCB) among 13,601 cocoa farmers, the study estimates a bivariate probit model to address endogeneity, complemented by PSM.
The results suggest that agricultural training is positively associate with the probability of achieving a high yield (≥ 500 kg/ha) by 15.54 percentage points. However, this effect is stronger for conditional probability for women (12.0 points) than for men (7.5 points) and for joint probability for women (6.9 points) than for men (4.8 points). Motorcycle ownership also has a positive effect, whereas primary education level, intensive use of organic fertilizers and production of Grade 2 cocoa are associated with a lower probability of reaching high yields.
Limitations include the cross-sectional nature of the data, potential measurement error from self-reported yields, and restricted external validity beyond the study region.
This study revisits the relationship between agricultural training and cocoa farmers’ yields in the Greater South of Cameroon where only 32% of producers have benefited from technical support. Its originality lies in the use of a large administrative dataset, a joint estimation strategy (bivariate probit and PSM), and an explicit gender-differentiated analysis.
