This study investigates the impact of agronomic practices and technology adoption on the extent of pre-harvest losses among maize and paddy farming households in Tanzania. Using data from the fifth wave of the 2019/2020 Tanzania National Panel Survey, we estimate pre-harvest loss as the difference between expected and actual production and construct a binary indicator that takes the value 1 if a household experienced crop loss and 0 otherwise. To assess the impact, we employ two estimation techniques: Propensity Score Matching (PSM) and the Generalized Additive Model (GAM). Results from the PSM reveal a negative and significant Average Treatment Effect (ATE), with marginal probabilities indicating that adopters of agricultural technologies have an approximately 15 percent lower predicted probability of experiencing pre-harvest losses compared to non-adopters. In the GAM analysis, intercropping, the use of organic and inorganic fertilizers, herbicides, mechanization, animal traction, and improved seeds are found to influence maize pre-harvest losses, with intercropping being the only practice associated with increased losses. For paddy, the use of organic fertilizer shows a positive association with pre-harvest losses, while animal traction, pesticides, and irrigation are associated with reductions in losses. Based on these findings, it is essential to promote the widespread adoption of effective technologies through targeted extension programs. In addition, training on the appropriate use of these technologies, coupled with government subsidies to support adoption, could play a critical role in reducing pre-harvest losses and enhancing food security.
