Vegetable crops have high farm gate values per unit land and contain the essential micronutrients, vitamins, phytochemicals and antioxidants critical for ensuring balanced diets and enhancing human nutrition. Vegetables are a major source of farm income for smallholders and thus contribute to improving livelihood status of smallholders in many developing countries. High postharvest losses of between 30-80% have been recorded in the vegetable sub-sector of most sub-Saharan African countries due to the highly perishable nature of produce and poor pre-and postharvest handling practices, resulting in reduced quality of produce and loss of net revenues for growers and other actors in the value chain. To this end, participation in targeted formal training programs on postharvest handling and practices of horticultural crops is considered a critical determinant of success to improve farmers’ awareness and knowledge of improved practices, and in turn increase revenue from vegetable crop sales.  This paper identifies the factors influencing farmers’ decision to participate in postharvest training programs for horticultural crops, and the implications of their participation on the revenue generated from vegetable production and marketing in Tanzania. Finally this study provides lessons for policy makers who work on Orange Flesh Sweetpotato (OFSP) in East African countries.  The study is based on a stratified random sample survey of 240 households in three major vegetable producing regions in Tanzania: Arusha, Tanga and Morogoro. 

The study assessed the casual linkages among factors that influence farmers’ decision to participate in a postharvest training program based on discrete choice models (i.e., Logit and Probit models) measured through Maximum Likelihood Estimation approaches.  Discrete choice models theoretically or empirically model choices made by farmers among a finite set of alternatives such that farmers make decision to participate in the training program or otherwise. 

First, factors influencing farmers’ participation decision were identified based on discrete choice models. Second, the revenue from vegetable crop sales was modeled as a sequential decision since farmers first decide to choose a treatment (participation in a training program on postharvest handling of horticultural crops or otherwise), which then endogenously impacts their revenue from vegetable crop sales. The decision to participate in the training program is endogenous in the sense that there might be some unobserved characteristics that influence both farmers’ participation decisions and vegetable crop income. Consequently, ordinary least-squares regression estimates cannot be used to identify the average treatment effect, as a result of which an alternative identification strategy should be employed. To control for endogeneity in the sample selection bias to the minimum possible, we subjected the data to a standard treatment effect as well as to Instrumental Variable regression and Heckman Selection Correction models.

The results suggest that farmers’ decisions to participate in postharvest training programs for horticultural crops significantly and positively impact their vegetable crop income by 1.6 times, along with other factors such as level of farmers’ education measured as number of years studied, farm size, type of recommended postharvest handling practices (i.e., precooling, value addition, storage facilities), sources of produce buyers at the farmgate level such as village collectors and wholesalers as well as regional effects.  Farmers’ decisions to participate in training programs are positively influenced by the following factors: (i) visits by government extension agents, (ii) information dissemination through radio, mobile phones, and printed media, and (iii) farmers’ attitudes toward adoption of new agricultural technologies. The results show that a literate female farmer living within a household with large family size tend to significantly participate in a postharvest training program.  Overall, the results find a robust and positive effect on household vegetable crop income from farmers’ participation in postharvest training programs, suggesting that there is considerable scope for improving the livelihood status of smallholder farmers through increased knowledge and skills acquired in recommended postharvest loss reduction practices.