Socio-economic analysis of market participation among small holder apple (Malus domestica) farmers in three key agro-ecological zones of Uganda

This study investigates the determinants of market participation among smallholder apple farmers in Uganda, focusing on the role of household income and education. Apple production in Uganda begun in the Kigezi highlands and later expanded to Mt. Rwenzori and Mt. Elgon highland regions offering potential for income diversification, food and nutritional security. Previous research focuses on profitability of the enterprise rather than market participation. Moreover, literature reveals that market participation increases household income. However, it remains unclear whether income, in turn, promotes market participation. Similarly, the influence of education on market participation is debated, especially when measured solely by years of schooling. This study was designed to respond to such propositions through a cross-sectional household survey using semi-structured questionnaire, conducted among 270 randomly selected apple farmers in the three key highland agroecological zones of Uganda. Exploratory data analysis (EDA) and an instrumental variable (IV) Probit model were used to analyse the data. Results show that 88.9% of the farmers market their apple fruits in local wet markets. Findings reveal that market participation is positively influenced by gender (favoring males), yield (P = 0.000), farmer group membership (P = 0.022), market assurance (P = 0.026), and extension support (P = 0.000). Conversely, higher household income and elevated education levels negatively influence participation at 1% and 10% significant levels respectively. The study recommends strengthening of farmer group marketing and partnerships to facilitate market assurance and access. We also advocate for integrated vocational and agribusiness training into higher education to align education with practical market engagements. This study is however limited to use of cross-sectional data and one instrumental variable which renders interpretation of results as association relationships rather than definitive causal effects. Future research should explore use of panel data with inclusion of several instruments to enable definitive causal statistical inferences. The intensity and frequency of farmers’ market participation should also be explicitly studied using Heckman or double hurdle models.