Multinomial Logit Analysis of Factors Affecting the Choice of Enterprise among Small Holder Yam and Cassava Farmers in Niger State, Nigeria


This study analyzed the factors affecting the choice of an enterprise among small holder yam and cassava farmers in Niger State, Nigeria. Data used for the study were obtained from primary source using a multi-stage sampling technique. Structured questionnaire was administered to 150 randomly selected yam and cassava farmers to elicit relevant information from respondents in the study area. Multinomial logit regression model was used to estimate factors influencing the choice of enterprise among small holder yam and cassava farmers in the study area. The findings revealed that most (65.33%) of the farmers chose sole yam enterprise while 4.67% and 30.0% of the farmers chose sole cassava enterprise and yam and cassava as mixed enterprise respectively. The study further revealed the mean of 1.83 tons of combined yam and cassava output of per annum and an average farm size of 2.84ha per farmer, an indication that the study covered small holderfamily managed farm units. The farmers were relatively young and with basic formal education. The multinomial logit regression model showed that income, farm size and output from the chosen enterprise had positive and significant effect on farmer’s choice of an enterprise. This implies that the probability of choosing yam or cassava enterprise increased with income earned from the enterprise, farm size and output from chosen enterprise. The partial elasticiticies of income and output for cassava and combined enterprises were elastic, while all other factors across the groups as classified were inelastic. The study therefore recommended that extension agents should create more awareness on different types, methods and techniques available for yam and cassava cultivation to further improve their adoption. Also, training and farm advisory services on improved management practices to boost yam and cassava production should be given to the farmers.

Journal of Agricultural Sciences 4(1):7-12
Multinomial Logit Regression Choice of an Enterprise