Participation in farmer organizations and adoption of farming technologies among rice farmers in Ghana

Purpose - The authors examine the factors influencing membership in farmer organizations (FO) and their effects on the decision to adopt farm technologies by rice farmers in Ghana. Design/methodology/approach - This study uses a farm survey of 900 households from Northern Ghana and a recursive bivariate probit (RBP) model that accounts for selection bias and endogeneity. Findings - The results indicate that the household head’s decision to adopt machinery and row planting increases by 38.4 and 25.3%, respectively, upon joining a farmer organization. Membership in farmer organization is positively influenced by off-farm income, asset value, farmer organization location and farmer location in Upper West region but negatively by males, age and total livestock units owned. Machinery adoption is positively influenced by membership in farmer organizations and respondent being male but negatively influenced by the years of schooling, farm size, farm distance and location of a farmer in Ghana's Upper East and West regions. Similarly, row planting adoption is positively influenced by membership in farmers' organization but adversely by farm size, farm distance and a farmer's location in Upper East region of Ghana. Research limitations/implications - It can be concluded that membership in farmers' organizations significantly impacts farm household head’s decision to adopt machinery and row planting in rice production, which potentially enhance crop productivity. Practical implications - These results show the importance of agricultural stakeholders in encouraging the formation and strengthening of farmer organizations to support the adoption of modern farming technologies. Originality/value - Developing literature has demonstrated that farmer organizations promote the adoption of agricultural innovations. However, most of these studies have concentrated on conventional agricultural innovations and have used methods that fail to account for potential selection bias. This paper fills this important gap.