Economic Determinants of IR2793-80-1 Rice Yield in Bunyala Irrigation Scheme, Kenya: Evidence from Comparative Production Function Modelling

Abstract

Rice productivity in Kenya remains below domestic demand despite continued investment in irrigation and improved varieties. This study analysed the economic determinants of IR2793-80-1 paddy yield among smallholder farmers in Bunyala Irrigation Scheme, Kenya, and examined whether the commonly used Cobb-Douglas production function adequately represented the observed input-output relationships. An explanatory cross-sectional survey design was adopted. A target sample of 333 farmers was determined, 266 questionnaires were administered, and 258 complete responses were retained for analysis. Paddy yield was measured in kilograms per acre; inorganic fertiliser was converted into formulation-adjusted nutrient quantity per acre; farm size was measured as acreage planted with IR2793-80-1; labour as person-days per acre; and productive credit as borrowed funds used directly for rice production per acre. Descriptive statistics and Spearman correlation preceded Ordinary Least Squares estimation of Linear, Cobb-Douglas and centred Translog production functions. The Cobb-Douglas benchmark explained 35.8% of variation in logged yield, whereas the centred Translog model explained 50.9%. The flexible terms significantly improved explanatory power (ΔR² = .151; F-change = 7.484, p < .001), while AIC and BIC also favoured the Translog specification. In the preferred model, land (B = .217, p < .001), labour (B = .151, p < .001) and productive credit (B = .030, p < .001) had positive first-order effects at the mean input combination. Fertiliser and credit exhibited significant positive curvature, land showed diminishing curvature, and the labour-credit interaction was negative and significant. The findings demonstrate that IR2793-80-1 yield is characterised by nonlinear and interdependent input relationships. The study recommends coordinated nutrient management, manageable acreage, timely labour organisation and production-linked credit rather than isolated increases in individual inputs.