This paper analysed the relationship between input use and yield of major crops grown in Kerala using spatial autoregressive stochastic frontier analysis (SAR-SFA). The data on yield of rice (three varieties) coconut, pepper and the respective inputs used in their productions in three agro-climatic zones (valley, coastal and hill) were used to study the spatial stochastic efficiency. The analysis revealed significant contribution of seeds on yield of autumn and winter rice in valley, while in the coastal region, hired human labour and manure/fertiliser contributed to higher yield of autumn rice and coconut. In hill district, all the inputs contributed significantly to the yield of summer rice despite differences in their individual contribution. The input inefficiencies were relatively higher for winter and summer rice in the valley and for coconut in coastal and hilly areas. The spatial analysis also revealed significant spatial autocorrelation, for autumn and winter rice depicting influence of neighbouring districts on the productivity.
Kumar et al. (Wed,) studied this question.