Technical Efficiency of Rice Farming and its Determinants in a Generalized Stochastic Frontier Framework: A Case Study from Assam, India

This study uses a generalized stochastic frontier framework to model a multiple-output rice production system and analyses farm-level technical inefficiency of rice farming in Assam, India. Primary data for this study was collected in 2009-2010 from 310 farm-households in four non-contiguous districts of Dhubri, Morigaon, Dibrugarh, and Cachar that are located in different agro-climatic regions of Assam. Based on a Cobb-Douglas production function for multiple rice varieties, we simultaneously estimate the Generalized Stochastic Production Frontier and examine effects of exogenous factors on farm-level technical inefficiency. We find that the average technical inefficiency of farms is 8.5% in our sample. Further, inefficiency is lower in the frequently flood prone areas, and availability of government support helps reduce such inefficiency as well. However, technical efficiency is higher for the Muslim farm-households, and it decreases with greater land fragmentation. We also find that the use of primitive technology like bullock reduces technical efficiency of rice farming.