Abstract
This study aimed to identify and explain the possibilities for improving the productivity and profitability of rice by increasing the productive efficiency of rice farmers and reducing yield gap of rice production in Bangladesh. The study was undertaken in three major rice growing areas representing three agro-ecological zones of Bangladesh namely- the old Meghna estuarine floodplain area in Comilla (AEZ-19); the old Brahmaputra floodplain area in Mymensingh (AEZ-9) and the Tista meander floodplain area in Rangpur (AEZ-2). The analysis was mainly based on a set of field level primary data collected from 300 sample farmers from three areas. The Stochastic Frontier production. function analysis was used to determine the technical, allocative and economic efficiency of rice production. Productivity of boro, aman and aus rice varied across areas and among farm categories. Boro rice showed the highest productivity (6198.78 kg) in all farm groups in all areas followed by aman rice (4073.73 kg) and aus rice (1583.66 kg). Net returns per hectare estimated for different rice crops were statistically different from each other. Net return per hectare was significantly higher for boro rice (Tk. 29995.58) followed by aman rice (Tk. 21749.26) and aus rice (Tk. 4536.57). All rice crops were profitable in respect of benefit-cost ratios (full cost basis) which were 1.77, 1.88 and 1.38 for boro, aman and aus respectively in all areas. The maximum likelihood estimates of stochastic frontier production function model showed that plot size, seed, urea, TSP, manure, pesticide cost, seedling age, land type and area (dummy) had a significant positive impact on boro rice production. For aman rice, plot size, urea, TSP, manure, pesticide and crop duration had positive and significant coefficients whereas plot size, TSP, pesticides and land rent had positively significant coefficients for aus rice. For all rice crops, the estimated coefficients of education and extension contact of the farmers had negatively significant effect upon the inefficiency effects. The mean technical efficiency of boro, aman, aus and for all rice crops were 0.85, 0.81, 0.75 and 0.80 respectively. In all areas for all rice crops, the farmers of Rangpur were technically efficient (87 percent) followed by Comilla (84) percent) and Mymensingh (82 percent). Large farmers were technically more efficient (87) than small (82) and medium (80) farmers for all rice crops. Likewise the mean allocative efficiency was estimated at 0.93, 0.91 and 0.89 for boro, aman, and aus rice farmers, respectively. The mean allocative efficiency for all rice crops was 0.91. In the Cobb-Douglas stochastic normalized cost frontier, the coefficients of output, plot size, seed price, TSP price, irrigation cost and area (dummy) were positive and significant at the aggregate level for boro rice. This means that increases in the magnitude of these variables would result in corresponding increases in the cost of production. The coefficient of output, plot size and human labour price were positive and significant for aman rice. The coefficient of output, plot size, TSP price, MP price, S price, pesticide cost, land rent and seedling age were positive and significant in the cost frontier for aus rice. The coefficients of age, household size and extension contact were negatively significant in the inefficiency effect models whereas education, experience and training (dummy) were negative but not significant at the aggregate level for all rice crops. The mean economic efficiencies estimated from Cobb-Douglas stochastic normalized cost frontiers were 79, 74, 67 and 73 percent for boro, aman, aus and all rice crops, respectively. For all rice crops, farmers of Comilla area attained the largest economic efficiency (81) percent) followed by farmers of both Mymensingh and Rangpur area (80 percent). The yield gap-1 was estimated at 1460.68 kg/ha resulting mostly from non-transferable components of technology and environmental factors. Yield gap-II was 1264.66 kg/ha. This was attributed mainly to technical inefficiency. The biotic and abiotic factors were jointly responsible for 27.08 percent yield loss causing a mean yield gap of 1344.18 kg/ha. The study suggests the existence of some gaps in rice yield, which may be reduced through policy interventions and adoption of improved technology. The bio-physical constraints limiting rice production were lack of good quality seed, excessive weed, poor seed germination, poor utilization of irrigation facilities etc, whereas the main socioeconomic constraints were lack of proper knowledge, high price of fertilizer and low price of rice.