Farm efficiency in Bangladesh

Abstract

This thesis examines farm-level efficiency of rice farmers in the High Barind region of Bangladesh by estimating technical, allocative and economic efficiency using farm level cross section survey data. Two contrasting methods for measuring efficiency are applied: the stochastic econometric frontier and Data Envelopment Analysis (DEA). These measures are used to investigate the factors associated with technical, allocative and economic inefficiency. First, technical efficiency is computed by estimating the translog stochastic frontier in which technical inefficiency effects are modelled as a function of socioeconomic, infrastructure and environmental degradation factors in a single stage estimation technique using maximum likelihood method. Technical and scale efficiency are calculated by solving output- and input-oriented constant returns to scale (CRS) and variable returns to scale (VRS) DEA frontiers. A Tobit model is used to evaluate factors associated with technical and scale inefficiency from both input-oriented and output-oriented CRS and VRS frontiers. Same factors are analyzed as in the translog stochastic frontier. The translog stochastic frontier results show that farm households are, on average, 79 per cent technically efficient. The output-oriented DEA frontier results show that the average technical efficiency estimates are 79 and 86 per cent under CRS and VRS assumptions and the average scale efficiency is 92 per cent. The average values for technical efficiency measures and scale efficiency from the input-oriented CRS and VRS frontiers are 79, 85 and 93 per cent respectively. The translog stochastic frontier exhibits decreasing returns to scale, whereas the DEA frontier exhibits decreasing, constant and increasing returns to scale. The technical inefficiency effects model in the translog stochastic frontier and Tobit analysis for DEA frontier show that irrigation infrastructure and environmental degradation are significant factors in determining technical inefficiency. We then measure technical, allocative and economic efficiency by estimating the Cobb- Douglas stochastic frontier following the Kopp and Diewert cost decomposition technique and by running input-oriented CRS and VRS DEA frontier models. We estimate the Tobit model to analyze the factors associated with technical, allocative and economic inefficiency from the DEA frontiers. In addition, we compare the results obtained from both the Cobb- Douglas stochastic frontier and DEA frontiers. The results from the Cobb-Douglas stochastic frontier shows that the average technical, allocative and economic efficiency of farm households are 80, 77, and 61 per cent respectively. The input-oriented CRS frontier results show that farm households have, on average, 86, 91 and 78 per cent technical, allocative and economic efficiency and the corresponding VRS frontier shows that farm households are, on average, 91, 87 and 79 per cent technically, allocatively and economically efficient. An evaluation of factors associated with technical, allocative and economic inefficiency from both the Cobb-Douglas stochastic frontier and DEA frontier reveals that irrigation infrastructure and environmental degradation are the most statistically significant factors affecting technical, allocative and economic inefficiency. This implies that diesel-operated pumps and environmental degradation are not only reducing output from given inputs but are also causing sub-optimal cost-minimizing input decisions. Assessing efficiency suggests that there is a considerable amount of inefficiency among farm households and there is room for enhancing rice production through the improvement of technical, allocative and economic efficiency without resort to technical improvements. Farm households could reduce their variable production costs, on average, between 21 - 31 per cent if they could utilize their inputs in a technically and allocatively efficient manner. An evaluation of factors associated with inefficiency concludes that government electrification programmes which convert diesel pumps into electricity-operated pumps for irrigation in rural areas and policies which lead to reduced environmental degradation would reduce inefficiency, thereby increasing rice production and the welfare of farm households.