Ver artículo

This paper investigates the impact of credit constraints on efficiency of Bangladeshi rice farms in a survey conducted in 2009. The DEA efficiency scores indicate substantial technical, allocative and cost inefficiencies. The bootstrapping method provides farm specific confidence intervals at the 95% significance level and the results show that confidence intervals are large. This finding indicates that differences in average efficiency scores of credit constrained and unconstrained farms are not statistically significant although the average efficiencies are higher in the group of unconstrained farms. To address the sample selection and the binary nature of households’ credit constraint status, we applied a Tobit model with sample selection to consistently evaluate the effects of credit constraints on households’ efficiencies. After effectively correcting for selectivity bias, households’ experience, on-farm training, off–farm income, yearly savings and farm size were the main determinants of inefficiencies. These findings suggest that improving credit access to farms in general but more specifically to the credit-constrained farm households should be targeted to improve farming efficiencies in Bangladesh.