Technical Efficiency Analysis of PRIME Branches in Monga Areas of Bangladesh: An Application of Data Envelopment Analysis

This paper evaluates the efficiency of Programmed Initiatives for Monga Eradication (PRIME) branches by using non-parametric Data Envelopment Analysis (DEA) in five districts of North-West region of Bangladesh. The production function approach is used for the assessment of efficiency scores of 149 branches under both constant and variable returns to scale. The results of the study revealed that there were considerable inefficiencies among PRIME branches for the year of 2010 to 2012. In addition, a second stage Tobit regression shows that the variation is also related to branch-specific attributes such as branch age, PRIME to total member ratio, borrower per staff, and location. Since PRIME is an ultra-poor program, it is suggested that achieving higher efficiency might take longer time as old branches were more efficient than new ones. It is, therefore, also suggested that by employing more skilled staff, borrower per staff as well as branches efficiency will be increased in the study areas. However, Kurigram was less scale efficient and Nilphamari was more technically efficient in contrast to Rangpur district. This result implies that for expanding PRIME branches in future, selection of appropriate location will help to achieve higher efficiency.