Application of Robust Data Envelopment Analysis to Estimate Technical Efficiency: A Case Study of Large-Scale Broiler Farms of Sari County in Iran

Uncertainty is a prominent feature of real world issues. Since the change in the input data, though small, can dramatically change the results of the Data Envelopment Analysis (DEA) models, applying ways to deal with uncertainty is an inevitable aspect of the use of these models. Therefore, this research used the combination of robust optimization with DEA to study the technical efficiency of broiler farms with the capacity of over 30,000 broilers in Sari County of Iran. The results of Robust Data Envelopment Analysis (RDEA) model indicated that given the uncertainty level of ε=0.1, as the level of conservatism increased from 0 to 8, the average of efficiency scores decreased from 96 to 93 percent and the number of efficient farms decreased from 23 to 18 units. In addition, the study results showed that the greatest difference between the amount of optimal and actual consumption of inputs related to the cost of taking drugs and vaccines. Given the largest part of drugs and vaccines costs spent when a disease spreads, applying appropriate policies to prevent the spread of epidemic diseases can greatly help to enhance the efficiency of inefficient farms. These policies may include choosing healthy broilers in the same age, using proper nutrition, providing comfortable aviary, choosing proper density of broilers, dissolving the carcasses of ill broilers, using disinfectants properly and also using experts to determine the appropriate time and amount of drugs and vaccines.