identifying the factors explaining technical efficiency in iran's banks in a multilevel / mixed effects model

this study examines technical efficiency in iran's banking system using unbalanced panel data for 17 state and private banks during the 1997-2009 period. therefore, technical efficiency of banks was measured through estimating a translog cost function in the form of a panel sfa error compound model. then, the effect of underlying factors including both individual and structural (ownership/activity) variables on banks' technical inefficiency, was measured in a multilevel / mixed effects model. the results showed that in spite of the improvement in state banks' efficiency within time, it showed a decreasing trend compared with that of private banks. furthermore, the results of multilevel/mixed effects model showed that 57 percent of the difference among banks were explained with respect to owenership and 15 percent were explained by activity and obligatory loans, respectively. finally, it was shown that automation is the most importance factor in improving banks' efficiency. jel classification: c12, c23, c24, c87