Estimating the Stochastic Frontier and Meta-frontier Input Distance for Electricity Distribution Companies in Iran and the Factors Affecting its Efficiency

Analysis of electricity distribution structure and the forces affecting its efficiency as well as electricity industry decomposition and electricity distribution companies independence and privatization have been very important subjects of study in Iran. In this article، Translog stochastic frontier input distance function estimated by Maximum-Likelihood method and data from 36 electricity distribution companies within 1381-1388 period is used to analyze the structure reformation effect on the efficiency of these companies. The results show that translog function specification is preferred to Cub-Dauglas function specification، and met-afrontier model is preferred to frontier models. In terms of circuit density، companies can be classified into two groups: low-circuit-density (group 1) and high-circuit-density (group 2). In the long-run and short-run horizon of electricity distribution companies، no economic perspective is available. Network load factor increase causes efficiency decrease in the first group and، in the long run، efficiency increase in the second group. Increasing the residential consumer’s share from all consumers lead to an efficiency decrease in group 2 but has an insignificant long -run effect on group 1 efficiency. The share of underground network length only has a significant effect on group 1 efficiency and leads to efficiency increase in this group. Increasing the circuit density leads to an increase of efficiency in the long run and short run، which is highly significant. Generally، in the long run، the return of increase to scale is possible and privatization does not have a significant effect on efficiency.