Efficiency estimation with heteroscedasticity in a panel data model

This paper deals with the estimation of efficiency in a panel data framework. A one-way error component model (ECM) is generalized to accommodate firm-specific variances for the inefficiency component. A multi-step procedure is developed to estimate the cost function and predict increase in cost due to inefficiency. The model is applied to examine efficiency of ten major investor owned electric utilities in Texas during 1966-1985. The production technology is represented by a translog cost function that is assumed to be the same for every utility (except for heterogeneous intercepts), but it is allowed to shift over time due to technical change. The data supports the ECM with firm-specific variances and rejects the homoscedastic ECM. Cost inefficiency is found to vary substantially across utilities ranging from O to 28%. Technical change is found to be quite small (less than ± 1%), especially during 1971-1985.