This paper generalizes the results of Hausman and Taylor (1981), Schmidt and Sickles (1984), Cornwell, Schmidt and Sickles (1990) and Park and Simar (1992) to the efficient IV estimation of panel models in which the random effects are correlated with a subset of the regressors. The model in which this estimator has particular promise is the stochastic frontier model in which it is posited that inefficiency is correlated with certain characteristics of the determinants of technology, or observable proxies for heterogeneity in the application of that technology, which renders the random components treatment of ethciency inconsistent. In the spirit of Robinson (1988) our semi parametric model assumes It particular form for the frontier production function while considering the joint density of the individual firm-specific effects and those regressors with which they are potentially correlated as unknown. Efficiency of the slope parameters and the asymptotic properties of the level of the frontier function are explored. We illustrate our new estimator in an analysis of productive efficieney between selected European and American airlines after domestic deregulation in the U.S. and prior to recent European reforms implemented ill the course of EC integration.
