COMPARISON OF FRONTIER EFFICIENCY METHODS - AN APPLICATION TO THE US LIFE-INSURANCE INDUSTRY

The objective of this paper is to provide new information on the perfo rmance of efficiency estimation methods by applying a wide range of ec onometric and mathematical programming techniques to a sample of U.S. life insurers. Average efficiencies differ significantly across method s. The efficiency rankings are well-preserved among the econometric me thods; but the rankings are less consistent between the econometric an d mathematical programming methods and between the data envelopment an alysis and free disposal hull techniques. Thus, the choice of estimati on method can have a significant effect on the conclusions of an effic iency study. Most of the insurers in the sample display either increas ing or decreasing returns to scale, and stock and mutual insurers are found to be equally efficient after controlling for firm size.