Measuring economic efficiency of United States life insurance industry: Econometric and methematical programming approaches

Despite their contribution to financial markets and major role in promoting the national welfare, the question of how efficiently life insurance firms have operated in the United States has been unresolved. The main purpose of this dissertation is to enrich the efficiency literature in this important industry. To gauge the economic or X-efficiency of individual firms as well as the whole industry, this study employs two different approaches: the econometric approach and the mathematical programming approach. Within each general approach, a variety of methodologies are applied to the efficiency estimation, encompassing almost all of the models developed in the efficiency literature. This study also tries to overcome the shortcomings of previous studies with new input and output measurement for the insurance industry. The results reveal that the life insurance industry operated at a significantly low level of efficiency for the period 1988-1992. The results also show that there is noticeable dissimilarity in the levels of efficiency for individual firms between these general approaches. On the other hand, the patterns of efficiency rankings of individual firms are well preserved within each approach. The study suggests the expense preference behavior hypothesis is not supported in the life insurance industry. Other interesting findings are that medium size firms operated less efficiently than small and large size firms for the period, and that small firms exhibit increasing returns-to-scale while large size firms display decreasing returns-to-scale. The usual half normal and exponential assumptions for the inefficiency term are not inconsistent with the sample used in this dissertation