This study analyzes the profit change of the top 100 US life in surers by applyinga newly developed mathematical programming method. The profit change is decomposed into a productivity effect(technical change and oper ating efficiencyeffects), an activity effect(product mix, input mix, and scale effects), and a priceeffect. The results indicate that for the period 1996-2000, the positive profit changehas been mainly due to scale and output mix effects, while the productivity change has affected the profit change negatively. We found noticeable difference in various effects between stock and mutual life insurers and that the expense preference hypothesis is not supported in the sample. The results also suggest that the agency system has a positive output mix effect, while the direct writing system shows positive input mix effect on the profit change.

Profit Change Decomposition , Life Insurance , Mathematical Programming Approach