Estimating Prudence

This study uses an expected utility model of life insurance demand to derive reduced form equations for the Leland-Kimball measures of absolute and relative prudence. The model is estimated for various demographic subgroups using survey data form the University of Michigan's Health and Retirement Study. The estimates of relative prudence range from 1.51 to 5.15, with males at the lower end and females at the higher end of the range. Additionally, prudence is higher among older and married cohorts than among younger and unmarried cohorts, respectively; and smokers, drinkers, and risk-takers appear to be less prudent than their respective counterparts.