The Interdependence of "P" and "E" in the Lpe Methodology : A Caution to Users About Possible Bias

This Paper examines the statistical properties of the LPE (Life-Participation-Employment) approach to computing worklife expectancy. The LPE method requires computing a joint probability of life, labor force participation, and employment; it is offered as an alternative to using worklife tables. While the LPE method has not displaced worklife tables, it has been accepted and adopted by 17% of forensic economists. Implicit in the acceptance of the LPE approach is the assumption that the joint probability of a person being alive, participating in the labor force, and then being employed is equal to the product of the three probabilities. This requires statistical independence, i.e., the probabilities of labor force participation and employment are independent. Tests of this proposition conclude that these probabilities are not independent and can be closely related when looking at disaggregated data. LPE users therefore should be advised that this statistic is not always robust; this raises questions of possible bias when computing the value of economic damages