[Journal:] Revista de Métodos Cuantitativos para la Economía y la Empresa [ISSN:] 1886-516X [Volume:] 16 [Year:] 2013 [Pages:] 47-67
We model the mortality behavior of the general population in Mexico using data from 1990 to 2009 and compare it to the mortality assumed in the tables used in Mexico for insured lives. We fi a Lee-Carter model, a Renshaw-Haberman model and an Age-Period-Cohort model. The data used are drawn from the Mexican National Institute of Statistics and Geography (INEGI) and the National Population Council (CONAPO). We also fit a Brass-type relational model to compare gaps between general population mortality and the mortality estimates for the insured population used by the National Insurance and Finance Commission in Mexico. As the life tables for insured lives are unisex, i.e. they do not differentiate between men and women, we assume various sex ratios in the mortality tables for insured lives. We compare our results with those obtained for Switzerland and observe very similar outcomes. We emphasize the limitations of the mortality tables used by insurance companies in Mexico. We also discuss the bias incurred when using unisex mortality tables if the proportion of male and female policyholders in an insurance company is not balanced.
mortality rates Lee-Carter longevity dynamics Brass-type model insured population