Smoothing Age and Spatial Patterns

Abstract

A comparison of an observed pattern of age-specific rates or probabilities with the corresponding model schedule fitted pattern identifies idiosyncrasies in the observed data and points to possible data errors or to irregularities created by an insufficiently large sample. Actuaries calculating life insurance policies or annuities, for example, would want to smooth irregular patterns to ensure that age-specific probabilities of dying, do not show, say, that an average 45-year old female had a higher risk of dying within the next year than did an average 46-year old female. Confronting such an irregularity, an actuary is likely to smooth out the suspicious behavior with a model mortality schedule, for example, the eight-parameter Heligman-Pollard (1980) model mortality schedule.