A proposed method for fitting the mortality pattern in Mexico 2000, 2005 and 2010

We present a methodology based on Cubic Splines to fit the general mortality curve in Mexico by sex at all ages for the years of 2000, 2005 and 2010. This method consists in fitting a polynomial of degree three in the first four ages (fitting them perfectly); between ages 4 and 80 we fit a cubic spline using five nodes; we fit another third-degree polynomial beginning at age 80. We employ 24 parameters to estimate the mortality curve in the full age range. This model presents an excellent goodness of fit in men, which increases over the years; for women, the goodness of fit is, in general, better than men’s. This methodology is considered to be a relatively parsimonious estimation of mortality; it maintains the mortality curve’s changes in concavity; and it allows the realization of mortality projections in different contexts and causes of death.