Proyección de los indicadores de mortalidad para España

Resumen

This study presents the results obtained by applying the Lee-Carter model and its (two terms and cohort effect) on the probabilities of death for the Spanish population between 1991 and 2010. These models were fit to the data published by the Spanish National Statistical Institute (INE) separately for men and women. Moreover, predicted mortality and probabilities of death were obtained using ARIMA time series between 2011 and 2030. From the comparison of the goodness-of-fit of models we concluded that the Lee-Carter model with cohort effect provides better results than the Lee-Carter model with one or with two terms, therefore it can be used to predict probabilities of death. The comparison shows that, despite the increased complexity of extended models, they can explain past trends and therefore future in a better way. Two indicators of mortality were calculated: Life expectancy and the Gini index. The improvement of the both indicators reflects the increase in the standard of living in the Spanish population in recent decades.