Considering developed countries, Italy Japan, Spain and France are believed to be among the most long-lived in the world. This belief is strongly underpinned by period life expectancy statistics. However it has already been suggested that longevity rankings of countries could be dramatically different if a cohort perspective is used. In this paper we reinforce the evidence that cohort approach based ranking of countries is not consistent with the usual one. In order to this, we use an EM algorithm to estimate life expectancy for not yet completed cohorts.
During the considered period all the analysed countries increase the value of their life expectancy at birth, and in 1950, the last year of the study, southern populations recover the initial deficit, but they are still at disadvantage compared to the northern ones and the ultimate trend suggest that not necessarily the gap will be filled.