Principal components analysis of ASFR : application to the recent fertility schedules observed around the world

Mathada Sivamurthy, Applied Statistics Research Centre, Bangalore-560094(India)
Chetna M. Sivamurthy, Applied Statistics Research Centre

Applying Principal Components Analysis to the Five year age group ASFRs (age-specific fertility rates) in the countries around the World for the years around 1960, 1970 and 1980, it was found by Sivamurthy that the first Three Principal Components explained more than 95 percent of the variations in the observed values of the ASFRs. Since the fertility situations around the World have changed greatly during the recent decades, an attempt is made here to apply the same technique to the ASFRs for the years around 2001. It is shown that the first Three components again explained about 92 percent of the variations in ASFRs. Based on this analysis, a revised Three parameter version of Sivamurthy’s PC model to represent the ASFR schedule, is presented. A new method is suggested for fitting the model to a given ASFR schedule. This will be useful for fertility estimation and projection.