An Empirical Analysis of the Effect of Fertility Measurement Choice on Subnational Population Projections: A Case Study of 47 Prefectures in Japan

Masakazu Yamauchi, National Institute of Population and Social Security Research, Japan

In this study, we consider four kinds of fertility measurements—age-specific birth rate, child–woman ratio, general fertility rate, and standardized birth rate—and investigate how the choice of fertility measurement affects the outcomes of subnational population projections. We develop four different models of cohort component projections, which are identical except for the fertility measurement. Actual survival and net migration rates are used to provide projections for the population aged five years and older; the population under five years old is projected by using not only actual survival and net migration rates but also assumed fertility measurement values. We use these models to produce 15-year population projections in five-year age groups for 2000, 2005, and 2010 for each of the 47 prefectures in Japan. This study shows that the choice of fertility measurement does not directly affect the outcome of subnational population projections.