AGGREGATION ERRORS IN COHORT-CHANGE RATE METHOD

Abstract

The cohort-component method is widely used for population projection of local governments in Japan. The purpose of the paper is to examine aggregation errors caused by the cohort-component method, which can account for age distribution. First, we show that the populations estimated at aggregated level frequently are less than those at disaggregated level by using municipality data in Ibaraki Prefecture. Second, we prove this tendency through a simple mathematical model using Lexis diagram by focusing on internal migration.