[2Q6-GS-11] AI and Society:

[2Q6-GS-11-02] Estimation of Population Growth and Decline Ratios Using Population Distribution Shapes

〇Kent Takai¹, Jumpei Hirai¹, Shiho Saito¹, Yuki Yamagishi^1,2,3, Shota Ura³, Keisuke Moriyama³, Ichiro Uwafuji⁴ (1. Shizuoka Institute of Science and Technology, 2. RIKEN, 3. Shizuoka City, 4. Shizuoka University)

Keywords:Population Decline, Cohort Component Method, Population Estimation

The representative method for population estimation is the cohort-component method, which is also used in Japan for future population projections. Fundamentally, in addition to the population by gender and age, it requires “age-specific fertility rates and sex ratio at birth,” “survival rates by gender and age,” and “international migration rates by gender and age.” Generally, it is difficult to adapt to the Resident Registration System data that is publicly available from each municipality. Nonetheless, the Resident Registration System is compiled more frequently than national censuses and often provides data by single year of age, making it a crucial source of information for inferring a city’s economic growth, i.e., the factors of population increase or decrease. Therefore, this study attempts to estimate the ratios of increase and decrease within the framework of log-linear regression, similar to the future mortality rate prediction model (Lee-Carter model), using the population distribution by single year of age and the shape of that distribution, namely, the moments in the probability distribution of age.