Development of current estimated household data and agent-based simulation of the future population distribution of households in Japan

Abstract

In response to the declining population and aging infrastructure in Japan, local governments are implementing compact city policies in the form of location normalization plans. To optimize the reorganization of urban public infrastructure, it is important to provide detailed and accurate forecasts of the distribution of urban populations and households. However, several local governments do not have the necessary data and forecasting capability. Moreover, current forecasts of gender- and age-based population data only exist at the municipal level, and household data are exclusively available by family type at the prefecture level. Meanwhile, the accuracy is limited with an assumption of the same rate of change of population in all municipalities and within each city. Therefore, the aim of this study was to develop an agent-based microsimulation household transition model, with the household as the unit and agent. Household data was estimated for all cities in Japan from 2015. Estimated household data comprised of family type, house type, and address, age, and gender of household members, obtained from the national census, and building data. The resulting household transition model was used to forecast the attributes of each household every five years. Simulations in Toyama and Shizuoka Prefectures, Japan from 1980 to 2010 provided highly accurate estimates of municipal-level population by age and household volume by family type. The proposed model was further applied to predict the future distribution of disappearing villages and vacant houses in Japan.