Spatial distribution patterns of the elderly ratio and their association with the density of the elderly population

Abstract

Using the elderly ratio as the sole variable for the analysis of the spatial distribution pattern of the aged population has the danger of underestimating the seriousness of the aging problem, particularly in urban areas that usually have a larger elderly population than rural areas but a low elderly ratio due to the large size of the base population. To improve the understanding of the spatial distribution of the aging population, a choropleth map was first constructed with respect to the elderly ratio and the density of the elderly population to highlight any distinctive features in their spatial distribution patterns and the relationship between them. We then investigated the spatial correlation between the two indices to classify the individual areas into different categories of aging society using the bivariate local Moran statistic. The results highlighted some aspects of aging society that would have been difficult to detect using only the elderly ratio. These include the existence of neighborhoods with high elderly ratios and high densities of elderly population in the urban center, and high elderly ratios and low densities in rural communities. Additionally, the presence of neighborhoods with low elderly ratios and high densities of elderly population surrounding the urban center were detected, which were difficult to identify using only choropleth maps. Finally, this paper discusses the impact of the modifiable areal unit problem on the analysis and concluded that the data aggregated at the basic grid square (1-kmxl-km) scale would be suitable for the analysis of aging society conducted at the prefectural level.