The Japanese government supplies post-disaster public housing (PDPH) to victims who are unable to rebuild homes damaged by a disaster. The efficient supply of PDPH requires an accurate prediction of housing demand, based on survey results. However, non-negligible households often change their residential preferences from the initial survey, and these changes make the demand prediction difficult. This study aims to examine the transition of victims’ residential preferences in Mashiki, Japan, post the 2016 Kumamoto earthquake. The study utilizes a three-wave panel data analysis comprising two sets of household survey data and the applicant data for PDPH to investigate disaster victim behavior with greater accuracy and provide a predictive tool for disaster management. The results indicate that households that initially preferred “rebuilding/repairing their homes” or “relocating to another place,” had young householders, did not live in post-disaster temporary housing, and had a high annual income tended to cancel moving into PDPH. These results can be instrumental in predicting the transition of residential preferences in future disasters. This study can establish patterns that, in turn, would help local governments predict the demand of PDPH provided as a form of disaster relief.

日本政府向无法重建受灾破坏房屋的受害者提供灾后公共住房（PDPH）。PDPH的有效供应需要根据调查结果准确预测住房需求。但是，不可忽略的家庭通常会从最初的调查中改变他们的居住偏好，而这些变化使需求预测变得困难。这项研究旨在研究2016年熊本地震后日本Mashiki受害者住所偏好的转变。该研究利用三波面板数据分析（包括两组家庭调查数据和PDPH的申请人数据）来更准确地调查灾难受害者的行为，并为灾难管理提供预测工具。结果表明，最初喜欢“重建/修理房屋”或“搬迁到其他地方”的家庭，有年轻的住户，没有住在灾后的临时住房中，并且年收入较高，倾向于取消搬入PDPH。这些结果可能有助于预测未来灾难中居民偏好的转变。这项研究可以建立模式，进而帮助地方政府预测作为救灾形式提供的PDPH的需求。