A combination of pharmaceutical and non-pharmaceutical interventions as well as social restrictions has been recommended to prevent the spread of coronavirus disease 2019 (COVID-19). Therefore, social contact surveys play an essential role as the basis for more effective measures. This study attempts to explore the fundamental basis of the expansion of COVID-19. Temporal bidirectional causalities between the numbers of newly confirmed COVID-19 cases (NCCC) and individual mobilisations with consumption motives across prefecture borders in three metropolitan regions in Japan were analysed using vector autoregression models. Mobilisation with consumption in pubs from Kanto to Tokai contributed to the spread of COVID-19 in both regions. Meanwhile, causal mobilisation with consumption motives in Kansai also contributed to the expansion of COVID-19; however, the pattern was dependent on the industrial characteristics of each prefecture in Kansai. Furthermore, the number of pub visitors in Kanto immediately decreased when NCCC increased in Kanto. In contrast, the causal mobilisations for the expansion of COVID-19 in the Tokai and Kansai regions were unaffected by the increasing NCCC. These findings partially proved the validity of the conventional governmental measures to suppress pub visitors across prefectural borders. Nevertheless, the individual causal mobilisations with consumption motives that contributed to the increasing COVID-19 cases are not identical nationwide, and thus, regional characteristics should be considered when devising preventive strategies.
Keywords: COVID-19; Japan; consumption; epicentre; individual mobilisation; prefecture.