Disasters and impact estimation the great East Japan earthquake and the tsunami in Tohoku

The objective of this paper is to use the synthetic control method (SCM) with an event study framework to estimate the economic effect of the Great East Japanese Earthquake and Tsunami in Tohoku on the Fukushima, Iwate, and Miyagi prefectures in Japan. Using data from 41 prefectures, the SCM was applied to provide a proxy for normal performance in these prefectures in terms of Prefecture Per Capita Income and Gross Prefecture Product. Using these metrics, the Miyagi and Iwate prefectures had positive and statistically significant increases in the three years that followed the event, whereas Fukushima performed in line with our synthetic control estimates. Our results indicate that the impact of the tsunami and earthquake from an economic standpoint was relatively short-lived and this geologic disaster was positively corrected with short- and intermediate-term economic growth in the three prefectures most affected. Our findings provide a contrast to studies on the 1995 earthquake in Kobe, Japan, in which DuPont and Noy (Econ Dev Cult Change 63: 777–812, 2015) illustrated that even massive infusions of resources could not overcome the negative economic impact of that particular geological event over the short- and medium-term. Koshimura et al. (Phys Eng Sci 373: 140–373, 2015) indicated that the tsunami and earthquake of Tohoku led to a paradigm shift in Japan’s disaster management policy, and it seems as though it was effective. The Japanese Government’s plan ‘Act on Development of Tsunami-resilient Communities’ called for a combination of urban planning, housing reconstruction, structural prevention/mitigation, and tsunami disaster mitigation plans all seem to have been effective in accomplishing their goals to support people affected by the event, reconstruct homes and cities, and to revitalize industries and livelihoods. There were some limitations to our study, which include reliance on certain prefectures to construct the synthetic control and whether other areas that were included in the construction of the synthetic control were also impacted by the event. In the study, we targeted the prefectures most affected by the event and conducted robustness checks to ensure that the most appropriate synthetic control was used based on a comparison of prefecture level predictor variables such as the average monthly rental cost, population, and average income.