Long-term business impacts of disasters on firm sales: a longitudinal structural equation modeling analysis using six years of panel data following the 2016 Kumamoto Earthquake

Understanding long-term economic recovery from major earthquakes requires tracking aggregate outcomes and identifying the evolving mechanisms through which disaster damage constrains business performance. However, previous studies have predominantly relied on macro-level indicators or cross-sectional evidence, neglecting the dynamic pathways from earthquake damage to business outcomes. This study proposes a unique longitudinal structural equation modeling approach to quantify and decompose the multi-year impact of the 2016 Kumamoto Earthquake on business-level sales. Using repeated business surveys for 2016 to 2021, we constructed a longitudinal dataset of 513 businesses operating in directly affected areas. The estimated standardized total effect of earthquake-related physical damage on business-level sales decreased over time, from -0.293 in 2016 to -0.115 in 2021. The dominant constraint on sales shifted from supply disruptions in 2016 to an earthquake-related decrease in demand in subsequent years. The results also suggest that business continuity planning mitigates immediate post-disaster supply disruptions. Additionally, some businesses with substantial post-earthquake capacity losses may have entered the COVID-19 period before fully recovering their sales, leaving behind residual earthquake-related effects on sales. By characterizing how the dominant constraints on business performance evolve over time, these findings present a useful analytical basis for designing more adaptive post-disaster recovery strategies for future large-scale disasters.