Households' perceptions of earthquake risks and protective measure adoption intentions

Thesis (Ph.D.)--University of Washington, 2021Preparedness levels vary among individuals who face common threats of natural hazards in the region in which they live. The studies in this dissertation explore the factors that may explain such differences, examining the relationships between key predictors of protective action decisions and protective behavior intentions in the context of Western Washington. Specifically, this dissertation aims to answer the following research question: How do Western Washington residents understand earthquake risks and form intentions to adopt protective measures? Chapter 2 investigates the effect of perceived efficacy of seismic reinforcement on people’s intentions to pay for such protective measures using the Meta-Cognitive Model (MCM) approach. Using original data collected in King County, WA, the study examines the association between perceived efficacy and intended payments for seismic retrofits jointly with one’s metacognitive confidence in their assessment of efficacy using interaction effects. The hypothesized moderating effect of metacognitive confidence is not confirmed, but perceived efficacy is found to be an important determinant of intentions to pay for retrofits. Chapter 3 provides an international comparison of perceptions of earthquake risk, efficacy of earthquake early warning (EEW), and willingness to pay (WTP) for EEW between the city of Sendai, Japan and Seattle, WA that have similarities and differences in terms of earthquake hazards and availability of EEW. Risk perceptions and perceived effectiveness of EEW in personal protection are found to be significant determinants of WTP for EEW in both cities, while the association of fear with earthquakes is a significant predictor of WTP only in Seattle. In addition, this study uses a double hurdle model to examine if people’s WTP is a two-stage decision that involves (a) an assessment of the determinants of the decision to pay nothing versus anything at all, and (b) among those who decide to pay something, an assessment of the factors that are associated with how much one is willing to pay. The study finds that the set of variables predicting whether one decides to pay or not does not predict the amount of payment. This suggests that indeed, two sets of decisions may be involved. Finally, Chapter 4 uses a choice experiment to investigate (a) how the key predictors of protective action affect the probability of one choosing to adopt a resource-intensive protection measure over taking no action, and (b) whether residents of Western Washington are willing to pay for resource-intensive measures. Perceived efficacy of each protective measure is found to be positively associated with the probability of one choosing the respective measure over taking no protective action. However, residents of King County are generally not willing to pay for resource-intensive measures, suggesting a need for policy to increase uptake of protective action in the region