When Disasters Strike, Context Shapes Our Choices

Why It Matters

Disasters don’t just destroy buildings – they reshape how people think about risk. But our choices after catastrophe are not purely rational; they are shaped by the context we live in.

Key Takeaways

• After the 1995 Kobe earthquake, city residents reacted differently to insurance than those in less urban areas.
• People rely on mental shortcuts, with “urban-ness” influencing decisions as much as actual risk.
• Policymakers and insurers must address these biases to close the protection gap and strengthen resilience.

The Challenge of the Protection Gap

When the ground shakes or the sea surges, economic losses mount quickly. In Asia-Pacific, one of the world’s most disaster-prone regions, an alarming 91% of losses from natural disasters are not insured. Japan faces some of the highest seismic risks globally, yet its national earthquake insurance scheme – affordable and backed by government reinsurance – still struggles to attract full participation.

From an economist’s perspective, this is puzzling. If premiums are fair and cover expected losses, demand should be near universal. Yet take-up has remained well below 100% for decades. One explanation lies in how people perceive risk. Even in Japan, where earthquakes are frequent, the public tends to underestimate the likelihood of a major event.

This study investigates how those perceptions shift after catastrophe, specifically how Japanese households responded to the 1995 Great Hanshin-Awaji Earthquake, better known as the Kobe earthquake.

Kobe’s Lessons Beyond Kobe

The Kobe earthquake struck on 17 January 1995, killing more than 6,400 people and levelling much of the city. It was Japan’s first major urban earthquake since the introduction of residential earthquake insurance in 1966.

Surprisingly, the study found that the earthquake spurred higher insurance purchases not only in Kobe itself but also in other major cities such as Tokyo and Nagoya, even though those areas were untouched. In contrast, similar surges were not observed in less urban regions following earthquakes elsewhere.

The implication? It was not just the scale of destruction but its urban setting that shaped behaviour. People in other big cities saw Kobe and thought, this could be us. The “representativeness heuristic” – judging risks by similarity to familiar contexts – drove city dwellers to reassess their vulnerability.

Interestingly, population size mattered more than density. Larger cities experienced bigger shifts, possibly because extensive media coverage and collective memory made the disaster feel closer to home. By contrast, rural communities, which shared fewer similarities with Kobe’s urban landscape, did not adjust their behaviour in the same way.

Heuristics at Work

Traditional economic models assume people act rationally, weighing costs and benefits. But behavioural science shows otherwise. We rely on heuristics – mental shortcuts that simplify decision-making.

Two key heuristics play a role in disaster insurance:

• Availability: people overestimate risks of vivid, recent events.
• Representativeness: people judge risks based on whether a past event resembles their own circumstances.

The Kobe earthquake illustrates representativeness. Because it occurred in a dense urban environment, residents of other cities updated their risk perceptions. By contrast, when major earthquakes struck rural areas, city dwellers barely changed their insurance behaviour – the context felt too different.

The study also ruled out a simple “fear effect” or broad change in risk attitudes. After Kobe, demand for auto insurance actually fell and lottery purchases rose, the opposite of what one would expect if people had simply become more risk-averse. This reinforces the conclusion that context, not generalised fear, was the driving factor.

Beyond Short-Term Reactions

One striking result is that the increase in earthquake insurance did not quickly fade. In many cases, take-up continued rising years after Kobe. This challenges the assumption that post-disaster surges in insurance are temporary spikes driven by fleeting anxiety. Instead, context-based heuristics can help people recalibrate risk more sustainably.

The researchers also employed a new Bayesian method to pinpoint which regions showed the sharpest shifts. Unsurprisingly, Tokyo stood out, despite being hundreds of kilometres from Kobe. Its sheer population size amplified the “urban effect”, with media saturation and social networks reinforcing the lesson that a catastrophe in one major city could happen in another.

Business Implications

For insurers, the findings suggest that designing effective catastrophe protection is not just about pricing policies correctly. People do not respond only to statistics; they respond to context.

This means communication strategies should vary. In big cities, messaging must counter fatalism – the sense that damage will be so vast that insurance is pointless – and instead highlight how collective preparedness and financial protection can reduce chaos. In smaller towns, where people may feel more directly exposed, insurers can emphasise personal security and peace of mind.

Policymakers face a similar challenge. Closing the protection gap requires more than infrastructure investment or public awareness campaigns. Behavioural biases must be recognised. Options include making earthquake insurance the default when people buy fire insurance, offering targeted subsidies for low-income households, and tailoring public messaging to the lived experiences of different communities.

Ultimately, disaster resilience is as much about psychology as economics. Recognising how context shapes choices can help close the gap between available protection and actual coverage.

Authors and Sources

Authors: Kazuhiko Kakamu (Nagoya City University), Shinichi Kamiya (Nanyang Technological University), Petra Staufer-Steinnocher (Vienna University of Economics and Business), Takashi Yamasaki (Kobe University), Noriyoshi Yanase (Keio University)

Original Article: North American Actuarial Journal

---

For more research, click here to return to NBS Knowledge Lab.