Vulnerability game

The resilience of communities to natural disasters depends on mitigation actions, which can be undertaken by public institutions, communities, or individual citizens. However, in disaster-prone areas, trust is often lacking—not only in institutions but also among neighbors, hindering the implementation of collective mitigation strategies. This paper investigates whether trust can be strategically incentivized to support the adoption of group-based mitigation in such contexts. We design a repeated coordination game in which participants first choose the number of players to group with (2, 3, or 4), and then select a mitigation strategy: no action, individual mitigation, or grou p mitigation. Group mitigation offers the lowest cost but only succeeds if all group members independently choose it; its cost decreases as group size increases. Results reveal a shift in participants' preferences over time, with a progressive transition toward larger groups. Participants increasingly abandon individual strategies in favor of group mitigation. Our findings also show that trust is not a stable trait, but a strategic belief that adapts over time. Players’ choices change dynamically in response to observed payoffs, perceived cooperation probabilities, and perceived disaster risk. The game provides insights into when and why cooperation emerges or collapses, allowing for the pre-testing of institutional interventions before they are implemented in real-world contexts.