Decisions with weather warnings when waiting is an option

We use laboratory experiments to study decisions with a two-stage weather warning system that varies in the reliability of early warnings. The warning system is based on one used by the Met office since 2011: A risk matrix where the rows and columns represent probability levels and damage levels, respectively. Participants can choose between a safe but costly option and a risky but free option. They can make their decisions based on an unreliable early warning or wait for a more reliable warning, however, waiting makes the safe option more expensive. We find that increasing the reliability of the early warning generally leads participants to wait less and, surprisingly, the reliability of the early warning has a significant impact on the decision they make after waiting and receiving the more reliable warning.