Insurance decision against forest fire : An econometric analysis combining experimental and real data

Storm and fire are responsible for 70% of total forestry damage due to natural hazards in Europe. They are associated to high costs for forest owners. However, very heterogeneous behaviors in terms of forest insurance demand arise across Europe. For example, only 6% of the French private forest area is insured against fire and/or storms. In this paper, we focus on the private forest owners' insurance decision against fire risk. For that purpose, we collect hypothetical data on willingness-to-pay (WTP hereafter) for insurance through a field experiment in which we test different scenarios of private insurance and public assistance. We also deal with real data about forest owners' insurance decision, characteristics of forests and owners. We simultaneously estimate real insurance coverage and willingness-to-pay for different insurance scenarios, using a selection equation for zero WTP values that we potentially explain by the phenomenon of protest responses against insurance in the expected utility framework. We find that real insurance provision is relevant to explain positive WTP, and we show that unobservable determinant of insurance coverage may explain protest responses. These results confirm the advantage of including observed choices in experiments on insurance demand.