In this article, we consider wildfire risk management decisions using a dynamic stochastic model of homeowner interaction in a setting where spatial externalities arise. Our central objective is to apply observations from the social science literature about homeowner preferences to this economic externality problem and determine how assumptions about insurance, information and starting fuel loads affect outcomes and the effectiveness of policy. Three new features of our approach are, first, to assess fuel treatment behavior under potential misinformation scenarios, second, to allow for heterogeneous starting fuel loads across ownerships, and, finally, to evaluate the effectiveness of insurance and direct regulation at improving outcomes. Among other results, we find that risk-adjusted insurance may not create incentives for fuel treatment when government suppression exists, and in games with heterogeneous starting fuel loads, the social costs from misinformation can persist over a greater range of fire probability and damage function parameter values. These results suggest that, even as information about wildfire improves, the social costs inherent in private decisions will be more persistent than previously thought on landscapes where fuel stock differs across ownerships. (C) 2013 Elsevier B.V. All rights reserved.
