Investigating intended evacuation from wildfires in the wildland-urban interface: Application of a bivariate probit model

With evidence of increasing wildfire risks in wildland-urban interface zones in the U.S. West and elsewhere, understanding intended evacuation behavior is a growing issue for community planners. This research investigates intended evacuation behavior due to wildfire risks, using mail survey data collected from over 1000 households in the East Mountain area outside Albuquerque, New Mexico (USA). Respondents were asked whether they would evacuate under both voluntary and mandatory evacuation orders. Bivariate probit probability models are used to jointly investigate the subjective belief structure of whether or not the respondent is concerned about wildfire risk, and the intended probability of evacuating as a function of risk perception, and a variety of socioeconomic and demographic variables (e.g. gender, political affiliation, length of residence, owning animals and amenity ratings).