A multilevel analysis of private-vehicle evacuation clearance times along the US Gulf Coast

This study uses multilevel regression analysis to examine the effect of social characteristics and the built environment on clearance time under an evacuation scenario. The primary unit of analysis is the US Census tract (N = 1660), nested within 31 incorporated places spanning five US states. The dependent variable is an estimate of clearance time in hours derived using network analysis techniques within a geographic information system. We find that tracts with a more peripheral location, more female residents, a higher proportion of Hispanic residents, and higher median household incomes are associated with higher clearance times, on average. Our research suggests the relationship between suburbanization and clearance time is complex and evolving, mediated by past investments in the built environment and shifting social conditions. In addition to facilitating the evacuation of areas with low access to personal vehicles, urban planners and emergency management officials should also consider how the degree of connectivity in the street network impacts congestion and clearance time.