Fault Determination for Crashes in Vermont

Abstract

Previous research has shown that both foreign and out-of-state drivers are at a safety disadvantage when compared with local drivers. But what defines “local”? This paper examines the effect of distance from home with regard to fault in crashes. With the estimate of the distance between drivers’ homes and crash locations, the effect of increased distance from home on crash fault determination was tested with 5 years of data from the Vermont state crash database. A set of logistic regression models showed that greater distances from home increased the odds of fault for drivers, particularly in single-vehicle crashes. The odds increased further when distance interacted with variables such as road geometry, vehicle ownership, and poor road surface condition compared with distance that was held constant. Increased distance from home alleviated effects of certain factors that have been shown to be hazardous in previous research. Holding distance constant, being a non-owner, and driving on Fridays, Saturdays, Sundays, or during summer increased one's odds of fault; however, as distance increased, the effects of these conditions decreased odds of fault. This phenomenon, although present, was markedly less influential in two-vehicle crashes. Overall, drivers more than 50 mi from home were predicted to have increased odds of fault of about 50% to 200% compared with local drivers.