Modeling dependence of accident-related outcomes using pair copula constructions for discrete data

Abstract

This paper investigates the relationship between accident-related outcomes and per capita income, and explores the interdependency between them by using vine pair copula constructions. Equations for number of accidents, number of fatalities, and number of people injured are estimated using a provincial level data of Thailand in 2011. We discovered that there exists an inverted U-shaped relationship between accident-related outcomes and per capita income. Moreover, it was found that the accident-injury pair had stronger concordance and tail dependence, whereas the accident-fatality and fatality-injury pairs had weaker concordance and tail dependence. Our findings provide useful insight and information to policymakers who can then use the same to select appropriate road safety measures. © Springer International Publishing Switzerland 2014.