• This study describes relationships between user’s profile, BRT system’s features and fear of crime (FoC). • A hybrid choice model was used to predict fear of crime at three locations. • Gender and income were the strongest predictors of fear of crime. • Frequent users report higher FoC at bus stops, and sexual harassment victims report higher FoC at buses. • This study provides useful data on the relationship between FoC and user’s travel behavior. Security perception and Fear of Crime (FoC) in urban scenarios have the potential to affect travel behavior, changing people’s travel choices and patterns. In this sense, the feeling of being “safe” or “at-risk” in public transportation not only depends on observable factors like illumination, travel companionship or transport crowding, but also on unobservable individual-specific latent attributes, among which fear of crime constitutes a major issue to consider in transport security policy-making. This study aimed to describe the relationships among sociodemographic features, travel situations, system-design features, and the Fear of Crime at three different locations (buses, bus stops, and stations) of the BRT system of Barranquilla (Colombia). Using an image-based survey applied in Barranquilla, data was collected from a full-sample of 500 adult users (64% females and 36% males) of the system. It was estimated a hybrid choice model to include location-based fear of crime in public transportation perception of risk, along with attributes related to (e.g.) surveillance, crowding, time of the day, and companionship. The relation between fear of crime and perception of security in public transportation is negative and highly significant. FoC inside the bus has the highest negative effect on the utility, followed by bus stops and BRT Stations, for which it remains non-significant. Gender (being a female user) was the strongest FoC predictor. Moreover, it was found that the higher is the user’s income, the lower is the fear of crime reported for each one of these three types of location. Overall, the inclusion of the fear of crime perception in the estimation of the hybrid choice model enhances the model fit. Also, the user-related findings of this study at different locations provide a better understanding of the decision-making process and the predictors of fear of crime in BRT systems.
Read full abstract