Effect Of Built Environment On Traffic Fatalities In Indian And US Cities

The main aim of this study is to model the correlation between built environment parameters and traffic fatalities in Indian and US cities. In the past models, built environment parameters have usually been studied along with traffic parameters. Macro-level safety models with focus on the built environment have also been attempted for the first time. Lastly, comparison of models across geographical locations (India and USA) has also been made. To test the framework in India, the cities of Agra (a high crash-rate city) and Ludhiana (a low crash-rate city) have been analyzed. For comparison, the US cities chosen were Chicago (a high crash-rate city) and Baltimore (a low crash-rate city). Two Negative Binomial regression models have been developed for each city. The first involves population data. The entire city is divided into grids (based on latitude/longitude) and the number of traffic fatalities in each grid is correlated with the lengths of different types of roads and the number of intersections in that grid. The second model involves survey data. Number of fatalities in small buffer regions is correlated with 16 other built environment parameters in these buffer regions. Results showed that lengths of highways and arterial roads were significant in increasing the expected number of fatalities in Indian cities. For the cities in USA, the length of highways in a region was found to be insignificant. This strengthens the argument that highway design is a major contrasting element between Indian and US cities. In the second model, pavement quality showed varying impacts in Indian and US roads mostly due to unsafe designs in India whereas road markings showed similar and intuitive effects in both India and the USA.