Pedestrian and Car Occupant Crash Casualties Over a 9-Year Span of Vision Zero in New York City

Abstract

Vision Zero has been increasingly embraced by jurisdictions across the United States. Existing research primarily focuses on the theoretical principles and effectiveness of specific engineering measures. However, there is limited understanding of the holistic effects of Vision Zero treatments, in the context of street type and urban environment. We developed a street typology framework to categorize street segments using four design and operational features: street width, traffic direction (one- versus two-way), number of travel lanes, and presence of on-street parking. We applied a sample-based partitioning around medoids algorithm to classify 90,327 street segments in New York City. This process results in six distinctive types of street segment. To integrate neighborhood-level factors (e.g., land use variables and sociodemographics), we aggregated street segments of a given street type for each neighborhood. Negative binomial regression models were developed for pedestrian and car occupant crash injuries and fatalities separately for three periods: 2014 to 2016, 2017 to 2019, and 2020 to 2022. Our findings showed that street-segment groups with narrower, two-way sections and greater tree canopy coverage were significantly associated with a lower risk of casualties for both pedestrians and motorized users. Street-segment groups located in neighborhoods with a larger percentage of African American and Hispanic American residents experienced a significantly greater risk of casualties. Vision Zero treatments had mixed effects on safety outcomes. Streets treated with leading pedestrian intervals showed a lower risk of casualties. Neighborhood- and arterial slow zones were associated with a lower risk of car occupant casualties.