Background:
Residents of high-poverty city neighborhoods have higher rates of obesity and lower rates of physical activity than other urban residents. Standard urban form measures cannot explain these disparities; by these measures, poor urban neighborhoods tend to be more walkable than nonpoor neighborhoods. In our study site of New York City, for instance, poor neighborhoods have higher population density and better access to subway and bus systems and are more likely to mix commercial and residential land use. Explaining this “inner-city paradox” requires health researchers to go beyond these widely-used indicators to consider additional streetscape features that may promote or discourage physical activity among city residents.
Objectives:
This mixed-methods study asks how highly-walkable poor and nonpoor urban neighborhoods differ in environmental conditions that might influence walking or biking. Our conceptual framework, based on Pikora et al. (2002), considers four broad types of factors: functionality, safety, aesthetics, and destinations. Controlling for neighborhood walkability using a standard scale (Frank et al. 2006), we compared conditions in poor and nonpoor census tracts, using a 20% poverty rate as the cut point.
Methods:
The research employs two kinds of evidence. First, we collected extensive information from public and private sources on New York City urban features such as retail and services, parks, street trees, sidewalk cafes, sources of air pollution, and historic districts and landmarked buildings. Second, we conducted systematic observation in a total of 38 blocks (76 block faces), drawn from pairs of poor and nonpoor tracts matched on walkability scores. Observers spent an average of about 75 minutes observing each block face; measures included street and sidewalk infrastructure, pedestrian and vehicle counts, vehicle speeds, and counts or checklists of amenities such as street trees, benches, and public art; disamenities such as overflowing trash cans, foul odors, or excessive noise; and social behavior such as people sitting in groups, standing and talking, or shopping at sidewalk vendors. Observers also recorded their subjective evaluations using an inventory developed by the Project for Public Spaces, a nonprofit organization which collaborated on this research. All neighborhoods studied were in the top quartile of neighborhoods according to our walkability measure.
Results:
Differences in functionality (ease and efficiency of pedestrian or bicycle travel) by neighborhood poverty were minimal, with the exception that infrastructure for bicycle travel (bicycle lanes, bicycle racks) was more common in nonpoor neighborhoods. Few differences related to personal safety and traffic safety were observed. The most consistent differences were in aesthetics and destinations. According to the GIS measures, highly walkable nonpoor census tracts had a higher density of street trees, sidewalk cafes, and basic retail and services. These tracts were also more likely to contain landmarked buildings or neighborhood areas designated as historic districts - designations likely to be associated with interesting architecture. No difference was found in proximity to air pollution sources.
The pattern of findings was similar for the observational measures. Observers’ counts of natural items (e.g. street trees, planters, window boxes) were significantly higher for nonpoor than poor blocks, and decorative architecture and lighting features were also more common. The observational count of disamenities was higher for poor than nonpoor blocks. Subjective ratings of “comfort and image” were also lower for blocks in poor than nonpoor neighborhoods. However, sidewalks in poor neighborhoods were more likely to serve as venues for both social and commercial activity. For instance, people sitting in groups were more often observed in poor neighborhoods. There were significantly more sidewalk vendors, and more people observed to be shopping at sidewalk vendors, in poor than nonpoor neighborhoods.
Conclusions:
This study explored differences between poor and nonpoor urban neighborhoods that appear similar according to standard measures of walkability. Nonpoor neighborhoods tend to be more pleasant venues for pedestrians because they contain more natural features such as street trees, more inviting destinations, and aesthetically more interesting architecture. However, our observational evidence suggests that streets in poor neighborhoods have more settings for social and commercial activity. This difference is likely to reflect differences in income; sidewalks represent lower-cost alternatives to the commercialized spaces within which both social and commercial activity tend to take place in more affluent neighborhoods. Future research relating neighborhood features to pedestrian behavior should incorporate measures of amenities and disamenities. Researchers should also bear in mind that people may use the streets in different ways depending on their socioeconomic status and other life circumstances, and that these different patterns of use may moderate the effects of street features on pedestrian behavior.
Support:
This study was conducted at Columbia University as part of the Institute for Social and Economic Research and Policy’s summer intern program with support from NIEHS grant # 5R01ES014229.
