How Residents Perceive Neighborhood Scale: An Examination of Individual and Contextual Factors
Purpose: Neighborhood is a social and geographic concept that plays an increasingly important role in research and practice that address disparities in health and well-being of populations. However, most studies of neighborhoods as well as community initiatives geared toward neighborhood improvement make simplifying assumptions about boundaries, often relying on census geography to operationalize the neighborhood units. This study was intended in part to evaluate differences in scale between commonly used artificial neighborhood units (typically census blocks or tracts) and the scale of neighborhood as actually perceived by residents. In addition, the study examined the individual and neighborhood contextual factors that affect perceived neighborhood scale.
Method: This study used data from Annie E. Casey Foundation’s Making Connections (MC) program. The Making Connections program takes place in selected target areas of 10 cites (Denver, Des Moines, Hartford, Indianapolis, Louisville, Milwaukee, Oakland, Providence, San Antonio, and Seattle/White Center). The samples for Making Connections survey were designed to give equal probabilities of selection to all households within each target area. A total of 6,224 households completed in-person interviews and a map-drawing exercise. Neighborhood scale was measured by having respondents draw maps of their neighborhoods during the survey interview. These paper maps were digitized with GIS tools. Individual variables included demographic and socioeconomic characteristics as well as measures of longevity within the neighborhood, homeownership status, neighborhood participation, and fear of crime. Measures of residential context included street connectivity, mixed land use, the presence of physical disorder, residents’ collective efficacy, poverty rate, population density, and the presence of vacant and multifamily housing units. Multilevel modeling was used to estimate a two level model in order to examine both individual differences and contextual effects on perceptions of neighborhood scale.
Results: The median resident map size was approximately 30 percent smaller than the median census tract, but 25 percent of residents viewed their neighborhood as quite small (less than one-fifth of the typical census tract). Multi-level modeling showed significant within context variation in perceived neighborhood scale. Longer term residents with higher education and income and who were more engaged in the neighborhood held more expansive views. But there were also contextual influences with higher density and mixed use areas associated with smaller perceived neighborhoods, and collective efficacy contributing to more expansive neighborhood definitions.
Implications: Much of the research and community practice tends to rely on artificially imposed neighborhood units. These artificial units may misrepresent resident experience, but GIS tools can be used to craft more authentic neighborhood definitions for research and practice.