This paper investigates whether tobacco retailers cluster around schools in New York City. Retail clustering is assessed through two measures: (1) proximity of retailers to schools and (2) density of retailers around schools. Through exploratory spatial statistics and OLS regression, we analyze how retailer density and proximity to schools vary in relation to socio-demographic differences across NYC community districts. As a secondary objective, we test whether patterns of retail clustering differ when assessed by proximity versus density. We identify areas of positive spatial autocorrelation in lower Manhattan and the Bronx, indicating that some NYC community districts have significantly higher density of tobacco retailers around schools than others. The density and proximity of tobacco retailers to schools co-vary with population density, commercial land use and broad indicators of social disadvantage including health insurance coverage. We also find that patterns of retail clustering differ when assessed using either density or proximity, suggesting that land use interventions that restrict tobacco retail proximity to schools might not address retailer density around schools.
Copyright © 2012 Elsevier Ltd. All rights reserved.