DOI: 10.1093/ntr/ntad161 ISSN:

Spatial Clustering of Tobacco Retailers Near U.S. Public Schools

Sarah M Halvorson-Fried, Amanda Y Kong, Heather D’Angelo, Paul L Delamater, Kurt M Ribisl
  • Public Health, Environmental and Occupational Health



Previous studies have found that tobacco retailers cluster near schools. However, all retail outlets may locate near each other and near schools due to existing infrastructure and zoning policies. We assessed whether tobacco retailers cluster near schools in the United States more than expected when accounting for existing retail locations.


We identified 322,056 probable tobacco retailers, 95,110 public schools, and >3.8 million businesses comparable to tobacco retailers in land use and business type. We created 500 simulated tobacco retailer datasets by randomly selecting from the larger list of businesses. For each simulated dataset, we calculated distance from schools to the nearest tobacco retailer (proximity) and count of tobacco retailers within 800 meters of schools (density). Observed proximity and density values were compared to 95% coverage intervals from the 500 simulations. We stratified analyses by urbanicity, percentage of students in the free and reduced-priced lunch program (FRLP), and percentage of Hispanic/Latino, non-Hispanic Black, and non-Hispanic white students.


Tobacco retailers were closer to schools in rural areas, cities, and towns and more dense around schools in rural areas, cities, and suburbs compared to random locations in potential retail space. Schools with more students receiving FRLP had higher density than expected while schools with fewer students receiving FRLP had lower density than expected. Within rural areas, clustering did not vary across sociodemographic group. Within non-rural areas, there were inequities in clustering by racial, ethnic, and socioeconomic school composition.


Tobacco retailers cluster near schools after accounting for existing business patterns. There are inequities in clustering by sociodemographic school composition.


This study provides compelling evidence that tobacco retailers cluster near U.S. public schools and there are racial, ethnic, and socioeconomic inequities in clustering, even when accounting for overall retail location patterns. Given that public schools tend to reflect neighborhood demographics, policies to limit tobacco retailers near schools may reduce both school-based and neighborhood-based inequities

More from our Archive