Investigating the Spatiotemporal Relationship between Neighborhood Demographics and Tobacco Outlet Density

Background

Tobacco smoking and its related morbidity and mortality disproportionally impact communities of color and low-income populations. Residing in an area with a high concentration of tobacco retailers is a health disparity experienced through the over-exposure of dangerous products to specific sociodemographic groups. Living in close proximity to a tobacco outlet has been associated with increased chances of smoking and reduced success when trying to quit. Given this, the spatial distribution of tobacco outlets is important to consider when thinking of how to best support community health.

Previous research has found a link between household income, percentage of African American residents, and percentage of Hispanic residents and a greater density of tobacco outlets; however, studies have equivocated on the relative strength and direction of these relationships. The current study seeks to untangle the relationship between race, ethnicity, income and tobacco outlet density using a longitudinal design.

Methods

Census tract level data were combined from 2 sources at 2 waves, 2010 and 2015: 1.) US Census Bureau provided demographic and income data; 2) NJ Department of the Treasury provided a listing of all cigarette retailers. The 11,578 and 10,918 outlets from 2015 and 2010 respectively were geocoded (99% success rate) and joined to the US Census data at the tract level.

A dynamic spatial autoregressive panel model was used to estimate the relationship between demographics, income, and changes in tobacco outlet density in to control for both time and spatial autocorrelation.

Results

Results from the full model indicated census tracts with lower median income (β=-.115, p<.001) saw greater increases in tobacco outlet density. Both percentage of African American residents and Hispanic residents were not significantly related to tobacco outlet density. However, interaction terms between the demographic and tobacco outlet density showed significant negative effects, high-income, high-African American (β=-.035, p<.05) or Hispanic residents’ (β=-.099, p<.001) census tracts had significantly lower concentrations of tobacco outlets.  Prior tobacco outlet density (β=.646, p<.001) also predicted later density.

Conclusion

Findings here suggest much of the relationship between tobacco outlet density and race or ethnicity is driven by income. Low economic capital is associated with a neighborhood having high concentrations of tobacco outlets. Once a tobacco license is granted, there is a high likelihood that more will locate nearby. Policy approaches to improving the health of communities through reducing exposure to tobacco have shown success through the application of environmental strategies to reduce tobacco availability. Strategies related to licensing may put a cap on the number of licenses, restrict outlets’ proximity to certain establishments, or tie the number of licenses to population size. All of these should ensure outlet density is reduced in low income areas that may have high concentrations of tobacco retail.