Commercial and industrial zoning and neighborhood disparities in billboard exposure: a field-verified statewide census in Rhode Island, USA
Abstract Background. Billboards stand on land that municipal zoning permits, and residents see their advertising without choosing to. Studies in large US cities have found more outdoor advertising in lower-income and minority neighborhoods. Most compared selected neighborhoods rather than every billboard in a jurisdiction, and zoning may account for part of the pattern. We estimated how billboard exposure was distributed across a US state and how much of its association with neighborhood disadvantage was accounted for by commercial and industrial zoning. Methods. Between June and August 2024 we visited all 582 billboard panels listed in Rhode Island by Lamar Advertising and Clear Channel Outdoor, and coded the content of the 507 still standing and visible. We linked panels to 782 census block groups, American Community Survey estimates, statewide municipal zoning and land use maps from 1995 to 2020. Using negative binomial models, we divided the association between neighborhood disadvantage and panel counts into the part accounted for by commercial and industrial zoning and a remainder. We examined municipal clustering, alternative specifications and traffic volume. Because the zoning hypothesis arose during analysis, we treat this decomposition as exploratory. Results. The most disadvantaged fifth of residents received 36.0% of annual billboard impressions (95% CI 18.2 to 57.3). For state environmental justice designation, the pre-specified primary exposure, the total association with panel counts was imprecise (incidence rate ratio 1.82, 95% CI 0.97 to 3.40), although its indirect effect through zoning was positive. Each standard deviation of disadvantage was associated with more panels (1.66, 1.22 to 2.27), and adjustment for zoning reduced this ratio to 1.24 (0.91 to 1.68). Zoning accounted for an estimated 57.8% of this association (24.6 to 160.5), mostly through industrial zoning, and between 27.8% and 61.7% across alternative specifications. Without Providence, which held 157 of the 507 panels, the total association was weaker and imprecise (1.37, 0.95 to 1.97). Content shares differed by under one percentage point between designated and non-designated block groups. Racial and Hispanic composition were not associated with panel counts net of disadvantage and zoning. Conclusions. In Rhode Island, billboard exposure rose with neighborhood disadvantage. Commercial and industrial zoning accounted for between about a quarter and three fifths of that association, depending on specification. Neighborhoods differed in how much billboard advertising they received more than in what it advertised.
Authors
- Marcos Luna (ORCID: https://orcid.org/0000-0002-7866-1891)
- Erica Walker (ORCID: https://orcid.org/0000-0001-5066-1403)
- Sai Venkat Mandalapu (ORCID: https://orcid.org/0009-0001-0180-4145)
- Amol Rathore
- Sage Lefebvre (ORCID: https://orcid.org/0009-0008-3009-1798)
- Huyen Nguyen
- Julia Yakirevich
Institutions
- Brown University (US)
- Yale University (US)
- Salem State University (US)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-13
- DOI
- https://doi.org/10.21203/rs.3.rs-11542124/v1
- Primary Topic
- Environmental Justice and Health Disparities
- Type
- preprint
Funders
- Brown University