PCD logo

Measuring and Mapping Alcohol Outlet Environment Density, Clusters, and Racial and Ethnic Disparities in Durham, North Carolina, 2017

Mike Dolan Fliss, PhD, MPS, MSW1,2; Mary E. Cox, MPH2; John W. Wallace, PhD, MSPH3; Matthew C. Simon, MA, GISP3; Kendall B. Knuth, MPH2,4; Scott Proescholdbell, MPH2 (View author affiliations)

Suggested citation for this article: Fliss MD, Cox ME, Wallace JW, Simon MC, Knuth KB, Proescholdbell S. Measuring and Mapping Alcohol Outlet Environment Density, Clusters, and Racial and Ethnic Disparities in Durham, North Carolina, 2017. Prev Chronic Dis 2021;18:210127. DOI: http://dx.doi.org/10.5888/pcd18.210127.

PEER REVIEWED

The figure shows alcohol outlets of Durham County, North Carolina in 2017 with the city of Durham inset at top right. The inset shows historically redlined regions in 1933, established by the Home Owners’ Loan Corporation (HOLC). Measures from maps of present-day alcohol outlets and historic racial divides show similar disparities. The largest off-premises alcohol outlet cluster in 2017 also appears among the most HOLC grade D (ie, “hazardous”) neighborhoods from 1933. Concentrations of non-Hispanic Black residents are also shown.

High-resolution JPG for printimage iconimage icon

Static display of Durham County, North Carolina’s on- and off-premises alcohol outlets in 2017. Historic redlined areas from the same region are inset, established in 1933 by the Home Owners’ Loan Corporation (HOLC). Data sources: 2017 American Community Survey (ACS) 5-year block group estimates, 2013–2017; North Carolina Alcoholic Beverage Control license database. [A text version of this figure is available.]

Top

Background

Excessive alcohol consumption is responsible for more than 95,000 deaths in the US each year (1). Policies limiting high densities of alcohol outlets (places where alcohol can be sold or consumed) can curb excessive alcohol consumption (2). Denser alcohol environments are associated with multiple chronic disease pathways (3), neighborhood-level social effects (4), and increased rates of alcohol-related morbidities and mortalities, such as from motor vehicle crashes, pedestrian injuries, and various types of violence (5).

Community partners in Durham County, North Carolina, explored the alcohol and tobacco exposure environment during community-wide conversations about gentrification and neighborhood change (6), laying the groundwork for our study. Given the community context and the role of racism as a fundamental cause of health disparities (7), we supplemented the community-led analysis of racial and ethnic disparities with measures of spatial access and distance to the nearest alcohol outlet. After conversations with community partners, we contextualized these metrics by using maps to emphasize spatial associations of historic racial disenfranchisement and present-day alcohol outlet clusters.

Using the Centers for Disease Control and Prevention (CDC) publication, Guide for Measuring Alcohol Outlet Density (8), Zhang et al identified outlet clusters in Atlanta, Georgia, and found that reducing outlet density was associated with reductions in violent incidents (9). We expanded on that study by using alcohol outlet density metrics (spatial access index, minimum distance) to describe racial and ethnic disparities in the alcohol outlet environment in the city of Durham, North Carolina. We report on-premises (eg, sit-down restaurant) and off-premises (eg, liquor store or gas station) results separately because off-premises settings carry unique and increased population health harms (10). Durham is the name of both a North Carolina county and the largest city within that county. Hereafter, Durham refers to the county, except when stated otherwise.

Top

Data and Methods

We derived alcohol outlet locations in Durham from the North Carolina Alcoholic Beverage Control license database for 2007 through 2017, representing over 165,000 outlet licenses of more than 60 permit types. Permits were filtered to 38 types of alcohol outlets that sell to individuals, excluding catering, shipping, and wholesalers.

Distances between populations and outlets were calculated by Euclidean (ie, straight line) methods, because of the ease of communication and because it is unclear whether relationships of alcohol outlets and populations operate by driving along roads in a smaller, walkable city. Distance-to-nearest-outlet calculations reached outside of county boundaries to maintain accuracy for residents living near borders (eg, when a Durham resident’s nearest outlet was in a neighboring county). A spatial outlet cluster was defined similarly to that in the Zhang study (9) but with parameters modified for a smaller city context. After comparing the Durham context to Atlanta, we defined clusters on the basis of an overlapping 0.15-mile radius (ie, spatial buffer) around outlets active on January 1, 2017. Clusters were those overlapping areas with at least 5 outlets for both premise types. This method yielded 8 off-premises clusters and 12 on-premises clusters in 2017.

Population data were derived from block group estimates of the 5-year 2017 American Community Survey (https://www.census.gov/programs-surveys/acs). These estimates were distributed into smaller US Census block centroids (center point of shape) for increased accuracy in distance calculations using the 2010, 10-year census block and block group population proportions. Combined race and ethnicity data from the US Census Bureau were used to calculate aggregate demographic-specific outlet density measures. We acknowledge limitations to this approach for measuring disparities, including not having distinct categories for some racial and ethnic groups, such as Hispanic populations including Latino speakers of other languages. Average measures of spatial exposure access (inverse distance-to-nearest 7 outlets) and minimum distance were stratified by race and ethnicity. Home Owners’ Loan Corporation (HOLC) redlining maps and spatial polygons for Durham in the 1930s were gathered from the Richmond University Mapping Inequality project (11) to contextualize disparities.

Statistical analysis was performed in R (12) with some spatial visualization completed in QGIS (13); both are free and open-source software.

Top

Highlights

To measure racial and ethnic disparities in alcohol outlet density, we assessed the transferability of alcohol density research in Durham, North Carolina, by applying techniques previously used in Atlanta, Georgia. Multiple quantitative methods were used to measure distance-to-nearest-outlet, spatial exposure index, and demographics within outlet clusters. Our study contextualizes present-day racial and ethnic disparities with historic disenfranchisement pathways by overlaying 1930s redlining data with alcohol outlet locations and clusters. Results indicate similar disparities in maps of the present-day alcohol outlets and historic racial divides from redlining.

Top

Action

The demographic characteristics of residents living within alcohol clusters differed from the population of Durham County. Non-Hispanic White residents are 42% of the Durham County population but make up 47% of the on-premises clusters and only 26% of the off-premises clusters. In contrast, Hispanic residents are 14% of the county population and 12% of the on-premises cluster population but 21% of the off-premises cluster population. Non-Hispanic Black residents are 38% of the county population and 33% of the on-premises cluster population but 47% of the off-premises cluster populations (Table).

The average distance to the nearest on-premises or off-premises outlet for non-Hispanic Black residents was 0.48 miles, closer than 0.74 miles for non-Hispanic White residents. Hispanic residents’ average distance to outlets was 0.66 miles, closer than 0.76 miles for non-Hispanic White residents. Non-Hispanic Black and Hispanic residents were also exposed to denser off-premises alcohol environments (higher spatial exposure index) than non-Hispanic White residents. Hispanic residents were exposed to denser on-premises environments as well, whereas non-Hispanic White residents experienced a much less dense off-premises environment (Table). Because thresholds of effect or dose–response curves for outlet spatial exposure indexes and health outcomes are not known, it is unclear how much more of a health effect is indicated by an off-premises score of 14 for Hispanic residents, as compared with a score of 13 for non-Hispanic Black residents, or a score of 10 for non-Hispanic White residents. Hispanic residents had the greatest spatial exposure indices for both on-premises and off-premises outlets.

Adverse effects of alcohol uniquely burden communities of color (14). Partial explanations might be multigenerational effects of planning and zoning, resource distribution, neighborhood investment decisions, and access to and use of health care. The effects of historical redlining, a widespread practice in the 1930s of denying home loans and underinvesting in Black neighborhoods, might still contribute to health disparities observed in communities of color. Although evidence is mounting on the relationship of redlining to present-day health disparities, some authors suggest it is understudied (15). During the 1930s and 1940s, banks drew maps of major cities for the purpose of loan planning. Maps were labeled in colors, with some areas grade A (best, traditionally colored in green), grade B (still desirable, colored in blue), grade C (declining, colored in yellow), and grade D (hazardous, in red). We followed this traditional color scheme on our map.

HOLC typically marked Black neighborhoods as grade D. This systematized a process of denying housing loans in Black neighborhoods, which blocked Black residents’ access to capital and equity building through homeownership. This perpetuated a cycle of lower wealth accumulation in Black communities, underinvestment by businesses and government, and racial segregation. Redlining, therefore, could be expected to both lead to a present-day concentration of Black residents (if the area had not been recently gentrified) and unique vulnerabilities to undesirable land use patterns and businesses.

Visual inspection of Durham’s HOLC redlining maps shows the largest off-premises outlet clusters in D-grade areas of primarily Black residents (Figure). Structural racial disenfranchisement from the 1930s is still spatially associated with present-day alcohol outlet clustering in Durham. Historic, local-level maps may help interested community groups and policy makers understand how the now-outdated planning processes might have contributed to racial and ethnic disparities in alcohol environments today. Cluster-specific calculations may be combined with person-centered measures to tell a broader story about alcohol outlet exposure and racial and ethnic disparities.

Although novel behavior-level interventions exist (16,17), the Community Preventive Services Task Force (18) has recommended evidence-based interventions, including increasing both the price of alcohol and dram shop (eg, bars, pubs, taverns) liability laws, that may be applied at the population level. Population-level surveillance of alcohol outlet density may inform local interest groups, policy makers, and others interested in preventing excessive alcohol use and racial and ethnic disparities related to the availability and accessibility of alcohol. Combining maps and measures of present-day disparity and historical spatial associations may promote deeper, more meaningful discussion among groups. Redlining map archives of many large US cities are freely available online (11).

We encourage researchers not only to calculate overall alcohol outlet density but also to stratify density and disparity calculations by on-premises restaurant districts and off-premises settings. If these settings are combined, calculations may unintentionally hide meaningful disparities in outlet density exposures that may contribute to subsequent disparities in health outcomes.

Top

Acknowledgments

This work was supported by the Centers for Disease Control and Prevention, cooperative agreement no. 1NU38OT000297-01-00 and the North Carolina Division of Public Health, Overdose Data to Action, cooperative agreement no. 5 NU17CE925024-02-00. Epidemiologists from the North Carolina Division of Public Health were members of the study team and helped contextualize study results. Authors from the University of North Carolina at Chapel Hill were funded through a North Carolina Division of Public Health subcontract under Contract no. 5118396. This work does not represent official positions of the North Carolina Division of Public Health or the Centers for Disease Control and Prevention.

We thank our community partners at Together for Resilient Youth and our colleagues at North Carolina Alcoholic Beverage and Control for data sharing. No copyrighted materials or tools were used in this report.

Top

Author Information

Corresponding Author: Mike Dolan Fliss, PhD, 170 Rosenau Hall, CB #7400, 135 Dauer Dr, Chapel Hill, NC 27599‑7400. Telephone: 919-699-559. Email: fliss@email.unc.edu.

Author Affiliations: 1University of North Carolina Injury Prevention Research Center, Chapel Hill, North Carolina. 2North Carolina Division of Public Health, Injury and Violence Prevention Branch, Raleigh, North Carolina. 3North Carolina Institute for Public Health, University of North Carolina, Chapel Hill, North Carolina. 4Council of State and Territorial Epidemiologists, Applied Epidemiology Fellowship, Atlanta, Georgia.

Top

References

  1. Esser MB, Sherk A, Liu Y, Naimi TS, Stockwell T, Stahre M, et al. Deaths and years of potential life lost from excessive alcohol use — United States, 2011–2015. MMWR Morb Mortal Wkly Rep 2020;69(30):981–7. CrossRef PubMed
  2. Campbell CA, Hahn RA, Elder R, Brewer R, Chattopadhyay S, Fielding J, et al. ; Task Force on Community Preventive Services. The effectiveness of limiting alcohol outlet density as a means of reducing excessive alcohol consumption and alcohol-related harms. Am J Prev Med 2009;37(6):556–69. CrossRef PubMed
  3. Stahre M, Roeber J, Kanny D, Brewer RD, Zhang X. Contribution of excessive alcohol consumption to deaths and years of potential life lost in the United States. Prev Chronic Dis 2014;11:E109. CrossRef PubMed
  4. Cambron C, Kosterman R, Rhew IC, Catalano RF, Guttmannova K, Hawkins JD. An examination of alcohol use disorder symptoms and neighborhood disorganization from age 21 to 39. Am J Community Psychol 2017;60(1-2):267–78. CrossRef PubMed
  5. Snowden AJ, Hockin S, Pridemore WA. The neighborhood-level association between alcohol outlet density and female criminal victimization rates. J Interpers Violence 2020;35(15-16):2639–62. CrossRef PubMed
  6. Boone WJ. Regulating outlets that sell alcohol in predominantly African American and Hispanic neighborhoods. In: Giesbrecht N, Bosma LM, editors. Preventing alcohol-related problems: evidence and community-based initiatives. Online: APHA; 2017:397–408.
  7. Phelan JC, Link BG. Is racism a fundamental cause of inequalities in health? Annu Rev Sociol 2015;41:311–30. CrossRef
  8. Centers for Disease Control and Prevention. Guide for measuring alcohol outlet density. 2017. https://www.cdc.gov/alcohol/pdfs/CDC-Guide-for-Measuring-Alcohol-Outlet-Density.pdf. Accessed November 30, 2020.
  9. Zhang X, Hatcher B, Clarkson L, Holt J, Bagchi S, Kanny D, et al. Changes in density of on-premises alcohol outlets and impact on violent crime, Atlanta, Georgia, 1997–2007. Prev Chronic Dis 2015;12:E84. CrossRef PubMed
  10. Nesoff ED, Milam AJ, Branas CC, Martins SS, Knowlton AR, Furr-Holden DM. Alcohol outlets, neighborhood retail environments, and pedestrian injury risk. Alcohol Clin Exp Res 2018;42(10):1979–87. CrossRef PubMed
  11. Nelson RK, Winling L, Marciano R. Mapping inequality: redlining in New Deal America. https://dsl.richmond.edu/panorama/redlining/. Accessed March 23, 2021.
  12. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/. Accessed March 23, 2021.
  13. QGIS Development Team. Get Started Using QGIS. Open Source. https://www.qgis.org/en/site/forusers/index.html. Accessed March 23, 2021.
  14. Glass JE, Rathouz PJ, Gattis M, Joo YS, Nelson JC, Williams EC. Intersections of poverty, race/ethnicity, and sex: alcohol consumption and adverse outcomes in the United States. Soc Psychiatry Psychiatr Epidemiol 2017;52(5):515–24. CrossRef PubMed
  15. Gee GC. A multilevel analysis of the relationship between institutional and individual racial discrimination and health status. Am J Public Health 2008;98(9 Supplement 1):S48–56. CrossRef PubMed
  16. Al-Hamdani M, Smith SM. Alcohol health-warning labels: promises and challenges. J Public Health (Oxf) 2017;39(1):3–5. CrossRef PubMed
  17. Ross CS, Brewer RD, Jernigan DH. The potential impact of a “no-buy” list on youth exposure to alcohol advertising on cable television. J Stud Alcohol Drugs 2016;77(1):7–16. CrossRef PubMed
  18. Task Force on Community Preventive Services. Recommendations for reducing excessive alcohol consumption and alcohol-related harms by limiting alcohol outlet density. Am J Prev Med 2009;37(6):570–1. CrossRef PubMed

Top

Table

Return to your place in the text

Table. Cluster and County Demographics, Spatial Exposure Index, and Distance-to-Nearest-Alcohol-Outlet in Durham County, North Carolina, 2017a
Factor On-Premises Off-Premises Durham County
Cluster demographics, no. (%)
Non-Hispanic White 3,939 (46.9) 2,316 (26.3) 112,697 (42.1)
Non-Hispanic Black 2,724 (32.5) 4,164 (47.3) 36,077 (37.5)
Hispanic 1,008 (12.0) 1,874 (21.3) 100,260 (13.5)
Totalb 8,391 (100) 8,794 (100) 267,587 (100)
Distance to nearest outlet, miles
Non-Hispanic Whitec 0.76 (0.74–0.78) 0.74 (0.72–0.76)
Non-Hispanic Blackc 0.66 (0.64–0.69) 0.54 (0.52–0.56)
Hispanicc 0.59 (0.56–0.63) 0.48 (0.45–0.51)
Spatial exposure index, miles
White non-Hispanic 11.6 (11.2–12.0) 10.1 (9.8–10.4)
Non-Hispanic Black 11.1 (10.7–11.5) 12.6 (12.3–12.9)
Hispanic 12.9 (12.2–13.6) 13.9 (13.4–14.5)

a Demographic data derived from block group estimates of the 2017 5-year American Community Survey (https://www.census.gov/programs-surveys/acs).
b Total includes other races and ethnicities besides the 3 listed, so total counts and percentages do not add to 100.
c Generalized linear model estimates (95% CI) are outside the premises-specific range.

Top

Error processing SSI file

The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors’ affiliated institutions.