Characterizing the unhealthy consumer food environment of dollar stores and exploring differences by neighborhood racial composition

Abstract

Objective:

To characterize the overall availability, price, and promotional placement of food and beverage products at dollar stores and explore differences in the food environment by neighborhood racial composition in Atlanta, Georgia USA.

Methods:

A cross-sectional assessment of the food environment was conducted at 25 dollar stores. Measures included availability, affordability, and promotion of fresh produce, salty snacks, sweet snacks, sugar-sweetened beverages (SSBs), and water. Store neighborhoods were categorized as majority-Black (MB) or non-majority Black (NMB) using American Community Survey data. Kruskal Wallis and chi-squared tests were used to test for differences across neighborhood racial composition.

Results:

Only two stores sold fresh produce, whereas all offered and most promoted sweet snacks, salty snacks, and SSBs. Compared to NMB neighborhoods, prices for SSBs and salty snacks were significantly lower in MB neighborhoods (p<0.05).

Conclusions and Implications:

The dollar store food environment lacks fresh produce and is comprised of largely unhealthy food options. Findings suggest dollar stores in MB neighborhoods may provide lower prices for unhealthy food and beverages.

Introduction

Dollar stores, or small box discount establishments that sell assorted inexpensive items continuously at a discounted price, are an increasingly dominant player in the food environment. There are over 35,000 dollar store locations across the United States (US)—more than McDonalds, Starbucks, Target, and Walmart combined.¹ With dollar stores steadily increasing food and beverage offerings in addition to traditional every-day discount items, research suggests US household food expenditures have shifted considerably toward dollar stores over the past decade.²

Dollar stores recently captured the attention of community members and policymakers at the local and national level over concerns about their proliferation, location, and offerings of unhealthy products.^1,3 At least 60 municipalities have enacted local policies against dollar stores ranging from total bans on new dollar stores to requirements about the amount of shelf space dedicated to healthy, fresh food.^1,3 Among them is Atlanta, Georgia which adopted a dollar store dispersal ordinance in 2019 requiring at least 1-mile separation between new and existing dollar stores.^1,3

There is an ongoing debate about whether the ubiquity and location of dollar stores positively or negatively impacts food access.⁴ Dollar stores tend to locate in neighborhoods with low incomes, lower education levels, and in neighborhoods comprised predominately of people of color.^1,4,5 Such neighborhoods have also been historically deprived of healthy food access and full-service grocery stores due to structural racism via discriminatory practices including retailer redlining where retailers opt not to operate in certain areas based on racial/ethnic composition.⁵ Dollar stores may be beneficial as they appear to operate where other food stores have not been successful; however, concerns about their healthfulness, impact on the broader retail food environment (e.g., grocery stores and independent food stores), and contributions to inequities in access to healthy foods persist.^4,6–10

Most research on dollar stores focuses on their access and expansion^2,5,6,9–11 with few studies^12,13 comprehensively examining their in-store food environment, including the availability, price, and promotion of different product categories. Those that have tend to capture only a portion of dollar stores in a given city^13–18 and often focus only on fresh produce or healthy food availability,^15–18 leaving unhealthy products largely uncharacterized. While there is evidence to show that dollar stores tend to locate in majority Black and Latine U.S. neighborhoods,^11,12 no studies have examined whether their targeted location practices come with differences in product availability and pricing depending on neighborhood racial composition. Given these gaps, our study objective was to characterize the overall availability, price, and promotional placement of food and beverage products at dollar stores in Atlanta, Georgia USA and explore differences in the food environment by neighborhood racial composition.

Methods

Dollar store data for this analysis derives from a larger study of convenience and small food stores in the City of Atlanta, Georgia—the 39th largest city in the US with a population of 499,127 in 2022.¹⁴ This population is almost equally split between those who identify as Black or White, and the area is largely physically divided by racial group (Figure 1).^13,14 We focused on store locations within the city limits of Atlanta and used the City’s Business License and Permitting Portal to identify dollar stores within the City. After removing duplicates, 34 dollar stores were deemed eligible for data collection. Of these, 2 were permanently closed, 4 refused data collection, and 3 were not open upon multiple visits (e.g., undergoing store remodel). Thus, the final analytic sample includes 25 stores.

Figure.

Map of included dollar stores in the City of Atlanta by neighborhood racial majority* (n=25).

*Each store was categorized as being located in either a majority Black (MB) (50% or more of the population identify as non-Hispanic Black or African American), or non-majority Black (NMB) (<50% of the population identify as non-Hispanic Black or African American) neighborhood.

The gray lines indicate the 219 individual census tracts (note that some census tracts are very small and may not appear clearly on the map).

We collected data at dollar stores using an observational store assessment tool.¹⁹ Several previous store food environment assessments informed items for the tool.^20–23 Specific food and beverage items included in the tool were finalized through an initial round of pilot testing to update the most common package sizes and brands among for small food stores in the city. Trained staff and graduate research assistants conducted in-store data collection from March-May 2022. Data collectors approached store staff upon entry to explain the study and obtain permission to collect data. Stores were visited between 10:00 AM and 4:00 PM. If a store was closed at the time of visit, data collectors revisited up to 3 times at later dates.

Measures included assessment of the indoor and outdoor appearance, layout of the store, and the food environment (see Supplementary File). Food environment assessments focused on 5 food and beverage categories that ranged in healthfulness: fresh produce, salty packaged snacks, sweet packaged snacks, sugar-sweetened beverages (SSBs) and unsweetened water. For each food and beverage category, availability, price, and promotional placement were recorded. Availability was measured both as the presence of certain categories (yes/no) as well as linear shelf space in feet for each category. We used a tape measure to record the length of the shelf to the nearest inch and then multiplied by the number of fully-stocked shelves with the product category stacked vertically. Empty shelves were not included and partially filled shelves were separately measured and recorded. Price for each item was recorded in US dollars (USD). Promotional placement was measured as whether or not the product could be seen from entrance. Product placement at the front of the store is one of the in-store locations associated with more customer purchasing,^24,25 therefore we use the term “promotional placement” to refer to this marketing strategy throughout.

Each store’s neighborhood racial composition was obtained by mapping the store’s census tract to its 5-year estimates of race/ethnicity from the US Census Bureau’s 2021 American Community Survey (ACS). Race is self-reported in the ACS survey from a list of choices including American Indian and Alaska Native, Asian, Black or African American, Hispanic or Latino, Native Hawaiian and Other Pacific Islander, White, or ‘some other race.’²⁶ Each store was then categorized as being located in either a majority Black (MB) (50% or more of the population identify as non-Hispanic Black or African American), or non-majority Black (NMB) (<50% of the population identify as non-Hispanic Black or African American) neighborhood.

We used descriptive statistics to summarize the availability, price, and promotional placement of products within each of the 5 food and beverage categories. For availability, we calculated the proportion of stores with the product category available as well as the average linear shelf space for each category across stores. Then, we also calculated shelf space ratios (average total healthy product shelf space divided by average total unhealthy product shelf space) and the average price per ounce in USD for each item of the 5 food and beverage categories. Food and beverage items that were largely unavailable across stores (available in <50% of stores), including Nestle water brands (available in n= 2 stores), Deer Park water brands (n= 2 stores), Honey Buns (n=7 stores), and Oreos (n=5 stores), were dropped from the price analyses. Given the small sample size and non-normally distributed data, we used non-parametric Kruskal Wallis and chi-squared tests to assess differences across neighborhood racial composition. All analyses were conducted using Stata 16.²⁷

This study was deemed exempt by the Emory University Institutional Review Board.

Results

Neighborhood and Store Characteristics

There were 100 MB and 119 NMB census tracts, and most dollar stores were located in MB neighborhoods (80%) (Figure). All stores accepted Supplemental Nutrition Assistance Program (SNAP) benefits, and all 3 national dollar store brands were represented (Table 1). No stores accepted Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) benefits, which was verified using the State Department of Health’s WIC authorized vendor list.²⁸

Table 1:

Characteristics of Dollar Stores (n=25) in Atlanta, Georgia

Neighborhood Characteristics	Number of Stores
Neighborhood Racial Composition,* n (%)
Majority Black	20 (80.0)
Non-Majority Black	5 (20.0)
Business Characteristics	Number of Stores
Store Brand, n (%)
Dollar General	9 (36.0)
Family Dollar	11 (44.0)
Dollar Tree	5 (20.0)
Acceptance of SNAP, n (% yes)	25 (100)
Acceptance of WIC, n (% yes)	0
Store Infrastructure	Number of Stores
Refrigeration present, n (% yes)	25 (100)
Freezer present, n (% yes)	22 (88.0)
Prepared food availability, n (% yes)	0 (0)
Food and Beverage Availability	Number of Stores
Availability, n (% yes)
Fresh fruit	2 (8.0)
Fresh vegetables	2 (8.0)
Salty snacks	25 (100)
Sweet snacks	25 (100)
SSBs	25 (100)
Bottled water	25 (100)
Shelf space (linear feet), mean (SD)
Fresh fruit (n=2)	12 (0)
Fresh vegetables (n=2)	16.5 (0.71)
Salty snacks	226.6 (64.15)
Sweet snacks	351.32 (163.36)
SSBs	286.64 (93.63)
Bottled water	39.4 (20.41)
Average shelf space ratio, mean (SD)
Healthy Food:Unhealthy Food† (n=2)	2.3:578.0
Healthy Beverage:Unhealthy Beverage††	39.4:287.0
Food and Beverage Prices	Price/Oz‡
Price/Oz in US Dollars for food groups,‡ mean (SD)
Fresh fruit
Apples (price/pound) (n=2)	$4.95 ($0)
Bananas (price/each) (n=1)	$0.35
Fresh vegetables
Carrot/celery sticks (price/16oz container) (n=1)	$1.75
Vegetable mix (price/8oz container) (n=1)	$4.95
Salty snacks (n=21)	$0.50 ($0.15)
Sweet Snacks (n=16)	$0.63 ($0.32)
SSBs (n=21)	$0.12 ($0.05)
Bottled Water (n=19)	$0.07 ($0.03)
Food and Beverage Promotional Placement	Number of Stores
Items seen from entrance, n (% yes)
Produce	2 (8.0)
Salty snacks	21 (84.0)
Sweet snacks	24 (96.0)
SSBs	20 (80.0)
Bottled water	14 (56.0)

^*The majority Black or non-majority Black neighborhood category was determined for each census tract based on whether 50% or more of the population identified as Black, or less than 50% of the population identified as Black based on the 2021 American Community Survey database.

^†Calculated by adding the average shelf space allocated for fruits and vegetables and the average shelf space allocated for salty and sweet snacks.

^††Calculated by adding the average shelf space allocated for water and the average shelf space allocated for SSBs.

^‡Prices are reported as price in USD per ounce unless otherwise noted.

SNAP indicates Supplemental Nutrition Assistance Program; SSBs, sugar-sweetened beverages; WIC, Special Supplemental Nutrition Program for Women, Infants, and Children

Food and Beverage Availability, Price, and Promotional Placement

Table 1 provides a full description of food and beverage availability, price, and promotional placement. Only 2 stores (8%) sold fresh produce, whereas every store offered sweet snacks, salty snacks, and SSBs. A large amount of store shelf space was dedicated to salty and sweet snacks (average= 578 feet), followed by SSBs (average= 287 feet). The average combined shelf space for unhealthy snacks and SSBs per store is equivalent to more than 1.5 times the length of a football field. Healthy-to-unhealthy product shelf space ratios highlight the disproportionate priority for unhealthy products, as for each foot devoted to bottled water on average there were 7 feet given to SSBs; at the 2 stores that sold produce, for each foot devoted to fresh produce on average there were 200 feet given to sweet and salty snacks.

For prices, water was less expensive than SSBs. In the 2 stores that offered fresh produce, those items were visible upon entering the store. Aside from fresh produce, products that were most commonly visible upon entering the store were sweet (96%) and salty (84%) snacks, while water was the least likely to be visible upon entering (56%).

Differences in the Food Environment by Neighborhood Racial Composition

The USD price per ounce for SSBs and salty snacks were significantly lower in MB compared to NMB neighborhoods (SSBs: 0.11 ± 0.03 vs. 0.19 ± 0.08, p=0.02; salty snacks: 0.48 ± 0.16 vs. 0.57 ± 0.03, p=0.03) (Table 2). There were no other significant differences observed in product availability, price, or promotional placement across neighborhood racial composition. However, there were several notable descriptive differences. Compared to NMB neighborhoods, MB neighborhoods had more shelf space dedicated to sweet snacks (difference= 125.4 feet), less shelf space dedicated to water (difference= −8.0 feet), less shelf space dedicated to SSBs (difference = −36.9 feet), and less front of store water placement.

Table 2:

Differences in Availability, Price, and Promotional Placement of Food and Beverage Categories by Neighborhood Racial Composition (Majority vs. Non-majority Black)

Food Category	Availability (Total Shelf Space (linear feet))			Price (Average Price (USD/Oz))			Promotional Placement (Visibility from Entrance)
	Majority Black† (n=20)	Non-Majority Black† (n=5)	P-Value	Majority Black† (n=20)	Non-Majority Black† (n=5)	P-Value	Majority Black† (n=20) N (%)	Non-Majority Black† (n=5) N (%)	P-Value
Fresh fruit ‡	12 (n=1)	12 (n=1)	--	--	--	--	1 (100)	1 (100)	--
Fresh vegetables ‡	17 (n=1)	16 (n=1)	--	--	--	--	1 (100	1 (100)	--
Salty snacks	227.1 ± 68.6	224.6 ± 48.6	0.73	0.48 ± 0.16	0.57 ± 0.03	0.03	17 (85.0)	4 (80.0)	0.79
Sweet snacks	376.4 ± 161.0	251.0 ± 145.9	0.14	0.64 ± 0.34	0.63 ± 0.20	0.95	19 (95.0)	5 (100)	0.61
SSBs	279.3 ± 85.3	316.2 ± 102.8	0.42	0.11 ± 0.03	0.19 ± 0.08	0.02	16 (80.0)	4 (80.0)	1.00
Water	37.8 ± 19.3	45.8 ± 25.8	0.43	0.07 ± 0.03	0.08 ± 0.02	0.74	10 (50.0)	4 (80.0)	0.23

^†The majority Black or non-majority Black neighborhood category was determined for each census tract based on whether 50% or more of the population identified as Black, or less than 50% of the population identified as Black based on the 2021 American Community Survey database.

^‡Kruskal Wallis and chi-square tests were not performed for fresh fruit and vegetables due to the small sample size (n=2 stores).

SSBs indicates sugar-sweetened beverages; USD, United States Dollar

Discussion

This study characterized the consumer food environment in dollar stores in Atlanta, Georgia, and explored differences by neighborhood racial composition. Our results show that dollar stores offer and promote a range of unhealthy products, and that there may be relevant differences in the food environment between MB and NMB neighborhoods. Among the 5 product categories assessed, SSBs and salty snacks had significantly lower prices in MB neighborhoods, and while not reaching statistical significance, additional descriptive differences suggest more sweet snack availability and less water availability and promotion among MB stores. The findings in this study are timely as US municipalities consider dollar store policies and are important for informing food retail policies that can better support healthy choices among consumers with low incomes.

Unhealthy products were ubiquitous in this sample of dollar stores, which is concerning given their large and growing number of US store locations.^1,5 This was evident as only 2 of the 25 stores had fresh produce, and a large amount of shelf-space was dedicated to SSBs and unhealthy snacks. Of the 2 stores that offered fresh produce, both were Dollar General brand, and may be part of Dollar General’s pilot program which aimed to expand fresh offerings to 5,000 stores by January 2024.²⁹ Retailers have long used strategic product placement to promote certain items and increase sales.^24,25,30 Snacks and SSBs were highly promoted in dollar stores, often placed on a shelf or display visible upon entering the store. Overall our findings align with previous dollar store literature demonstrating that dollar stores rarely sell fresh produce,^12,16,31 and have high availability of unhealthy food options.^12,13

Although we did not observe consistent statistically significant differences across all outcomes—likely due to the small sample size—we found that dollar stores in MB neighborhoods had significantly lower prices for salty snacks and SSBs. Those stores also had more shelf space dedicated to sweet and salty snacks and less shelf space and promotional placement dedicated to water. Dollar stores tend to locate in MB neighborhoods where healthy food access is lower,^1,5,32,33 and thus may contribute both to existing obesogenic food environments and to exacerbating health inequities between racial groups.

Dollar stores within the 3 major brands (Dollar General, Dollar Tree, Family Dollar) have largely similar formats (size, layout, product availability) serviced through their own warehouses, contributing to lower operating costs overall.³⁴ Yet, the variation found in product categories across neighborhoods in this study suggests that dollar stores may have the ability to adjust product availability, store layout, and prices depending on the neighborhood location–and thus may be doing so in a way that amplifies rather than narrows nutrition inequities. As a large player in a food system that contributes to structural racism,^35,36 advocates and policymakers are increasingly calling upon dollar stores to create healthier store environments and promote nutrition equity.^1,7 For example, emerging literature suggests localities could attempt to partner with dollar store corporations to improve existing dollar stores in a way that aligns with neighborhood priorities, such as offering healthy, fresh food options.^34,37 This could be accomplished through store-level interventions and local policies, such as community benefits agreements. In line with equity frameworks for policy making, voices of those most impacted by dollar stores should play a central role in these efforts.^38,39 In fact, several localities have introduced and/or passed community-driven policies that aim to improve dollar stores as indicated in reports and policy documents in cities such as Birmingham, Alabama; New Orleans, Louisiana; Chicago, Illinois; and Baltimore, Maryland.^3,7,40,41

Our findings should be interpreted in the context of study limitations. This exploratory analysis had a small sample size and was limited to 1 geographic region at one time point. Therefore, it was under-powered to examine differences across residential racial composition and results may not be generalizable to other regions. However, our comprehensive in-store assessments of availability, price, and promotional placement characteristics across both healthy and unhealthy product categories help provide initial insight into potential differences in the consumer food environment that could be further examined in future work.

Implications for Research, Practice, and Policy

This study adds to a growing body of literature that supports calls for policies and partnerships to improve the food environment in dollar stores. Policies targeting the healthfulness of dollar store food environments at local, state, and federal levels are crucial for creating lasting changes in the food environment. This is particularly important for health equity given dollar stores’ tendency to locate in neighborhoods of color and low-income neighborhoods, and that we found some unhealthy products were more available and cheaper in MB neighborhoods.^1,3 Future research with a larger sample size may build upon this work by characterizing the dollar store food environment and testing policy interventions in other settings, across different geographical locations (e.g., urban and rural), and neighborhood sociodemographics. Future policy action and interventions to address the unhealthfulness of dollar store food environments are needed to effectively incorporate community voices and ensure equitable solutions are identified.

Of note, all dollar stores in our sample accepted SNAP benefits, while none of the stores accepted WIC benefits. Local-level policies, such as a staple foods ordinance⁴² (which would outline stocking and depth of stock requirements for all food stores) or conditional use ordinances³ (which would require new dollar stores applying to open to meet certain conditions, such as healthy food stocking) could be important tools for policymakers to help encourage dollar stores to stock healthy food items such as fresh produce. Discussions around changes to state and federal nutrition assistance program requirements, such as expanding SNAP retailer stocking requirements, should also continue to be explored to ensure SNAP-accepting retailers like dollar stores improve the stock of healthful food items.^43,44