Restricting Tobacco Retail Outlets Around Middle and High Schools as a Way to Reduce Tobacco Marketing Disparities: A Simulation Study

Abstract

Introduction

Communities with more people of color and economically disadvantaged residents are disproportionately exposed to tobacco marketing from tobacco companies. This study examined if banning tobacco retail outlets (TROs) within 1000 ft of schools would reduce these marketing disparities through a greater reduction in the amount of tobacco advertising around schools in these communities.

Methods

Data from objectively audited advertisement data from 106 convenience stores and gas stations around 42 middle and high schools located in the four major metropolitan areas of Texas were linked with schools’ enrollment data. ArcGIS (Aeronautical Reconnaissance Coverage Geographic Information System) was used to simulate a 1000-ft ban of tobacco sales around the schools. Independent sample T-tests and Mann–Whitney U tests were used to test mean differences where appropriate.

Results

Schools with a higher enrollment of Hispanic/Latino (mean = 171.6, SD = 96.9) and economically disadvantaged students (mean = 168.9, SD = 102.3) were surrounded with significantly greater advertising at TROs than schools with lower enrollment of these groups (mean = 82.8, SD = 49.1 and mean = 89.2, SD = 50.6, respectively). A simulated 1000 ft ban of TROs around schools led to greater advertising reduction around schools with a higher enrollment of Hispanic/Latino students (13.3%–29.4% reductions) in comparison to schools with lower Hispanic/Latino student enrollment. However, the more economically disadvantaged schools had a smaller reduction in the number of advertisements (5.9%–21.9% reductions) in comparison to schools with less economically disadvantaged students.

Conclusion

The implementation of a ban of tobacco sales at TROs within 1000 ft of schools is one policy approach to reduce youth exposure to tobacco marketing, particularly among students of color.

Study Implication

Tobacco retail outlets (TROs) around schools with a higher enrollment of Hispanic/Latino and economically disadvantaged students had significantly more tobacco advertisements in comparison to schools with lower enrollment of these student groups. A simulated ban of TROs within 1000 ft of schools led to greater advertising reduction around schools with a higher enrollment of Hispanic/Latino students. For schools with more economically disadvantaged students, the ban led to a smaller reduction in advertisements in comparison to schools with less economically disadvantaged students. This proposed place-based strategy could be a successful means to reduce tobacco advertising and marketing disparity among communities of color.

Background

Following the ban of cigarette advertisements on TV, radio, billboards, transit ads, and brand-sponsored youth focused events, retail locations are now the primary channel of advertisements for tobacco companies in the United States.¹ Through this advertising channel, tobacco companies target economically disadvantaged and communities of color with more tobacco retail outlets (TROs) and products. In their study conducted at the census tract level in Erie County, New York, Hyland and colleagues reported greater numbers of tobacco retailers on roadways in the lowest income quartile compared with the highest quartile, as well as in quartiles with the greatest number of African Americans (vs. lowest).² A nationwide assessment of all 64 909 census tracts in the continental United States showed similar findings; tracts with greater proportions of residents who were Black, Hispanic, and women with lower education levels had a higher number of TRO density.³

Tobacco companies also target economically disadvantaged and communities of color with a disproportionate amount of tobacco marketing.^4–6 A study of 84 randomly selected neighborhoods in the Omaha Metropolitan Area of Nebraska showed that an increase of $10 000 in median household income was associated with a 14.3% reduction in the number of tobacco marketing items per square mile in a neighborhood.⁵ A nationwide study of 2230 tobacco retailers in the United States found that stores in neighborhoods with the highest (vs. lowest) concentration of African American residents had twice the odds of displaying a price promotion advertisment.⁶ These patterns are not confined to the United States, as research in other developed countries like Australia and Canada show that areas with more economically disadvantaged individuals have more tobacco retailers and promotions.^7–9 Differences in the type of products marketed in communities have also been observed. Research has shown that residents in predominantly white and higher income neighborhoods are exposed to more advertisements of non-combusted, potentially lower risk products, while communities of color and economically disadvantaged neighborhoods are bombarded with advertisements for combusted tobacco products.^10,11 Taken together, these studies show that disparities exist in the number and density of tobacco retailers and tobacco marketing across communities and in the types of tobacco products advertised.

Youth are also a prime target of tobacco companies.^12,13 Documents obtained from tobacco companies show that they target convenience stores, grocery stores and other vendors near schools and playgrounds when marketing their products.^12,13 Tobacco companies provide more advertisements and display spots for their products in stores where adolescents often visit.¹⁴ Approximately 64% of US adolescents report seeing cigarette advertisements all or most of the time when they visit convenience stores, supermarkets, and gas stations.¹⁵ Research has shown that the closer retailers are to public schools, the greater the probability of displaying exterior advertisements.¹⁶ A study of selected middle and high schools in Texas showed that TROs within 1000 ft of the schools had significantly more tobacco advertisements around them in comparison to TROs that are farther away from the schools.¹⁷ Moreover, disparities also exist in the brand and type of tobacco marketed at retail outlets around school neighborhoods.¹⁸ Youth and Black students are often targeted with more menthol cigarette advertising than other racial/ethnic groups.^19,20 A 2005–2006 study conducted among 156 schools in California shows that the percentage of menthol advertising increased by 5.9 percentage points for every 10 percentage point increase in the proportion of Black students.¹⁸ Therefore, not only are students and school neighborhoods the target of tobacco companies, students of color are faced with more tailored and branded tobacco advertisements as well.

The targeting of neighborhoods surrounding schools is especially concerning as youth are particularly vulnerable to tobacco advertising, as it shapes their attitudes and encourages tobacco initiation.²¹ Studies have shown that students who attend schools located in neighborhoods with a high density of tobacco outlets were more likely to be current smokers and more likely to make purchase attempts.^22,23 A school-based longitudinal study of adolescents (ages 11–14 years) who had never smoked showed that the incidence of initiation was 29% among those who visited convenience, liquor or small grocery stores at least twice a week while it was only 9% among those that visited less than twice per month.²⁴ Given that students are vulnerable to tobacco marketing,^23,24 more studies and interventions aimed at reducing student’s exposure to tobacco advertising are needed.

One of the proposed policy solutions to reduce exposure, as well as disparities in the location and marketing of tobacco products, is to ban the sales of tobacco at retail outlets within 1000 ft of schools.^25,26 Studies have found that banning tobacco product sales near schools is effective in bringing greater reduction in retailer density in lower income and racially diverse neighborhoods than in higher income and white neighborhoods.^26,27 Our study goes further by exploring the impact of these bans on tobacco advertisement reduction around schools with a greater percentage of economically disadvantaged students and Hispanic/Latino (hereafter referred to as Hispanic) students. We examined the amount of tobacco retail marketing at TROs around schools in major metropolitan areas of Texas by percentage enrollment of Hispanic and economically disadvantaged students to understand how a ban of tobacco sales within 1000 ft of schools would impact disparities in tobacco marketing. Using data from TRO audits, we objectively assessed if schools with a greater enrollment of Hispanic students and economically disadvantaged students had significantly more tobacco advertisements when compared with schools with lower enrollment of these groups. We then simulated a 1000 ft ban of TROs around schools to determine if this type of ban would lead to a greater reduction in tobacco marketing at schools with greater enrollment of Hispanic students and economically disadvantaged students in comparison to schools with lower enrollment of these student groups.

Methods

Study Design and Setting

Objective advertisement and price promotion data were collected at TROs within a half mile (i.e., 2640 ft; 804.7 m from the centroid) of middle and high schools that participated in the Texas Adolescent Tobacco and Marketing Surveillance System (TATAMS). TATAMS is a longitudinal surveillance study designed to examine the impact of tobacco marketing on tobacco use behaviors among adolescents attending public, charter, and private schools in the four largest metropolitan areas (Houston, Dallas/Fort Worth, San Antonio, and Austin) in Texas.²⁸ The design of the TATAMS study is described elsewhere.²⁸ Seventy-nine schools were initially recruited to participate in TATAMS and during the TRO data collection effort, an additional 25 schools were added to account for the transition of students from middle to high schools expected in this longitudinal study.

The addresses of the TROs were obtained from the Texas Comptroller’s Office in 2014 and the outlets included in this study were audited for tobacco marketing from January to April 2017. The distances of the TROs from the schools were determined via ArcGIS, using projection NAD 1983 Texas Centric Mapping System Albers. The half mile distance is thought to be a realistic walking distance and has been used in previous studies.^23,27 Data collectors were trained on how to identify tobacco marketing and to conduct the audit using the FileMaker Go application, hosted on iPod touches.²⁹ The protocol and data collection tool for this study were adapted from previous research^30,31 and informed by previous studies.^24,32 Data collectors were required to complete at least five practice audits and their data were reviewed by a trained research assistant who provided extensive, detailed feedback. After completing at least two practice audits with no corrections required, data collectors began the TRO audits. The TRO audit tool captured information on all outdoor and indoor advertising and price promotions (hereafter referred to as advertising) for cigarettes, e-cigarettes, cigarillos/little cigars, and smokeless tobacco, including whether the advertising was for a flavored or non-flavored product. In the event that an audit was not possible, the data collector indicated why the audit could not be complete (e.g., “No, store does not exist” to “No, asked to leave before completing survey). Only stores with completed audits were included in this study. To assess inter-rater reliability, a sub-sample of 56 TROs were assigned to two different data collectors. Of the 56 assigned, 33 had complete data and an average of 88% agreement across all measures collected at the store.

The full audit included 261 stores: 89 (42.6%) were located in Austin, 65 (31.1%) in Houston, 31(14.8%) in Dallas, and 24 (11.5%) in San Antonio. The audit included drug/pharmacy stores, grocery stores, wine/liquor stores, tobacco/smoke shops, and convenience stores with or without gas stations. However, the present study included only tobacco advertisement data from convenience stores and gas stations located around schools with available enrollment information. Convenience stores with/without gas stations were selected because of the high prevalence of marketing in these stores (84.2% of the tobacco advertisements were found in these stores) compared to drug stores, pharmacies, and grocery stores and the frequency of youth visits to these types of stores.³³ After excluding TROs that were not convenience stores with/without gas stations (n = 104) and convenience stores with/without gas stations with incomplete audits (n = 25), we also excluded TROs with incomplete/incorrect addresses that could not be geocoded (n = 2) and TROs around schools without available student enrollment information (n = 24), which left a final sample size of 106 TROs around 42 middle and high schools included in this study.

This study was reviewed and approved (HSC-SPH-19-0313) by the University of Texas Health Science Center’s Committee for Protection of Human Subjects.

Measures

School Enrollment Data

The percentage enrollments of Hispanic and economically disadvantaged students for schools that participated in the TATAMS study were obtained from the 2016 to 2017 Texas Education Agency (TEA) school enrollment data.³⁴ School-level economic disadvantage is a measure of the percentage of students at the school who are eligible for free meals, reduced-price meals, or who have other economic disadvantage. The median values for the percentage enrollments of Hispanic and economically disadvantaged students were obtained and used as a cut-point. The schools were grouped into (a) >median percentage enrollment or (b) ≤median percentage enrollment for each variable. For schools in our study, the median percentage enrollments for Hispanic and economically disadvantaged students were 65.5% and 75.0%, respectively. Previous research has utilized similar cut-points.³⁵

Tobacco Advertisement Data

At each of the TROs, data collectors documented information on all indoor advertisements, outdoor advertisements, and flavor of tobacco advertisements (menthol flavor, non-menthol flavor, and non-flavored) for all product types (cigarettes, cigars, cigarillos, smokeless tobacco, and e-cigarettes). For the advertisements with multiple tobacco products, each of the products was counted separately. The documented advertisements in each of the TROs were grouped by location of the advertisement (outdoor and indoor), by tobacco product (cigarettes, cigar products, smokeless tobacco, and e-cigarettes), and by flavor (non-flavored including tobacco flavor or non-flavored e-liquid, menthol, and flavored non-menthol). Then, these were summed to obtain 10 advertisement measures/variables at each TRO: (1) outdoor advertisements, (2) indoor advertisements, (3) cigarette advertisements, (4) e-cigarette advertisements, (5) cigar advertisements, (6) smokeless tobacco advertisements, (7) menthol flavored tobacco product advertisements, (8) non-menthol flavored tobacco product advertisements, (9) non-flavored tobacco product advertisements, and (10) all tobacco product advertisements. The TROs were assigned to the schools around which they were audited. The 10 advertisement variables at each TRO were summed for each school.

Data Analysis

Geocoding

To geocode the schools and TROs, the TIGER/Line Shape files for counties were obtained from the U.S. Census Bureau department of Commerce public file³⁶ and imported into ArcMap vs 10.2.2. Address locators for the school and TRO addresses were created using ArcMap vs 10.2.2. An address locator essentially turns the textual descriptions of the addresses into geographic features. State plane projected coordinate system was used. Schools and TROs were mapped as point features. In order to accommodate the distance between the school centroid and boundary, an 1150 ft buffer was created around the school point address instead. This 1150 ft buffer distance essentially represents an approximate 1000 ft ban and has been utilized in previous studies.^26,27 Retail outlets that sell tobacco were overlaid over the school map for each county.

Analysis: Observed Mean Difference in the Number of Advertisements by School Enrolment

The 10 advertisement variables were first transformed by taking their square roots. Independent two sample T-tests (t-test)³⁷ and Mann–Whitney U test (for variables that were not normally distributed) were then used to determine if there was a significant difference in the mean of each of the 10 advertisement variables around schools with greater enrollment of Hispanic students when compared with schools with a lower enrolment of Hispanics, representing the differences based on the data collected. This same procedure was followed to test for differences by economic disadvantage. This mean difference will be referred to as the observed mean difference. The “proximity tool set” and “multiple buffer analysis” procedures on ArcGIS were then used to remove TROs within 1000 ft of schools (i.e., 1150 ft buffer^26,27) to represent the ban on TROs.

Analysis: Simulated Mean Difference in the Number of Advertisements by School Enrollment after Simulation of a Ban

The means of the advertisement variables were also evaluated for these groups after implementation of the simulated 1000 ft ban (Tables 2 and 3), this difference will be referred to as the simulated mean difference. A two-sided type I error rate of 0.05 was utilized for all analyses. The assumptions of normality (using Shapiro-Wilks test), and equality of variance (using Levene’s test) were verified for the transformed advertisement variables before the test.^37,38 For variables that were not normally distributed before (outdoor, e-cigarettes, and flavored tobacco advertisements) and after the ban (outdoor, e-cigarettes, flavored tobacco advertisements, cigars, and smokeless tobacco), the non-parametric Mann–Whitney U test³⁹ was used in conducting the mean difference test. The percentage reduction for all the advertisements was obtained and reported for the school enrollment groups (Table 4).

Table 2.

Observed Mean Differences in Tobacco Advertising by Percentage Enrollment of Hispanic and Economically Disadvantaged Students (n = 42 Schools; n = 106 TROs)

Advertisements	Observed mean differences in the amount of tobacco advertisements
	By Hispanic Students Enrollment			By economically disadvantaged students enrollment
	≤65 % enrolment (n = 21)	>65 % enrolment (n = 21)	T-test or Mann–Whitney U test	≤75 % enrolment (n = 22)	>75 % enrolment (n = 20)	T-test or Mann–Whitney U test
	Mean (Std.)	Mean (Std.)	P-value	Mean (Std.)	Mean (Std.)	P-value
Total Ads	82.8 (49.1)	171.6 (96.9)	0.001	89.2 (50.6)	168.9 (102.3)	0.003
Indoor	79.1 (47.9)	163.7 (90.5)	0.001	85.1 (48.8)	161.3 (95.7)	0.003
Outdoor*	3.7 (5.5)	7.9 (8.4)	0.054	4.1 (5.5)	7.6 (8.6)	0.15
Cigarettes	48.5 (32.2)	108.6 (63.1)	0.001	50.8 (29.7)	109.1 (66.5)	0.001
E-cigarettes*	8.0 (7.1)	13.3 (13.5)	0.35	8.6 (7.9)	13.0 (13.5)	0.42
Cigars	9.5 (10.3)	18.4 (12.6)	0.005	9.5 (11.1)	18.9 (11.8)	0.004
Smokeless	16.7 (12.4)	31.3 (26.5)	0.09	20.3 (13.0)	28.1 (28.3)	0.83
Menthol	33.3 (22.6)	74.9 (44.6)	0.001	37.5 (22.3)	72.3 (48.7)	0.005
Non-menthol*	4.5 (4.9)	10.2 (6.2)	0.004	4.8 (5.8)	9.9 (5.7)	0.004
Non-flavored	44.3 (25.8)	86.3 (48.2)	0.001	46.5 (26.3)	86.0 (49.9)	0.003

Tested using Mann–Whitney U test.

Table 3.

Simulated Mean Differences in Tobacco Advertising by Percentage Enrollment of Hispanic and Economically Disadvantaged Students After Simulation of 1000 ft TRO Ban (n = 42 Schools; n = 87 TROs)

Advertisements	Simulated mean differences in the amount of tobacco advertisements after ban
	By Hispanic student enrolment			By economically disadvantaged student enrolment
	≤65% enrolment (n = 21)	>65% enrolment (n = 21)	T-test or Mann–Whitney U test	≤75% enrolment (n = 22)	>75% enrolment (n = 20)	T-test or Mann–Whitney U test
	Mean (Std.)	Mean (Std.)	P-value	Mean (Std.)	Mean (Std.)	P-value
Total Ads	68.6 (50.5)	131.7 (104.2)	0.08	60.8 (54.0)	143.4 (96.6)	0.001
Indoor	65.3 (49.3)	124.9 (97.4)	0.08	57.6 (51.8)	136.3 (90.2)	0.001
Outdoor*	3.3 (5.3)	6.8 (8.4)	0.13	3.1 (5.4)	7.2 (8.3)	0.03
Cigarettes	40.0 (32.2)	85.9 (67.1)	0.04	34.1 (30.4)	94.7 (62.9)	0.001
E-cigarettes*	6.2 (6.4)	9.4 (11.0)	0.62	5.2 (6.3)	10.7 (10.7)	0.08
Cigars*	8.1 (10.7)	14.1 (13.8)	0.12	6.5 (11.4)	16.2 (12.1)	0.002
Smokeless*	14.2 (12.8)	22.3 (25.0)	0.58	14.9 (14.5)	21.9 (24.6)	0.52
Menthol	27.5 (21.9)	57.0 (46.7)	0.049	25.3 (23.1)	61.0 (44.6)	0.001
Non-menthol*	4.0 (5.1)	7.4 (7.1)	0.12	3.1 (5.9)	8.3 (5.9)	0.003
Non-flavored	36.4 (27.0)	66.4 (52.1)	0.10	31.8 (27.6)	72.9 (48.5)	0.001

Tested using Mann–Whitney U test.

Table 4.

Total Number and Percentage Reduction of Tobacco Advertising Around Schools Before and After the Simulated 1000 ft TRO Ban

Percentage enrolment of Hispanic students
	Number of advertisements before ban (observed numbers)			Number of advertisements After ban (simulated numbers)			Absolute and percentage reduction in advertisements due to ban
	All Schools	≤65.5% Hispanic enrolment	>65.5% Hispanic enrolment	All schools	≤65.5% Hispanic enrolment	> 65.5% Hispanic enrolment	All schools n (%)	≤65.5% Hispanic enrolment n (%)	>65.5% Hispanic enrolment n (%)
Total	5341	1738	3603	4205	1440	2765	1136(21.3)	298(17.1)	838(23.3)
Indoor	5099	1661	3438	3993	1371	2622	1106(21.7)	290(17.5)	816(23.7)
Outdoor	242	77	165	212	69	143	30(12.4)	8(10.4)	22(13.3)
Cigarettes	3298	1018	2280	2644	841	1803	654(19.8)	177(17.4)	477(20.9)
E-cigarettes	448	169	279	327	130	197	121(27.0)	39(23.1)	82(29.4)
Cigars	587	200	387	468	171	297	119(20.3)	29(14.5)	90(23.3)
Smokeless	1008	351	657	766	298	468	242(24.0)	53(15.1)	189(28.8)
Menthol	2272	699	1573	1776	578	1198	496(21.8)	121(17.3)	375(23.8)
Non-menthol	279	85	194	224	76	148	55(19.7)	9(10.6)	46(23.7)
Non-flavored	2743	930	1813	2158	764	1394	585(21.3)	166(17.8)	419(23.1)

Percentage enrolment of economically disadvantaged students
	Number of advertisements before ban (observed numbers)			Number of advertisements after ban (simulated numbers)			Absolute and percentage reduction in advertisements due to ban
	All schools	≤75.0% economically disadvantaged	>75.0% economically disadvantaged	All schools	≤75.0% economically disadvantaged	>75% economically disadvantaged	All schools n (%)	≤75.0% economically disadvantaged n (%)	>75% economically disadvantaged n (%)
Total	5341	1963	3378	4205	1337	2868	1136(21.3)	626(31.9)	510(15.1)
Indoor	5099	1873	3226	3993	1268	2725	1106(21.7)	605(32.3)	501(15.5)
Outdoor	242	90	152	212	69	143	30(12.4)	21(23.3)	9(5.9)
Cigarettes	3298	1117	2181	2644	751	1893	654(19.8)	366(32.8)	288(13.2)
E-cigarettes	448	189	259	327	114	213	121(27.0)	75(39.7)	46(17.8)
Cigars	587	210	377	468	144	324	119(20.3)	66(31.4)	53(14.1)
Smokeless	1008	447	561	766	328	438	242(24.0)	119(26.6)	123(21.9)
Menthol	2272	826	1446	1776	556	1220	496(21.8)	270(32.7)	226(15.6)
Non-menthol	279	91	188	224	59	165	55(19.7)	32(35.2)	23(12.2)
Non-flavored	2743	1023	1720	2158	700	1458	585(21.3)	323(31.6)	262(15.2)

Results

Sample Characteristics

One hundred and six TROs (convenience stores with and without gas stations), located within 2640 ft (804.7 m) of 42 middle and high schools were included in this study. A total of 75 were located around schools with above median percentage enrollment of Hispanic students when compared with a total of 31 TROs around schools at the median or below (Table 1). Similarly, 71 TROs were located around schools with greater economic disadvantage while 35 TROs were located around schools at the median or below. A total of 16 schools (out of the 42) were classified as both lower Hispanic student enrollment and lower economic disadvantage. A total of 15 schools were classified as both higher Hispanic student enrollment and higher economic disadvantage. Six schools were classified as higher Hispanic enrollment/lower economic disadvantage and five schools were classified as Lower Hispanic enrollment/higher economic disadvantage.

Table 1.

Demographics of Schools by Percentage Enrollment of Hispanic and Economically Disadvantaged Students in the Schools

	≤65.5% Hispanic Student Enrollment (n = 21)	>65.5%Hispanic Student Enrollment (n = 21)	≤75.0% economic disadvantage (n = 22)	>75.0% economic disadvantage (n = 20)
Race/ethnicity
% White	33.3	5.3	32.7	2.0
% Black	20.7	8.4	14.8	17.1
% Hispanic	33.8	80.9	42.1	73.3
% Others	10.7	5.1	9.9	5.7
SES
% Economically disadvantaged	42.9	75.1	40.8	83.9
Enrolment size (n)	38 169	26 791	41 833	23 127
No. of TROs	31	75	35	71

NB: Enrollment %s does not add up to 100 because data were suppressed if numbers were too small and therefore could be identifiable at the school level.

Observed Mean Difference in Tobacco Advertisements by School Enrollment

The mean number of observed total advertisements, indoor, cigarettes, cigars, menthol flavored, non-menthol flavored and non-flavored tobacco advertisements were significantly higher around schools with a higher enrollment of Hispanic students in comparison to schools with lower enrollment of Hispanic students (Table 2). For each of these variables, schools with greater enrollment of Hispanic students had approximately two times greater mean numbers of advertisements within 1000 ft when compared with schools with lower enrollment of Hispanic students. There were no significant differences in the mean of e-cigarette or smokeless advertising.

The mean number of observed total advertisements, indoor, cigarettes, cigars, menthol flavored, non-menthol flavored, and non-flavored tobacco advertisements were significantly higher around schools with higher enrollment of economically disadvantaged students in comparison to schools with lower enrollment of economically disadvantaged students. As in the comparisons by ethnic composition, more socioeconomically disadvantaged schools had approximately two times greater mean numbers of advertisements as compared to schools that were less socioeconomically disadvantaged.

Simulated Mean Difference in Tobacco Advertisements After Simulation of the TRO Ban Around Schools

The simulation of the 1000 ft ban of TROs around schools resulted in a reduction of 19 outlets. After the removal of these outlets which represents a ban on tobacco sales, 87 TROs remained. The simulated ban reduced the disparity in the number of advertisements around schools based on the enrollment of Hispanic students slightly. Cigarettes and menthol flavored tobacco advertisements were still significantly higher around schools with a higher enrollment of Hispanic students after simulation of the ban (Table 3). For schools with a higher enrollment of economically disadvantaged students, all variables that were significant prior to the ban remained significant and outdoor advertising became significant after the ban (Table 3).

Percentage Reduction in Tobacco Advertisements After Simulation of the 1000 ft Ban

After simulation of the 1000 ft ban, the percentage reduction in tobacco advertisements for all variables was higher around schools with a greater enrollment of Hispanic students in comparison to the percentage reduction around schools with lower enrollment of Hispanic students (Table 4).

Unlike results by Hispanic student enrollment, the percentage reduction in tobacco advertisements was actually lower around schools with a greater enrollment of economically disadvantaged students in comparison to the percentage reduction around schools with lower enrollment of economically disadvantaged students, after the simulation of the 1000 ft ban (Table 4).

Discussion

Our study found that TROs around schools with a higher enrollment of Hispanic and economically disadvantaged students have significantly more tobacco advertisements, overall and for most types of tobacco advertising, in comparison to TROs around schools with lower enrollment of Hispanic and economically disadvantaged students. With a simulation of a ban of TROs within 1000 ft of schools, we observed a small reduction in this disparity in tobacco marketing by percentage enrollment of Hispanic students. Many studies have suggested that implementation of a 1000-ft ban of retail outlets that sell tobacco around schools will help in reducing the unequal tobacco marketing exposures directed at people of color and youth. To our knowledge, our study is the first to examine the impact such place-based tobacco-use prevention policy will have on the actual number of tobacco advertisements around these groups.

We observed that schools with a higher enrollment of Hispanic and economically disadvantaged students had more tobacco advertisements at TROs around them. The means of most of the advertisement measures were also higher around schools with a higher enrollment of Hispanic students and economically disadvantaged students, with the exception of e-cigarette and smokeless advertising. Other studies have also documented the disparity in tobacco marketing, but with a focus on advertising and retailer density in the broader community, rather than around schools.^2–4,40,41 A systematic review of studies on tobacco marketing concluded that neighborhoods with lower income residents have more tobacco marketing and retail outlets.^2,4,40,41 Studies have also shown that communities of color are exposed to a disproportionate amount of tobacco advertising and retailers.^2,4,40,41 Our study adds to this literature documenting the overexposure of vulnerable communities to tobacco marketing.

Some researchers have suggested that a 1000-ft ban of TROs around schools could help reduce tobacco-marketing disparities.^27,42 In our study, we observed a reduction in the disparity of tobacco marketing based on the enrollment of Hispanic students after implementation of the ban. For example, percentage reductions in advertising among schools with greater enrollments of Hispanic students ranged from 13.3% to 29.4% fewer advertisements after the simulation of the ban (with most reductions close to 25%), while reductions in schools with lower enrollments of Hispanic students had reductions ranging from 10.4% to 23.1% (with most reductions close to 17%). This means that students in schools with greater enrollments of Hispanic students will benefit from greater reductions in tobacco marketing around their schools. Further, the differences between the mean numbers of tobacco advertisements were no longer significantly different between the two school categories, with the exception of cigarette and menthol advertising, which remained significantly greater around schools with greater enrollments of Hispanic students. This finding suggests that this 1000 ft ban policy should be combined along with other advertising restrictions to reduce exposure to advertising for tobacco products like cigarettes and menthol flavored tobacco.

Interestingly, when simulating the ban on tobacco sales and examining reductions in tobacco advertising by school-level economic disadvantage, we found that schools that were less socioeconomically disadvantaged actually had greater reductions in tobacco advertising when compared with schools that were more economically disadvantaged. For example, percentage reductions around more economically disadvantaged schools ranged from 5.9% to 21.9% (with most reductions close to 15%), while percent reductions among less economically disadvantaged schools ranged from 23.3% to 39.7% (with most reductions close to 30%). Further, the mean differences between the school categories became greater, although the advertising reduced for both. These findings, contrary to what we expected, suggest that students in less economically disadvantaged schools would benefit from greater reductions in tobacco advertisement exposure from a 1000-ft TRO ban around schools as compared to students in more economically disadvantaged schools. It is important to note that though the relative reduction for total advertisements among schools with less economically disadvantaged students (31.9%) was greater than that for schools with more economically disadvantaged students (15.1%), the absolute reductions were similar between the groups (626 vs. 510). Further, the schools with less economically advantaged students had a lower mean number of advertisements and, as such, percentage reductions may be larger. However, the total number of tobacco advertisements remained higher around schools with more economically disadvantaged students after the simulated ban.

Other studies have examined the disparity-reducing potential of this tobacco control policy alternative, banning TROs, with a focus on outlet density.²⁶ Ribisl et al.²⁶ found that the number of TROs per 1000 people reduced from 1.28 to 0.36 in the lowest income quintile, while reducing from 0.84 to 0.45 in the highest income quintile with implementation of the 1000 ft ban around schools in New York. In Missouri, the same study found that the number of TROs per 1000 people reduced from 1.18 to 0.82 in the lowest income quintile, while reducing from 0.48 to 0.37 in the highest income quintile.²⁶ While reductions in density do result in reduced numbers of TROs themselves, our study suggests that total advertising exposure is still problematic. Our previous work shows that while a simulated ban of tobacco advertising within 1000 ft of middle schools and high schools would reduce tobacco advertising exposure, there are significantly more tobacco advertisements within 1000–2000 ft of schools.¹⁷ In addition, more economically disadvantaged schools did not benefit more in advertising reduction in comparison to schools with less economically disadvantaged students after the simulated ban. Taken together, these findings suggest that other tobacco control policies (e.g., expanding the buffer zone, reducing timing/placement of advertising in stores) will be needed to reduce advertising exposure among youth, particularly as youth are likely exposed beyond 1000 ft of their schools.

There is legal backing for the use of control measures such as the 1000 ft ban of TROs around schools in the 2009 Tobacco Control Act (TCA). The TCA reaffirmed States and local communities’ additional authority to use point-of-sale control strategies to regulate tobacco retail environment where necessary.^25,43 Researchers have suggested the use of licensing and zoning laws like conditional use permits (CUPs) to pursue public health policies such as these.^25,44 These policies can be introduced in the form of stand-alone ordinances that can be used to directly regulate tobacco retailing at the local level. With legal assistance and proper legislative language, these policies can be carefully drafted in a way to likely withstand and overcome potential constitutional challenges from the tobacco industry.^25,42,44 Retailer licensing laws have been used to prohibit new tobacco retailers from opening within 1000 ft of a school in places like Santa Clara County in 2010^45,46 and in Palo Alto in 2017.^47,48 While our study findings do suggest unequal impacts of the ban when considering ethnic composition of the school when compared with economic composition of the school, the bans did reduce absolute numbers around all schools, suggesting that this place-based tobacco reduction strategy could reduce exposure to tobacco advertising for all students, and increased benefit for students who attend schools with more Hispanic students.

Our study does have some limitations. First, only convenience stores with/without gas stations within a half mile of participating middle and high schools in the TATAMS study were included in this study. The TATAMS schools and hence the TROs, were limited to major metropolitan areas in Texas. Therefore, our results may not be generalizable to all counties in Texas or to the US as a nation. Nevertheless, our study is strengthened by the large number of TROs included across the four largest and most diverse metropolitan areas in Texas. These are TROs typically seen in urban and suburban areas with similar store layouts and advertising (e.g., 7-Eleven, Family Dollar, Citgo, etc.) Therefore, we believe our study provides a good representation of the TRO density and advertisements observed around middle and high schools, nationwide. In addition, we conducted a “census” of all tobacco advertisements in the stores within the study area. Also, private schools were not included in this study as we had only the enrollment data for public and charter schools. Further, our study used geocoded points for location since we do not have the exact boundaries around schools and retail stores. We tried to account for the distance between the school center point and its boundary by including an extra 150 ft around the school buffers. Though this buffer may be accurate for some schools, and less accurate for some others, we believe it does not bias the results of our study in either direction. Finally, we were not able to examine differences by the percentage of Black or white students due to limitations in sample size. Future work is needed that can explore these differences in other contexts where comparisons are more robust.

Conclusion

Disparity in tobacco marketing exists around neighborhoods by socioeconomic class and racial/ethnic composition. Studies have shown that economically disadvantaged neighborhoods and communities of color have disproportionate amounts of tobacco retailers.^2,3,40,49 Tobacco companies market their products in these communities through more targeted advertisements and price promotions.^4,5,18 In our study, we found that schools with a higher enrollment of Hispanic and economically disadvantaged students have a significantly greater numbers of observed tobacco advertisements within a half-mile in comparison to schools with a lower enrollment of Hispanic and economically disadvantaged students. Simulation of the 1000 ft ban reduced the tobacco marketing disparity observed by enrollment for schools with greater Hispanic student enrollment and while it did not reduce the disparities by economic advantage, all schools saw a reduction in tobacco advertising. Given these findings, we recommend the implementation of the 1000 ft ban of TROs around schools to reduce tobacco marketing around schools and to reduce disparities in targeted marketing around schools with greater numbers of Hispanic students. This strategy, together with other tobacco control measures, can help to reduce youth exposure to tobacco product advertising.

Funding

This work was supported by the National Cancer Institute and the FDA Center for Tobacco Products (CTP) [1 P50 CA180906-02] and the National Cancer Institute [R01-CA24988301A1].

Declaration of Interests

Dr. Harrell is a consultant in litigation involving the vaping industry. All other authors have no conflicts to report.

Data Availability Statement

Data available on request.