Background Flavoured tobacco is increasingly popular with youth. The twin cities of Minneapolis and Saint Paul, Minnesota restricted the sale of flavoured tobacco to adult-only tobacco shops in an effort to reduce youth access and exposure to these products. This observational study explores the impact of these policies on the retail environment.
Methods Between November 2015 and April 2017, observational assessments were conducted at convenience and grocery stores preimplementation and postimplementation of flavoured tobacco restrictions in Minneapolis (n=41), Saint Paul (n=37) and the comparison city of Brooklyn Park, Minnesota (n=14).
Results After policy implementation, significantly fewer of the convenience and grocery stores sold flavoured tobacco in Minneapolis (85.4% vs 39.0%, p<0.001, n=41) and Saint Paul (97.3% vs 8.1%, p<0.001, n=37). The average proportion of tobacco inventory that was unflavoured increased significantly in Minneapolis (42.9%±2.7% vs 56.7%±3.1%, p=0.002) and Saint Paul (40.8%±2.5% vs 70.3%±1.9%, p<0.001). Flavoured tobacco was available at significantly fewer convenience and grocery stores in Minneapolis (39.0%, p<0.001) and Saint Paul (8.1%, p<0.001) than in the comparison city of Brooklyn Park (100%, n=14). Most retailers complied with these policies by removing flavoured tobacco from their shelves.
Conclusions Policies that restrict the sale of flavoured tobacco to adult-only tobacco shops are associated with decreased availability of flavoured tobacco in convenience and grocery stores and increased proportion of tobacco inventory that is unflavoured.
- tobacco industry
- public policy
Statistics from Altmetric.com
The increasing popularity of flavoured tobacco is a public health concern.1 There is evidence flavoured products are preferred by youth tobacco users as most youth who report tobacco use started with a flavoured product,2 3 and many report flavours as a reason they use tobacco. Nationally, 81.5% of electronic cigarette, 78.9% of hookah, 73.8% of cigar and 69.3% of smokeless tobacco users under the age of 18 years report using these products because they are flavoured.3 Since 2011, sales of flavoured tobacco products have increased in the USA across nearly all product categories.1
A growing number of cities and counties across the country (eg, Providence, Rhode Island, New York City, New York and Oakland, California) have adopted policies that restrict the sale of flavoured tobacco. These policies take different approaches. New York City’s policy exempts menthol and restricts the sale of tobacco with other flavours to a handful of ‘tobacco bars’4 while Oakland’s policy restricts the sale of flavoured tobacco (including menthol) to tobacco stores that derive >60% of revenue from the sale of tobacco.5
Limited evidence exists on the impact flavoured tobacco restrictions. Recent studies suggest these types of policies decrease flavoured tobacco sales6 7 and reduce the likelihood youth will use tobacco.8 However, there are concerns these policies might be ineffective or less effective than they could be because the tobacco industry may find ways to evade these new restrictions. For example, tobacco companies might circumvent local restrictions by marketing the same flavoured products with new and ambiguous flavour names (eg, ‘Ba Boom’ or ‘Purple’; see online supplementary figure for examples).9
Supplementary file 1
To our knowledge, this is the first study to present observational evidence about how flavoured tobacco restrictions impact the retail landscape. Minneapolis, a city with 350 tobacco retailers and a population of 410 93910 became the first city in Minnesota to adopt a flavoured tobacco restriction in July 2015 (implemented January 2016). Saint Paul, Minnesota’s capitol city, with 260 tobacco retailers and population of 300 85110 followed in January 2016 (implemented April 2016). Both policies restrict the sale of flavoured tobacco to a limited number of adult-only tobacco product shops in each city (approximately 25 in Minneapolis and 15 in Saint Paul). In order to qualify as an adult-only tobacco product shop, a retailer must make >90% of revenue from the sale of tobacco and prohibit entry to anyone younger than 18 years. The policies apply to electronic cigarettes and exempt menthol. This study explores retailer compliance, defined as availability of flavoured tobacco, in Minneapolis and Saint Paul. It also examines the impact of local flavoured tobacco restrictions on the proportion of tobacco inventory that is menthol, other flavour and unflavoured; exterior advertising for tobacco and availability of tobacco with ambiguous flavour names. While the body of literature in this area is growing, at the time of this study was conducted, there was a lack of other published studies to drive hypotheses, making this study exploratory in nature.
We conducted observational retail assessments at a random sample of convenience and grocery stores with tobacco licences in Minneapolis and Saint Paul before (round 1) and after (round 2) policy implementation with further follow-up conducted approximately 14 months after policy implementation in Minneapolis only (round 3). Findings were compared with the neighbouring city of Brooklyn Park at the postimplementation time point only (figure 1). Brooklyn Park (suburb of Minneapolis with population of 79 149 and 50 tobacco retailers)6 was selected for comparison because the city has similar demographics to Minneapolis and Saint Paul but no flavoured tobacco restriction.
Convenience and grocery stores were randomly sampled from each city’s tobacco retail licensing list. In Minneapolis, we completed 42 of 48 assessments (87.5%, 5 stores were not assessed because they were out of business and 1 was not because collectors did not feel safe) preimplementation (round 1) and 41 of 42 assessments (97.6%, 1 store was out of business) 5 months postimplementation (round 2). We conducted another assessment (round 3) 14 months postimplementation and completed 40 of 41 assessments (97.6%, 1 store was out of business) because we found compliance rates were substantially lower in Minneapolis than in Saint Paul. One store had only partially complete data, and thus the pre/post N for round 1–3 comparisons in Minneapolis was 39. In Saint Paul, we completed 39 of 39 assessments preimplementation (round 1) and 37 of 39 (94.9%, 1 store was out of business and 1 no longer sold tobacco) assessments 2 months postimplementation (round 2). The pre/post N for round 1–2 comparisons in Saint Paul was 37. At the postimplementation assessment (round 2, June 2016), we also completed 14 out of 14 assessments in Brooklyn Park (figure 1).
Pairs of collectors used a paper and pencil data form to assess: (1) product availability by flavour (menthol, other flavour, unflavoured); (2) availability (yes, no, unsure) of tobacco products with ambiguous flavour names (eg, Ba Boom); (3) per cent of available tobacco inventory that was flavoured, unflavoured and menthol, which was assessed by observing and estimating what per cent of the store’s tobacco inventory fit into each category (<10%, about a quarter, about half, about three-quarters, 90% or more); (4) number of exterior tobacco advertisements by flavour type and (5) total number of exterior tobacco advertisements. While the policies in Minneapolis and Saint Paul did not restrict tobacco advertising, these measures were included to determine if the policies had any impact on the quantity or type of tobacco ads seen at retailers. Due to the small number of exterior ads observed overall, these variables were recoded as binary variables: exterior ad(s) present or absent.
Data collectors completed a training protocol that included instructions on answering each of the questions on the assessment form, assessing tobacco inventory estimates and counting tobacco ads. They were provided examples of the tobacco products to be assessed and an assessment guide to take with them into the field. The two data collectors worked together to complete each retail assessment, resolving any coding disagreements as they arose. Collectors did not seek permission from store staff before completing an assessment but were instructed to leave if asked, which did not happen at any of the locations assessed.
We calculated summary statistics for each measure. Chi-square tests (or Fisher’s exact tests where necessary, an expected cell count <5) and two-sample t-tests were used to compare the assessment items between Minneapolis and Brooklyn Park, as well as Saint Paul and Brooklyn Park, while McNemar’s tests (or McNemar’s exact tests where necessary, discordant pairs b+c<20) and paired t-tests were used to compare the assessment items between Minneapolis/Saint Paul rounds 1 and 2, and Minneapolis rounds 1 and 3. Tobacco inventory categories were recoded and analysed as a continuous variable (options: 0%, 10%, 25%, 50%, 75% and 90%). If a store did not sell that type of tobacco, their inventory was recorded as 0%. The proportion of ads for menthol tobacco was computed by dividing the total number of tobacco ads by the number of menthol-specific tobacco ads. Analyses were completed using R V.3.3.1 (The R Foundation for Statistical Computing, Vienna, Austria).
Availability of flavoured tobacco
After policy implementation in round 2, significantly fewer of the convenience/grocery stores sold flavoured tobacco in Minneapolis (85.4% vs 39.0%) and Saint Paul (97.3% vs 8.1%) (table 1). For the round 3 assessment in Minneapolis, flavoured tobacco availability was lower than in round 2 and significantly lower than in round 1 (85.4% vs 15.4%). Significantly more of the Brooklyn Park convenience/grocery stores sold flavoured tobacco than the stores in both Saint Paul and Minneapolis after policy implementation.
Presence of tobacco with ambiguous flavour names
In Minneapolis between rounds 1 and 3, there was a significant reduction in the per cent of stores that sold tobacco with ambiguous flavours names such as ‘TaTa’ and ‘Blue’ (80.5% vs 61.5%) (table 1). In Saint Paul, there was no significant change in the per cent of stores that sold products with ambiguous flavour names (67.6% vs 81.1%).
Compared with Brooklyn Park (57.1%), there was no difference in the per cent of stores in Minneapolis and Saint Paul that sold these products in round 2.
Exterior tobacco ads by flavour category
In Minneapolis and Saint Paul, the presence of exterior tobacco ads stayed relatively constant. Although not statistically significant, the proportion of stores with flavoured exterior ads in Minneapolis decreased from 12.2% in round 1 to 0% in round 3. The proportion of ads for menthol tobacco decreased significantly in Saint Paul but did not change in Minneapolis (table 1). Compared with Brooklyn Park, Minneapolis and Saint Paul retailers both had significantly more stores with unflavoured exterior ads in round 2.
Per cent of tobacco inventory by flavour category
In Minneapolis between rounds 1 and 2, the average per cent of flavoured tobacco inventory per store decreased significantly while the average per cent of unflavoured tobacco inventory increased significantly. The menthol inventory did not change. In round 2, compared with Brooklyn Park, Minneapolis stores had a significantly lower percentage of flavoured tobacco inventory, and higher percentages of menthol and unflavoured tobacco inventory (table 1 and figure 2).
In Saint Paul, the average percentages of flavoured and menthol tobacco inventory per store decreased significantly between rounds 1 and 2 while the percentage of unflavoured tobacco inventory increased. In round 2, Saint Paul stores had a significantly lower average percentage of tobacco inventory that was flavoured than Brooklyn Park, and a significantly higher per cent of inventory that was unflavoured. There was no difference in the per cent of menthol inventory (table 1 and figure 2).
Minneapolis and Saint Paul, Minnesota both restrict the sale of flavoured tobacco and we examined the impact of these policies on the retail environment. Key findings indicate these policies decrease the availability of flavoured tobacco, and most retailers complied with these policies. Study findings are supported by data from the City of Minneapolis, which indicate high compliance rates (84% in 2016 and 82% in 2017) (City of Minneapolis, personal communication, 2018). Saint Paul also conducts compliance checks but these data were unavailable at the time of writing (City of Saint Paul, personal communication, 2018).
In our study, retailer compliance was initially lower in Minneapolis than it was in Saint Paul (61.0% vs 91.9%). Minneapolis city staff responded by completing additional retailer outreach after round 2, which may have been responsible for improved compliance rates in round 3. This highlights the importance of incorporating a robust retailer education effort into the implementation of a flavoured tobacco restriction.
Policies resulted in an increase in the average per cent of tobacco inventory that is unflavoured, which is likely less appealing to youth.2 3 It was feared that because these policies restricted all flavours except menthol they might result in an increase in the presence of menthol tobacco or an increase in advertising for menthol tobacco. Results indicate this did not occur. While the policies in Minneapolis and Saint Paul were not associated with increased availability and advertisement of menthol tobacco, they also do not restrict the sale of these products. Menthol tobacco is a public health concern as these products are popular among youth11 and groups that are disproportionately harmed by tobacco such as African-Americans11 and Lesbian, Gay, Bisexual, Transgender, and Queer (LGBTQ) people.12 Furthermore, there is also evidence menthol could make tobacco harder to quit.13
In both cities, the average number of retailers with exterior tobacco advertisements did not change significantly after policy implementation, however in Minneapolis, the per cent of stores with flavoured tobacco ads decreased from round 1 to round 3. The proportion of exterior tobacco ads that were menthol decreased significantly in Saint Paul but did not change in Minneapolis. These results indicate flavoured tobacco restrictions have the potential to indirectly reduce certain types of tobacco ads. This is noteworthy because directly restricting advertising content is nearly impossible due to First Amendment protections.
These polices were not associated with an increased presence of products with ambiguous flavour names (eg, Purple). In Minneapolis, there was a significant reduction in the availability of products with ambiguous flavour names after the policy was implemented; however, in Saint Paul, there was no significant change in availability. In both cities, enforcement officials determine on a case-by-case basis whether or not an ambiguously labelled product is a violation of the flavour restriction. This determination is based on if there is supporting evidence that a product is or is not flavoured (such as packaging imagery or a description on the company website). A 2017 study measured the presence of flavouring chemicals in a sample of tobacco products with ambiguous flavour names.10 Nearly all of the ambiguously named products had a high concentration of the same flavour chemicals found in tobacco products with recognisable flavour names. Given this, there is a need to monitor the availability of these products and perhaps restrict their sale through improved policy language.
This study has limitations. First, we used observation to measure the variables of interest so the data may be subject to measurement bias, and the two data collectors did not make independent assessments so inter-rate reliability was not assessed. Second, we only conducted a second round of postimplementation assessments in Minneapolis because compliance was much lower than in Saint Paul during round 2. The comparison city was added to the study to strengthen the design only after the policies passed; therefore, we only conducted one round of assessments in the comparison city of Brooklyn Park after the policies were implemented in Minneapolis and Saint Paul. Our sample sizes were small, and our analyses were likely underpowered. Finally, due to the exploratory aim of the study to examine the effects of the policy on the retail environment, we performed several statistical tests which comes with a higher risk of type I error. Further research is needed to understand how these policies impact youth tobacco use. Despite these limitations, this study helps add to the limited literature about the effectiveness of flavoured tobacco restrictions by quantifying the real-world impact of these policies. To our knowledge, this is the first study to present observational evidence about how these policies impact the retail landscape. Results can be used to inform similar restrictions in other cities.
What this study adds
Flavoured tobacco is increasingly popular with youth. Cities and counties across the USA are restricting the sale of flavoured tobacco in an effort to reduce youth access and exposure to these products.
Policies that restrict the sale of flavoured tobacco are associated with decreased availability of flavoured tobacco and an increase in the average proportion of tobacco inventory that is unflavoured, which is likely less appealing to youth.
Most retailers complied with these policies by removing flavoured tobacco from their shelves. However, retailer education is important to ensure compliance.
The authors would like to acknowledge Becky Lien for her review of the analysis methods, as well as Chris Farmer-Lies, Kari Oldfield, Gabriel Glissmeyer, Sadie Lundquist, Amanda Swygard, Choua Lee, Kristen Werner, Joann Wade, Esha Seth, Chris Turner and Melissa Mady for their data collection assistance.
Contributors BB developed the study design, oversaw the data collection, led the development and writing of the article and is responsible for the overall content. SCC conducted data analysis and was involved in drafting and critically reviewing the article. AL provided input on study design and implementation and critically reviewed the article. CMM assisted with study design and critically reviewed the article. JD’S assisted with study design and critically reviewed the article. BAS provided input on study design and critically reviewed the article. All authors have reviewed and approved the ﬁnal version of the submitted article.
Funding This work was supported by ClearWay Minnesota and Blue Cross and Blue Shield of Minnesota.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.