Abstract
Our objective was to determine how point-of-sale tobacco marketing may relate to sales to minors. The authors used data from a 2007 cross-sectional study of the retail tobacco marketing environments in the St. Paul, MN metropolitan area matched with a database of age-of-sale compliance checks (random, covert test purchases by a minor, coordinated by law enforcement) of tobacco retailers and U.S. Census data to test whether certain characteristics of advertising or neighborhoods were associated with compliance check failure. The authors found that tobacco stores were the most likely type of store to fail compliance checks (44% failure), supermarkets were least likely (3%). Aside from a marginally significant association with Hispanic population proportion, there was no other association between either store advertising characteristics or neighborhood demographics and stores’ compliance check failure. Though our findings were null, the relationship between advertising and real youth sales may be more nuanced as compliance checks do not perfectly simulate the way youth attempt to purchase cigarettes.
Introduction
The tobacco industry is directly involved in point-of-sale marketing, working with merchants to advertise, promote, and showcase their products in stores in a variety of ways. Many tobacco vendors display advertisements both inside and outside of the store. There are shelving units for cigarette packs and cartons with tobacco branding as well as functional objects in stores such as welcome mats and clocks. These advertising items draw attention to the brand and can communicate discounts and specials available at the store. Tobacco companies actively engage tobacco vendors in marketing programs with lucrative incentives for the retailer (Feighery, Ribisl, Schleicher, & Clark, 2004). Often, retailers who participate in marketing programs agree to let the tobacco company control where shelving, displays, and/or products are placed within their store (Feighery, Ribisl, Clark, & Haladjian, 2003). Stores that sell tobacco are far more likely to receive incentives for displaying tobacco products or ads than for any other type of product (Feighery, Ribisl, Achabal, & Tyebjee, 1999).
The influence of the tobacco industry’s promotion tactics may result in creating store environments where it is easier for youth to illegally purchase tobacco. Tobacco companies often pay more for self-service displays compared to restricted displays and self-service displays may be more conducive to youth attempting to purchase cigarettes (Bloom, 2001). One author has suggested that clerks might be less likely to ask for proof of age if they work in stores, that are frequented by tobacco representatives who develop a regular relationship with the store’s clerks (Bloom, 2001). Previous research has shown that stores that display tobacco industry furnished age of sale signs (which may indicate a heavily incentivized store) are more likely to fail compliance checks compared to stores that have government age of sale signs (Cowling & Robins, 2000). Though it is unknown whether stores that receive more incentives or have greater amounts of tobacco advertising are more likely to sell cigarettes to minors, past research has shown that retail tobacco advertising is more prevalent in stores where adolescents shop frequently (Henriksen, Flora, Feighery, & Fortmann, 2002).
In the past two decades, it has become increasingly difficult for youth to purchase tobacco in stores, as measured by random test purchases. In 1992, Congress passed the Synar Amendment which mandated that all states and territories prohibit the sale of tobacco to minors by 1994, and enforce the law or risk the loss of federal funds for addressing chemical health issues in their state ( State Oversight of Tobacco Sales to Minors, 1995). In 1997, an average of 40.1% of U.S. retailers failed their Synar compliance checks; but by 2010, only 9.3% failed (Substance Abuse and Mental Health Services Administration, U.S. Department of Health and Human Services, 2011).
Youths’ perception of store advertising and promotion may undermine the Synar Amendment effort to reduce the supply of cigarettes to youth. When youth are shown photographs of stores with retail level tobacco advertising, the youth report that they perceive these stores as less likely to ask for proof of age and more likely to sell cigarettes compared to photographs of stores not displaying cigarette advertisements (Henriksen et al., 2002; Wakefield, Germain, Durkin, & Henriksen, 2006). Furthermore, exposure to higher levels of retail-level tobacco advertising increases the likelihood that adolescents initiate smoking (Feighery, Henriksen, Wang, Schleicher, & Fortmann, 2006; Henriksen, Schleicher, Feighery, & Fortmann, 2010; MacFadyen, Hastings, & MacKintosh, 2001; Schooler, Feighery, & Flora, 1996; Wakefield et al., 2006). If exposure to point-of-sale tobacco marketing actually enables youth to obtain tobacco from commercial sources more readily, this could lead to increased tobacco use since commercial access has been associated with greater tobacco use in youth (Forster, Chen, Blaine, Perry, & Toomey, 2003; Harrison, Fulkerson, & Park, 2000; Widome, Forster, Hannan, & Perry, 2007).
In this study, we first sought to describe compliance check failure patterns in a census of tobacco retailers in a Midwestern metropolitan area. Our main objective was to test the association between point-of-sale advertising intensity and likelihood that a store would fail a compliance check. We hypothesized that stores with greater amounts of in-store marketing and/or certain types of advertising and promotion would be more likely to fail compliance checks.
Method
Data Collection
The data for this study came from three sources: (1) Observations of the advertising environment in establishments, which we will refer to as “store assessments,” (2) a record of age-of-sale tobacco checks where an undercover minor working with law enforcement attempts to purchase tobacco, which we refer to as “compliance checks,” and (3) Demographic data from the Year 2000 U.S. census. This study was exempt from institutional review board review as it did not involve human subjects.
Store assessments
In 2007, the point-of-sale advertising and promotion environment was assessed in a census of all 655 licensed tobacco vendors Ramsey County, MN (which includes St. Paul, MN) and in the Dakota County, MN cities of South St. Paul, MN and West St. Paul, MN by trained data collectors. The data collectors visually inspected the retail establishments and systematically recorded observations on a standard data collection form.
The data collection forms used by the data collectors were developed based on a review of the literature (Barbeau, Wolin, Naumova, & Balbach, 2005; Henriksen, Feighery, Schleicher, Haladjian, & Fortmann, 2004; 2007; Laws, Whitman, Bowser, & Krech, 2002; Pucci, Joseph, & Siegel, 1998; Wakefield et al., 2006) and consultation with other groups who had conducted retail store observations focused on tobacco advertising and promotion including Massachusetts Operation Storefront, Operation Storefront (TX), Operation Storefront (WA), Operation Storefront (WY), and California Strategic Tobacco Retail Effort (STORE) Campaign. The assessments proceeded according to standardized protocol. At each assessment, the data collector notified store employees that they were doing an assessment of advertising and marketing practices. Data collectors were instructed not to specify that they were focusing on tobacco marketing and advertising practices unless they were specifically asked. The assessments typically took from 5 to 30 min to complete depending on the amount of advertising/promotion at the site.
There was a wide variety of types of vendors in our region that held tobacco licenses. Initially, for each establishment, the store assessors indicated one or more vendor category out of 22 choices on the assessment form. During the data cleaning process, we assigned each establishment to one of eight categories: (1) gas/convenience store (does not sell raw meat), (2) supermarket, (3) restaurants/bars and places with bars (which includes veteran clubs, billiard clubs, music venues, and bowling alleys), (4) small grocery store (which must sell raw meat and milk), (5) tobacco shop, (6) liquor store, (7) drug store (which primarily dispenses medications and personal care items), and (8) miscellaneous (which includes discount stores, gift stores, hotel/motel/resort, automotive shop/car wash, adult entertainment venues, golf courses, amusement parks, assembly halls, and establishments that were coded as unknown or other).
We focused on several aspects of stores’ tobacco retail environments that we hypothesized might be related to compliance check failure. For “Overall exterior/interior impression,” the assessor indicated “None,” “Discreet,” “Moderate,” or “In your face” when zero, less than 5%, 5–25%, or greater than 25% of the interior (such as counters or window interiors) or exterior (such as the outside of windows or doors) of the property was covered in tobacco advertising. Assessors also indicated whether there were tobacco advertisements located one foot or less from candy, any self-service tobacco products (those that could be purchased by a customer, without assistance from a store employee), ads six feet or less from the counter, ad coverage on doors and window exteriors, interior and exterior advertisement for tobacco merchandise, interior and exterior advertisements for price deals (such as rebates, buy-one-get-one free, etc.), if tobacco ads were only located in the areas where tobacco products were sold, and whether there were ads less than three feet from the ground (which would be easy for a young child to see).
Compliance checks
In Minnesota, as per Minn. Stat. § 461.12, the local tobacco licensing authority is required to conduct unannounced compliance checks (where a minor, over the age of 15 but under age 18, working under the supervision of law enforcement or agency staff, attempts to purchase cigarettes) at every licensed tobacco retailer at least once per year (Office of Revisor of Statutes, State of Minnesota, 2006). Minors were trained but did not follow a standard script. They were not allowed to use fake identification (ID) and if asked, would show their real ID card. A failure is defined as the sale of tobacco to the youth, regardless of whether the store clerk examined the minor’s ID. The name and address of the store as well as whether the sale was completed is recorded at each compliance check. Compliance check data, which are publicly available, were requested from the all cities in our study region. Some communities do more than one compliance check per store, per year. In these cases, we used the compliance check that was closest in time to the store assessment date.
Demographic/geographic data
The addresses of all tobacco vendors and schools within the study region were geocoded or labeled with the exact geographic coordinates (latitude and longitude). Using ArcGIS software (ESRI, Redlands, California), these geocoded stores were spatially joined with census block groups. We used the publically available U.S. Year 2000 Census data on these block groups’ demographic (age, race/ethnicity, below 150% poverty level, % on public assistance) composition. Buffers were created to determine which stores were within 400 m (a reasonable walking distance) from schools, which we hypothesized might be a confounder as some have observed that advertising is denser near schools (Pucci et al., 1998) and that youth smoking may relate to density and proximity of tobacco retailers near schools (Henriksen et al., 2008; McCarthy et al., 2009).
Venue Inclusion Criteria
The store assessment process began with enumeration of all of the establishments with current tobacco licenses in the 18 cities within the study region. Of the 655 licensed tobacco vendors in the region, both interior and exterior assessments were completed on 485 establishments (74.0%). Thirty-one establishments (4.7% of total) could not be observed because they had gone out of business before their licenses expired but prior to the store observations. Forty-one establishments (6.3% of total) did not sell tobacco despite having a license. Additionally, twelve establishments could not be located by staff (1.8% of the total), 21 (3.2 % of the total) could not be surveyed due to remodeling or because the store was still under construction, and for 51 establishments (7.8% of the total) observers could only do exterior observations for various reasons such as store staff asking the observers to leave or observers not feeling comfortable entering the establishment. There were 12 stores that were excluded due to incomplete or contradictory information in their assessment forms and two stores’ assessment forms were lost prior to data entry. Completion of a full store assessment was not significantly associated with whether stores passed their compliance check (p = .931). There were 18 venues where the compliance check record was incomplete for a variety of reasons (for instance, a store might have been closed at the time of the attempted compliance check). Thus, our analyses were conducted on 467 establishments.
Analysis
We described compliance check failure proportions for various types of stores and by census block group demographics. We then calculated failure proportions for various promotion and advertising characteristics. Chi-square tests were used to test whether failure was associated with a store being situated in a block group that was in the top decile of each demographic item. We used the top decile as a cutoff in order to examine more extreme examples of census block groups that had relatively high proportions of certain demographics. We also used t tests to examine whether the percentage of people in a block group for each demographic item was associated with compliance check failure. Finally, we used χ2 tests to examine whether specific advertising practices were associated with failure.
Results
Table 1 shows associations between store characteristics and compliance check failure. Tobacco shops were most likely to fail compliance checks with four of the nine failing. Supermarkets were least likely to fail with only 1 of the 30 in the sample failing. The prevalence of failing among chains was not significantly different from nonchain stores. When we compared stores that were situated in neighborhoods at the 90th percentile and above for each demographic characteristic to stores in neighborhoods there were below the 90th percentile, we did not find any characteristic of stores’ block groups to be associated with stores failing compliance checks (Table 1). However, when we tested associations between percentage of the population in each store’s block group with compliance check failure, we found that stores in block groups with a greater percentage of Hispanic residents were marginally significantly more likely to fail (p = .045). Table 2 shows mean percentages of demographic characteristics in stores’ census block groups for stores that passed and failed. Aside from percentage Hispanic, no other demographics characteristic was associated with failure in these tests.
Description of the Stores’ Compliance Check Failure by Store Type and Block Group Demographics
Note. aDue to small cell sizes, this p value is based on a Fisher’s exact test.
*“Top decile” refers to stores that are situated in a block that was in the top decile of each demographic item. For instance, if a block group’s proportion Black was 21% or greater it would have had a greater Black proportion than at least 90% of the other blocks in this study.
p Values are for likelihood ratio χ2 tests unless otherwise noted. Of a total of 467 stores, 48 failed.
Mean (%) Demographic Characteristics for the Stores’ Block Groups
Table 3 shows that we found no bivariable significant association between any of the advertising characteristics that we selected as likely predictors and store failure. Additionally, we tested whether stores that were within 400 m of a school were more likely to fail their compliance check and whether count of total ads or ads for menthol brands predicted failure, but these associations were not significant (data not shown). Due to there being no significant bivariable associations, we did not build multivariable models.
Associations Between Advertising Characteristics and Compliance Check Failure.
Note. aContinuity corrected test due to small cell counts.
p Values are for likelihood ratio χ2 tests.
Discussion
Contrary to our main hypothesis, we found no association between tobacco point-of-sale marketing and compliance check failure. Aside from our finding that there was a positive association between store failure and percent of the store’s census block group population that was Hispanic, there were no other associations between demographics and store failure. Tobacco shops were the type of store that was most likely to fail a compliance check with four of nine tobacco shops in our sample failing.
The association between Hispanic population and compliance check failure was only marginally significant. Given number of tests done for this study, we take a conservative interpretation of this finding and feel it may be spurious. We have considered several explanations for our null findings in addition to the possibility that tobacco advertising may truly not be related to sales to youth. It is possible that certain advertising strategies are actually associated with sales to youth, but that by the time we collected these data in 2007, those advertising methods had become uncommon. For instance, only three stores in our study had placed ads less than a foot away from toys or child items; only one store had exterior advertising for tobacco-branded merchandise, and thus we were unable to tests associations between these items and compliance check failure. Roeseler et al. found that in 2000 nearly 80% of stores in Los Angeles had ads at a child’s eye-level but by 2008 the number was down to about 30% (which is about the prevalence of ads less than or equal to three feet from the ground that we found; Roeseler, Feighery, & Cruz, 2010). The low level of the type of tobacco marketing that is likely to arouse interest in youth that we saw in our study in 2007 would appear to lend support to what Roeseler et al. reported.
Another possible explanation for why our hypotheses did not bear out relates to the compliance check procedure. Some have argued that the compliance checks are not a very valid measure of commercial tobacco accessibility for minors (DiFranza, Savageau, & Bouchard, 2001). Adolescents who wish to purchase cigarettes are savvy and choose outlets that they believe are less likely to ask for age ID and more likely to sell cigarettes to them (DiFranza et al., 2001). Or they may do things like purposefully dress to appear older, use a fake ID, or adopt other such strategies which are explicitly not to be used during compliance checks. Given these weaknesses in compliance checking as a measure, our study cannot answer the question of whether real youth, in real circumstances, would be more likely to successfully purchase tobacco in stores with certain advertising environments.
Finally, the lack of associations that we saw could speak to the success of tobacco retailer licensing in Minnesota. Only vendors with a current license can sell tobacco in the state. Licenses can be rescinded if stores repeatedly violate tobacco policy like youth access laws. For stores that earn substantial amounts of revenue through the sale of tobacco, the threat of losing a license might be more motivating than incentives from tobacco representatives to sell more tobacco, even in lower income or minority communities. However, this is a hypothesis that was not able to test and we cannot say whether our other researchers would see similar results in a state that does not license retailers.
The strengths of the present study are the large sample size of stores and the completeness of the data. Also, to our knowledge, this is the first study to examine associations between advertising and youth access compliance. We were limited in that we could not test whether a change in advertising leads to a change in compliance check failure since this study was cross-sectional. We were also limited by the reality that our study area has less racial/ethnic diversity compared to other urban centers. This may be a reason why we might not have been able to see an association between most of the demographic characteristics and compliance check failure. Additionally, related to the limited geographic area, it is possible that these establishments were supplied by a relatively small number of distributors and we may have had minimal variability in styles of promotion and incentives offered by distributors. However, this was not something that we measured. While we did not formally test reliability by having establishments assessed by more than one assessor, when comparing the ratings of all establishments, there was no significant difference in ratings between the three assessors. Despite this, the unquantified reliability of the store assessment is a limitation. Finally, we used Year 2000 US Census data to characterize the block groups, but the stores were assessed in 2007. It is possible that some census block groups’ demographics were mischaracterized because of this.
For the professional engaged in health promotion, it is important to note that while we saw no differences in compliance based on the factors we examined, commercial access (purchasing tobacco from stores) has been associated with greater tobacco use in youth (Forster et al., 2003; Harrison et al., 2000; Widome et al., 2007) and there is some evidence that compliance check enforcement can not only lessen commercial access for youth but is also associated with a lower youth smoking prevalence (DiFranza, Savageau, & Fletcher, 2009) and, thus, is important to maintain stores’ high compliance to age-of-sale laws. Additionally, for those professionals engaged in the implementation of tobacco policy, tobacco shops appeared to be a weak link in enforcing age-of-sale laws and perhaps deserve greater enforcement. With the new Family Smoking Prevention and Tobacco Control Act of 2009 (FDA regulation; The Family Smoking Prevention and Tobacco Control Act (FSPTCA), 2009), states now have the power to regulate the time, place, and manner (but not content) of point-of-sale advertising, which could limit youth exposure to tobacco marketing even further and change the point-of-sale landscape in ways that deter tobacco use and thus promote health.
Footnotes
Acknowledgments
The authors would like to thank Amanda Rudolph and Katie Engman for their work on this project. This material is the result of work supported with resources and the use of facilities at the Minneapolis VA Medical Center. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the U.S. government.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported through a grant from the Robert Wood Johnson Foundation’s Substance Abuse Policy Research Program (grant 65165) and the Minnesota Department of Health.