“If You Just Tell Me You’re 18, I’ll Still Sell to You”: A Qualitative Study of Underage Tobacco Product Purchasing Experiences in a Tobacco 21 Compliance Study

Abstract

Background

Tens of thousands of underage tobacco buy attempts are conducted each year for research, compliance, and public health surveillance. However, little research has qualitatively examined the perceptions and experiences of underage buyers participating in these programs. We sought to understand underage buyers’ experiences and gather recommendations for protocol improvements.

Methods

We conducted semi-structured interviews in the fall of 2022 to assess experiences with underage tobacco product purchasing. Participants (N = 19, 58% male, 42% White) were research assistants aged 18–20 in New Jersey, New York, or North Carolina. Interviews examined purchasing experiences in relation to store characteristics, clerk interactions, and buyer identities. We used deductive and inductive thematic coding to explore key themes related to buyer experiences.

Results

We identified four themes: (1) non-chain stores lacked consistency in verifying age; (2) female data collectors experienced uncomfortable situations more frequently than male data collectors; (3) not identifying with the store’s typical demographics impacted purchase attempts; and (4) participants suggested improvements for inspections and research during training.

Discussion

Retailer education, widespread adoption of ID scanners, and enforcement could increase standardization of ID requests and verification. Male and female buyers can be trained on what they might expect based on their gender, as well as how to maneuver through unwanted situations. Consideration of shared identity is important for future waves of data collection and research. Efforts to improve training include more extensive mock purchase training with supervisors well-versed in this area.

Implications

Electronic ID verification and promoting compliance at non-chain retailers could impact access to tobacco products for underage buyers. Training for underage buyers in research and compliance assessments should focus on ways to enhance data collectors’ confidence when making a purchase attempt, which may improve the validity of the rate of sales to individuals under 21.

Introduction

Reducing young people’s access to tobacco products remains an important part of evidence-based youth tobacco prevention strategies,¹ and key to their effectiveness is enforcement.² Tens of thousands of underage purchase attempts for compliance take place each year in the United States as part of (a) the Food and Drug Administration (FDA) Tobacco Retailer Inspections Program,³ (b) state enforcement to comply with the Synar Amendment,⁴ (c) local enforcement, and (d) mystery shopper programs required by legal agreements between retailers and enforcement agencies.⁵

A 2016 systematic review summarized 47 studies assessing buy protocols and the characteristics of buyers on purchase attempt outcomes. It found that mimicking real-world behaviors of youth drastically changed results.⁶ Research on underage purchase attempts has found that older buyers and female buyers had higher success rates, female clerks were more likely to sell, and most successful purchase attempts were in suburban and rural areas.⁷ Other research employing underage buyers has suggested that outcomes differ by the race of the buyer.^8,9 Researchers have also posited that a “familiarity effect” may occur when repeated attempts by the same buyer to the same retailer are made; repeated visits may indicate that the buyer is from the neighborhood, thereby increasing the likelihood of an underage sale.¹⁰

Qualitative studies of youth and young adults reporting access to tobacco products outside of research studies have been reviewed by Nuyts et al.,¹¹ and a recent paper explores young adults’ self-reported access to e-cigarettes.¹² A 2022 study looking at 140 countries and territories concluded that a commercial source was one of the main sources for obtaining cigarettes for adolescents who smoke followed by private exchanges.¹³ However, the literature on underage buyers’ perceptions of their own experiences participating in research purchase attempts is sparse. Two studies have examined only likelihood of smoking among minors completing purchase attempts.^14,15 The underage buyers in these studies indicated they were aware of how easy it was to purchase these products. Among buyers in a 1995 study who did not smoke before they began attempts, participating in cigarette purchasing research did not appear likely to increase their intention to use tobacco products.¹⁴ And, in a separate 2002 study, no data collectors initiated smoking at follow-up.¹⁵ These two studies are decades old, and no research that we are aware of has explored underage buyers’ experiences qualitatively or among underage buyers aged 18-20 participating in purchase studies.

In the last few years, purchase studies increasingly included older ages to reflect higher legal sales ages. The US federal minimum age of tobacco sales was changed in 2019 from 18 to 21 years. This change was designed to reduce rates of smoking in young people and address the epidemic of vaping among youth. Evidence suggests that increasing the age of sale will decrease the provision of tobacco products to younger youth from older peers.¹⁶ Nationally, the FDA started using older (age 18–20) buyers in September 2021 to test compliance.¹⁷

Thus, we sought to bring the expertise of underage buyers aged 18–20 participating in research into the literature. Specifically, the purpose of this study was to inform research and training protocols by assessing: (1) how underage buyers experience purchasing tobacco products; (2) differences by retail store types; and (3) differences in these experiences based on buyer identity.

Methods

This study was embedded within a larger project led by Rutgers University with sites in New Jersey, New York, and North Carolina. Underage buyers tested how well stores were identifying (ID’ing) persons aged 18–20 and refusing to sell tobacco products to underage buyers.¹⁸ All three states have vertical orientation for drivers’ licenses and state ID cards for cardholders under 21.¹⁹ The protocol for the study reported here was reviewed and designated as exempt research by the East Carolina University and Medical Center IRB (#22-001057).

Eligibility

To be eligible for this study, participants had to have been data collectors aged 18–20 during their employment and from one of three sites of the larger research project (Rutgers University in New Jersey, Columbia University in New York City, and East Carolina University in North Carolina). Data collectors currently or previously employed by the project with at least one purchase attempt completed were eligible.

Recruitment

To recruit participants, supervisors from each team forwarded a recruitment letter via email or provided a list of eligible data collectors to contact. Additional follow-up e-mails purposively solicited participation to balance site, gender, and racial/ethnic diversity. Of the 53 data collectors employed across the project, 19 (35.8%) agreed to participate in interviews. Participants were compensated with a $20 gift card. As shown in Table 1, each site had at least six members participate.

Table 1.

Underage Buyer Interviewee Characteristics, 2022, N = 19

Site	Total number of buyers	Minoritized racial/ethnic identity*	Female gender identity
	N	n (%)
Columbia	6	3 (50.0%)	3 (50.0%)
East Carolina	7	3 (42.9%)	3 (42.9%)
Rutgers	6	5 (83.3%)	2 (33.3%)

^*Minoritized racial/ethnic identities included three identifying as Black, four identifying as Asian including South Asian, two identifying as Middle Eastern/North African, and three identifying as bi or multi-racial.

Virtual Interview and Interview Guide

The first author conducted individual virtual interviews between September and November 2022 and audio-recorded conversations. Interview length ranged between 13 and 31 minutes, with an average of 18.5 minutes. Interview questions asked about typical purchasing experiences and any interactions that were considered atypical. Interviewees were questioned about the impact their experiences during purchase attempts had because of both external factors such as store type and location, as well as internal factors such as race, ethnicity, and gender. The final questions asked for advice on training for future data collection and advice for improving store compliance. Data saturation began to appear at the 17th interview, which was confirmed by interviews 18 and 19. The full interview guide is available as a Supplementary File. Recordings were professionally transcribed.

Data Analysis

Data were analyzed using inductive and deductive thematic coding²⁰ to reach theme crystallization.²¹ First, the primary analyst iteratively read all the transcripts to immerse himself in the data and create codes for data that appeared across multiple interviews. Second, the primary analyst met with the secondary analyst to discuss, refine, and identify relationships between the initial codes where codes could be combined. Third, the results were discussed with a larger team to check their perceptions of the data against the coding results. Fourth, we finalized the key themes that emerged and selected representative examples. We used NVivo v12/PC software. Regarding positionality, the primary analyst (TAW) was identified as a White male and was a previous underage buyer as part of the research team. The secondary analyst was a professor of health education and promotion (JGLL) who identified as a gay White man with a background in tobacco prevention and control research. The primary and secondary analysts met, discussed, and reflected on identity and experience across the study design, data collection, and analysis phases of the project to forefront how they could influence data collection (ie, interviewee responses) and interpretation (ie, theme crystallization).

Results

We identified four themes: (1) “They would just sell it”: non-chain stores’ lacked consistency in verifying age; (2) “I got catcalled”: female data collectors experienced uncomfortable situations more frequently; (3) “I was completely out of place”: data collectors not fitting the description of the store’s clerks and customers impacted visits; and (4) “How confident you look”: participants suggested improvements for inspections and research through training.

Theme 1: “They would just sell it”: non-chain stores lacked consistency in verifying age

Participants reported distinct differences in their experiences attempting purchases and how age verification was approached based on the type of store. Specifically, participants reported chain stores had standardized or similar routines where data collectors were often treated similarly. These stores were more likely to have a standardized procedure for age verification and use electronic age verification scanners. Data collectors described their general experience in chain stores as being treated in a predictable way:

I’d say every time I went to a [major drug store chain] it was the exact same thing. They would just take my ID, scan it, no, leave. (Participant 5, Female)

It doesn’t matter who you are, what you look like, what you say, they’re going to scan your ID. And then the computer doesn’t lie. It’s just going to tell them no. (Participant 3, Male)

Non-chain stores did not create as routine of an environment. Participants reported that they were not able to predict age verification in non-chain stores. There was more variability in individual clerks’ approaches to age verification, and participants reported they were more likely to complete a purchase in non-chain stores.

The bodegas [small convenience stores common in urban areas] -- it was more like they would either sell it or they would check it and just hand me back the ID; almost like kids might go into bodegas all the time to try to get tobacco products because that’s where they’re more likely to sell it. I kind of got that impression that they knew what to expect or they would just sell it, yeah. (Participant 10, Male)

The bodegas or the pop-up stores like non-chains, those are mostly where I get the “how old are you,” “do you need to see my ID,” “how old are you,” that back-and-forth game especially when it’s run by an older guy. (Participant 19, Male)

When asked about these differences in the process between chain stores and non-chain stores, participants indicated they thought that scanners were not used in non-chain stores nearly as frequently as chain stores but should be to keep the process consistent and uniform.

I think they [non-chain stores] should implement more of those scanners that they have for ID’s just because those stores that you have to scan it, it would always say no. It would always pop up that I was underage, and I would consistently get turned down. (Participant 2, Female)

Nonetheless, scanning the ID is not a perfect process. Buyers reported that a few clerks overrode the computer system, granting a sale even when the buyer was under the age 21.

I went to the [dollar store chain]… She scanned the item and then she scanned my ID, which usually is an indication that I’m not going to get anything because the computer doesn’t lie. So, the computer beeped at her… telling her that the purchase should not go through. She looked at me, she looked at the computer, she looked at my ID, and then she tells me, “These dang computers are dumb”, and overrides the system and lets me buy it. (Participant 3, Male)

Theme 2: “I got catcalled”: female participants experienced uncomfortable situations more frequently

The interview addressed the participants’ identities and how they felt it made an impact in two ways: how often they were sold to and their interactions with individuals they encountered during purchase attempts. Participants expressed how their experiences varied based on their own gender and racial/ethnic identity, as well as shared identities with store clerks. In our analyses, gender was the most prominent factor in patterning participants’ experiences. Data collectors noted that customers hanging out in or around the store could make purchasing more unpredictable, especially for women. Women specifically would face occasional uncomfortable interactions with other customers and occasionally clerks, with participants sharing experiences, explaining,

I feel like the girls typically have the crazier stories than the guys… I got catcalled… Some guy was like, “Let me buy you something. Let me buy you something…” definitely [being] a woman in sketchier places made a difference I think. (Participant 1, Female)

…there was one time where a cashier explicitly was like, “Oh, you’re so pretty, and I want you to come back. I’ll give you a free lighter.” (Participant 14, Female)

Data collectors also reported that the gender of the clerk would have an impact on both sales and interactions, specifically when it was a male clerk and a female buyer. Participants shared their thoughts on why, hypothesizing,

I think it’s because older men kind of see younger girls as being more pure. I’ve always gotten that sense. And so they’re really hesitant selling something that is considered to be bad or negative to a girl as opposed to a guy. (Participant 13, Male)

I think a lot of clerks more on like the male side always question if the female is going to buy cigarettes rather than a male I think. I think it’s more stereotypical that girls don’t smoke as much cigarettes, cigars or like tobacco free nicotine pouches than males… (Participant 8, Female)

Theme 3: “I was completely out of place”: not fitting the description of the store’s typical demographics impacted purchase attempts

Similarly to theme 2, race and ethnicity were reported to impact sales and interactions. Participants generally reported these impacts being dependent on the location of the store. If a store was in a neighborhood where there was a predominant race/ethnicity, an underage buyer of a different race than that of the predominant one may have a different experience.

…of course, in the country type stores I was more likely to be not looked at weird, or accepted, than if I went to a store of all Black people and I was the one White girl. But I think that would be the general case for anyone walking into a predominantly Black or White area of a different race. (Participant 18, Female)

I’d say, first off, my skin color because if I’m in a different area, like I said, inner city area, they think I’m just this random kid off the street that’s just looking to get a high or something. And then I’d say maybe worse in the suburban area … I’d be more likely to get ID’d and “What’s this kid about? What’s his deal?” Certain places were different based off of that. (Participant 15, Male)

Participants also noted that shared sociodemographic characteristics between themselves and clerks influenced not only perceptions but also experiences. Within the context of shared race/ethnicity, those factors would mostly result in interactions that could facilitate sales. However, participants reported it was more complex than simply sharing identities. There were possibilities of interactions that could decrease the likelihood of an underage sale as well as causing a more uncomfortable interaction. Participants reported that generally sharing characteristics with the buyer created a situation where an underage sale became more likely or at least resulted in friendly conversations.

So a lot of [location] is Hispanic and immigrants and I’m Hispanic … so it’s like when you get in a conversation and you can understand what they’re saying and then kind of speak back to them, I think they feel more inclined to be nicer to you and to be able to help you more and I think that definitely helped me in the beginning. (Participant 8, Female)

But the only time our shared experiences made it a little bit difficult for me to buy was when I walked into stores where the store clerks were Indian. I’m part [regional ethnicity] myself, and part [regional ethnicity] and Indian people can recognize that this guy is kind of [from] a region basically. … And most stores that I walked into where the store clerk was Indian, I was unable to buy. And I felt that they were looking at me a lot more closely than they would at someone from a different race or different background. (Participant 11, Male)

Another example of impacted experiences from shared factors were with age. Younger clerks were more likely to strike up a casual conversation with underage buyers and even bypassing checking IDs in some cases. One participant remembered his experience, telling us,

…with the age group it was almost like there was this level of comfortability between you two because of that shared factor that you were able to engage in conversation and kind of warm up to each other so that the sale could have gone easier for you. (Participant 16, Male)

Theme 4: “How confident you look”: Participants suggested improvements for inspections and research during training

Participants expressed the importance of including colloquial language, slang terms for tobacco products, and confidence in their training. Participants reported that it was important to ensure they knew the modern terminology of how people ask for certain products, such as tobacco-free nicotine pouches. They also suggested that an effective strategy to improve confidence was mock purchase attempts within teams, focusing on slang terms, pricing, packaging, and nicknames for certain products. Participants were asked at the end of the interviews if they had advice for future data collectors, and responses had a common theme, for example,

I think one of the key factors to determine whether or not you have the attempt [be] successful is really about how confident you look or can you pronounce all these products really fluently and when they ask you more questions, are you able to answer them right away and also are you familiar with the price of each product. (Participant 6, Female)

When asked for feedback regarding training, participants appreciated practice attempts so they could get a feel for the full process. There were also multiple people who reported that having supervisors or trained data collectors going with trainees to ride-along for their practice attempts was helpful. Those who did not have that opportunity found it would have been helpful had they received it.

I think the experience was most beneficial, especially the five stores in the beginning. To kind of go in and figure it out by yourself. There’s no way to really know what’s about to happen and what the experience will be like until you actually do it. (Participant 2, Female)

I think the buddy system we have now, how someone can request somebody ride along with them, I think that’s real nice. I definitely would have appreciated it when I was first starting out. (Participant 3, Male)

Discussion

Principal Findings

Our key takeaways from these interviews are four themes that indicate: (1) non-chain stores did not follow the same streamlined process as chain stores, causing more variability in how clerks ask for and assess ID; (2) female buyers received more unwanted attention from others in and around stores than their male counterparts; (3) shared factors relating to race and age mostly created interactions that might increase the likelihood of an underage sale; and (4) training to build confidence and learn product slang allowed the data collection to be more authentic. A cross-cutting finding was the lack of standardization during age verification procedures, which buyers thought might cause their identity to become more of a factor during purchase attempts as opposed to age only.

The first and third themes indicate that, given buyers’ experiences with non-chain stores, more widespread adoption of ID scanners and clerk training could increase standardization in ID requests and verification. ID scanners solve the issue of clerks not being able to tell from the birthdate if the customer is 21, remind clerks who forget that customers aged 18–20 are unable to purchase tobacco products, and may prevent racial, ethnic, or gender bias in ID requests. Indeed, prior research shows that electronic verification of IDs or use of ID scanners is associated with lower likelihood of underage sales.^22,23 Adoption of ID verification and clerk training can be part of evidence-based comprehensive community interventions.¹

The instances of unwelcomed interactions of male customers towards unaccompanied women identified in the second theme may not be a surprising finding to readers who identify as women, but it is relevant for training and safety protocols. Buyers can be trained in what they might expect based upon their gender, as well as how to maneuver through unwanted situations. Understanding and recognizing this harassment may be an important developmental skill for male data collectors to increase awareness of gender-based harassment and promote self-reflection. Data collection teams may want to offer a ride-along by another data collector and should ensure that safety procedures and training recognize the likely exposure to unwanted attention.

Our findings in theme four provide important recommendations for training programs used in inspections and research projects. Data collectors highlighted the importance of slang terms for products and phrasing of questions to the clerk. Efforts to improve these measures can include further mock purchase training with supervisors well-versed in this area, as well as having a supervisor participate in a ride-along during practice attempts so that data collectors can debrief after each attempt before starting their real purchase attempts.

Results in Context

This paper represents, to our knowledge, the first qualitative examination of underage researcher buyer experiences and is also the first to explore underage buyer experiences in the context of Tobacco 21 policy enforcement. Validity can be tied to the familiarity of a customer with products and the ability for a customer to appear like someone who uses tobacco products.⁶ Inspection programs should incorporate these two factors within protocols to maximize their validity. Regarding identity, prior research has indicated racial disparities in tobacco sales with Black minors having greater likelihood of successful purchases.^8,9 However, our qualitative work did not directly assess racial inequities, and we did not directly identify support for the “familiarity effect”¹⁰ where the same buyer returning to the same store was more likely to complete a purchase, ostensibly because of being perceived as a resident of the neighborhood. Our participants reported a more complex relationship with store clerks involving identity and context in ways that often increased the likelihood of an underage sale and occasionally did the opposite.

Strengths and Limitations

This study had several strengths. These included a qualitative approach to explore experiences and engagement with underage data collectors to learn what they recommend for training programs. Additionally, this study included data collectors from multiple settings that spanned urban (New York City) to relatively rural (Pitt County, NC) and multiple racial/ethnic and gender identities. The primary analyst was himself a data collector. Limitations included a focus solely on US east coast buyers that were a part of a single research study. Participating data collectors may not represent the experiences of non-participating data collectors. Because data collectors were recruited from all waves of data collection dating back to 2019, each participant’s recollection of events may have varied depending on the time of their most recent purchase attempt.

Conclusions

As tobacco use remains a leading cause of preventable disability and death, it remains critically important to enforce underage access policies to maximize effectiveness.² This study provides insight into the experiences of underage buyers in ways that can help improve research and interventions. This study offers important considerations for the design and implementation of inspection programs as well as research protocols to maximize their impact. Emphasizing variation of training dependent upon the identity of each data collector can prepare them for interactions that gender, race, or ethnicity may play a role in. These considerations can improve efforts to stymie underage access to tobacco products.

Funding

Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number R01CA231139 and by an Undergraduate Research and Creative Activities award from East Carolina University. AYK was additionally supported by the National Cancer Institute of the National Institutes of Health (P30CA225520) and the Oklahoma Tobacco Settlement Endowment Trust (TSET R23-02). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or funders.

Declaration of Interests

JGL has a royalty interest in a retailer mapping and audit software licensed by the University of North Carolina. The software was not used in this project. AYK serves as a paid expert consultant in litigation against the tobacco industry.

Author contributions

Tyler West (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Methodology [supporting], Project administration [lead], Writing—original draft [lead], Writing—review & editing [supporting]), Josephine Awadalla (Conceptualization [supporting], Formal analysis [supporting], Investigation [supporting], Methodology [supporting], Validation [supporting], Writing—original draft [supporting], Writing—review & editing [supporting]), Christopher Ackerman (Conceptualization [supporting], Formal analysis [supporting], Investigation [supporting], Methodology [supporting], Resources [equal], Supervision [equal], Validation [equal], Writing—original draft [supporting], Writing—review & editing [equal]), Mahdi Sesay (Conceptualization [supporting], Data curation [supporting], Formal analysis [supporting], Investigation [supporting], Project administration [supporting], Supervision [supporting], Writing—review & editing [equal]), Mary Hrywna (Conceptualization [supporting], Funding acquisition [supporting], Investigation [Supporting], Validation [supporting], Writing—review & editing [supporting]), Daniel Giovenco (Conceptualization [supporting], Methodology [supporting], Resources [supporting], Validation [supporting], Writing—review & editing [supporting]), Amanda Kong (Conceptualization [supporting], Investigation [supporting], Methodology [supporting], Validation [supporting], Writing—review & editing [Supporting]), and Joseph Lee (Conceptualization [supporting], Data curation [supporting], Formal analysis [supporting], Funding acquisition [supporting], Investigation [supporting], Methodology [equal], Project administration [equal], Supervision [equal], Validation [supporting], Writing—original draft [supporting], Writing—review & editing [equal])

Data Availability

Data from this qualitative study are not publicly available to minimize risk from deductive disclosure.