Would banning menthol cigarettes, flavored cigars, and flavored e-cigarettes prompt users to seek illicit channels for obtaining them in the United States?

Abstract

Background:

Banning flavors in tobacco and nicotine products may reduce youth initiation and prompt quit attempts but such bans may lead to illicit markets. We examined how likely current users would be to seek flavored products from illicit channels under various ban scenarios.

Methods:

Cross-sectional surveys of 2,552 current users of menthol cigarettes or flavored cigars and 2,347 users of flavored e-cigarettes were conducted between 2021 and 2022 in the United States. For each ban scenario, respondents reported if they would have intentions to seek the banned flavored products from any illicit channels and identified the specific illicit channel they would consider. Logistic regressions were used to estimate how the likelihood of having intentions to seek illicit channels was associated with demographics, ban scenarios, and status of tobacco use.

Results:

Under various ban scenarios, 24–30% of people who smoked said they would seek illicit channels to obtain the banned products compared with 21–41% of dual users and 35–39% of users of flavored e-cigarettes. Online retailers were favored by people who smoked while users of flavored e-cigarettes favored local retailers. Heavy users were more likely to say they would try illicit channels. Under bans restricting more types of flavored tobacco products, users would be less likely to try illegal channels.

Conclusions:

A significant proportion of users of flavored tobacco and nicotine products would not reject using illicit banned products. Tailored programs are needed to apply to the groups with a higher risk of seeking illicit channels for banned products.

Background

Restricting and banning the characterizing flavors in tobacco and nicotine products are key tobacco regulatory strategies for encouraging cessation and preventing initiation, especially among young adults and youth, and low-income and African-Americans [1–6]. In 2009, all characterizing flavors except menthol and tobacco were banned by the U.S. Food and Drug Administration (FDA) in cigarettes. Since 2019, some states (e.g., Massachusetts, New Jersey, New York) and municipalities (e.g., San Francisco) implemented further flavor restrictions on tobacco and nicotine products [7]. In 2020, the FDA implemented a policy to ban all capsule or cartridge-based e-cigarettes with added flavors other than menthol or tobacco. In April 2022, the FDA issued proposed rules to ban menthol in cigarettes and all characterizing flavors other than tobacco in cigars nationwide, but the rules are still not finalized yet [8, 9].

Despite these potential benefits, one of the negative concerns with flavor bans, particularly argued by the tobacco industry is that flavor bans may spark the creation of substantial illicit markets from which users of flavored tobacco products will seek to obtain banned products [10–13]. In 2023, the U.S. Customs and Border Protection intercepted a growing number of illicit shipments of e-cigarettes [14]. If established, illicit markets and related activities could cause several harms. First, products from illicit markets tend to be produced with low standards and thus are more harmful to health than products from the regulated market [15]. Second, the sale of tobacco and nicotine products to youth, whether illicit or not, would likely increase because the minimum age laws do not exist on the illicit market [16]. Third, an illicit market means business decreases for legal tobacco retailers and convenience stores and a loss of government revenue including tax revenue [17, 18]. For example, Massachusetts witnessed a decline in tax revenue and an increase in the illicit tobacco market following the state’s ban on flavored tobacco in 2020 [19]. Fourth, illicit tobacco markets and activities may facilitate other crimes and provide funding for additional criminal activities [20].

The argument for the creation of illicit markets may be exaggerated because there are several practical constraints to developing any substantial, new illicit market in the U.S. regarding where the illicit flavored tobacco products would come from, how they would be distributed and sold, how potential consumers would know where to obtain them, and how illicit markets would evade enforcement [21]. According to the FDA’s review [22], flavor bans are unlikely to significantly impact the illicit market. For example, there was no evidence that an illicit market for menthol cigarettes emerged in Canada following a ban on menthol cigarettes gradually implemented throughout Canada between 2015 and 2018 according to the number of illicit cigarettes seized and self-reported data on cigarette uses and purchasing [23, 24].

However, available empirical evidence is limited and mixed. For example, increased seized illicit flavored e-cigarettes, cigars, and menthol cigarettes in Massachusetts after a ban on the sale of menthol cigarettes and flavored e-cigarettes in June 2020 may indicate that some consumers of flavored tobacco products responded to the ban by turning to illicit markets [19, 25]. In San Francisco, a small proportion of young adults (5.3%) reported acquiring flavored tobacco products illegally after the implementation of a comprehensive ban on such products [26]. The prohibition of a product can often enhance its attractiveness, referred to as “forbidden fruit” [27], thus it is plausible that flavor bans increase demand for illegal tobacco and nicotine products [28, 29]. Overall, our understanding of the demand side is limited and it is unclear how people who currently use added-flavored cigarettes, cigars, and e-cigarettes would respond if flavored products were banned. Some might be prompted to quit all use, others might switch to still-legal versions of the tobacco products they consume, and others might switch to flavored versions of other tobacco products still legally available. But how many would be willing to seek out and use illicit flavored products and do so is not yet clear.

The main channels consumers might use to obtain prohibited flavored tobacco products would be illegal online sellers, with purchases delivered by mail or common carriers, purchases of illicit products from otherwise legally operating retailers, or illegal market sales such as illicit street sales at temporary public locations [30]. Some studies suggest that those who are more dependent (e.g., daily users compared with less frequent users) would be more likely to seek illicit products [11, 31–33], and it appears that individuals with more education might be more likely to engage in illicit purchasing of tobacco products than their counterparts with less education [31]. However, existing knowledge of which population subgroups among users of flavored tobacco products would be most likely to seek out illicit flavored products after they are banned is quite limited.

In this exploratory survey-based study, we examine how likely adults who currently use menthol cigarettes, flavored cigars, or flavored e-cigarettes would be to seek banned products from illicit channels if faced with different types of flavor bans, which adult subgroups would be most likely to do so, and which illicit trade channels they would be most likely to explore.

Methods

Data were collected on Amazon Mechanical Turk (MTurk) [34] between December 2021 and May 2022 from respondents who smoked flavored tobacco products and those who used flavored e-cigarettes. The appendix contains quality control data for the survey. The Institutional Review Board at the University of Memphis approved this study. The first group constituted adults who currently smoked menthol cigarettes or flavored cigars at least once a week in the past 30 days (N=2,552), and may or may not have used e-cigarettes concurrently. Respondents’ demographics and several variables related to tobacco use were collected. The second group constituted adults who currently used flavored e-cigarettes (N=2,347), in either menthol or other flavors such as fruit and candy, at least once a week in the past 30 days. Adults in this group may or may not have used tobacco or non-flavored e-cigarettes concurrently but did not smoke cigarettes or cigars or use other major tobacco products such as heated tobacco products or water pipes. Besides demographics, we collected respondents’ e-cigarette use status (intention to quit, use frequency, history, flavors, and devices), and whether they had previously smoked or never.

Five hypothetical ban scenarios were presented, including (1) E1: banning e-cigarettes with any added flavors except menthol or tobacco; (2) E2: banning e-cigarettes with any added flavors other than tobacco; (3) C: banning menthol cigarettes and all added-flavor cigars (with all flavored e-cigarettes remaining available); (4) E1C: banning e-cigarettes with any flavors except menthol or tobacco, menthol cigarettes, and all added-flavor cigars; and (5) E2C: banning e-cigarettes with any flavors other than tobacco, menthol cigarettes, and all added-flavor cigars.

To gain insights into possible illicit flavored use under each relevant ban scenario (i.e., the respondent was using one or more products that would be banned under the specific scenario), we asked respondents to assume that the banned flavored products would no longer be legally available from local retailers or other legal channels. We then asked them if they would seek to obtain these products from any illicit channels or not. Those who responded affirmatively were asked to identify the illicit channels they would try, including (1) local retailers who were illicitly selling banned products, (2) online sellers who violated the ban, and (3) illicit market sellers such as illegal street sellers.

We first summarized the percentages of respondents who endorsed that they would seek any illicit channels and try each of the three illicit channels aforementioned, under each relevant ban scenario and stratified for people who smoked and users of flavored e-cigarettes. Considering the non-representativeness of our sample, weights were added to ensure the demographics (gender, age, race/ethnicity, educational attainment, and household income) and tobacco use status of respondents match with the corresponding groups from the 2018–2019 Tobacco Use Supplement of the Current Population Survey (TUS-CPS) [23]. For the first group (i.e, the current users of menthol cigarettes or flavored cigars), the variables used for weighting included if smoked daily, if concurrently used e-cigarettes, if concurrently used non-flavored cigarettes or cigars, and if used menthol cigarettes, flavored cigars, or both. For the second group (i.e., the current users of flavored e-cigarettes), the variables used for weighting included if used e-cigarettes daily, if concurrently used non-flavored e-cigarettes, if used menthol e-cigarettes, other flavored e-cigarettes except menthol, or both menthol and other flavors, and if smoked before. These variables were chosen because they differentiate smokers and users of e-cigarettes into distinct groups that may be differently affected by various flavor bans [33, 35–37]. Second, logistic regressions (PROC LOGISTIC in SAS, version 9.4) were used to estimate how the likelihood of trying illicit channels was associated with ban scenarios, demographics, and their status of tobacco and nicotine use.

Results

As shown in Tables A1 and A2 in the Appendix, the majority of the respondents in our study were young adults and had a high level of educational attainment. After weights were added, the demographics and the status of tobacco and nicotine use of both those who smoked flavored tobacco products and those who used flavored e-cigarettes were consistent with the corresponding patterns in the sample of the 2018–2019 TUS-CPS. Overall, the majority of people who smoked smoked daily, either somewhat or not intending to quit smoking, used menthol cigarettes, not concurrently used non-flavored cigarettes or cigars, or any e-cigarettes. The majority of those who used flavored e-cigarettes either somewhat or not intended to quit using e-cigarettes, used e-cigarettes daily, used any flavors except menthol and had smoked before.

Table 1 shows that under various flavor ban scenarios, 24–30% of those who smoked flavored tobacco products had intentions to seek out illicit channels to obtain the banned products compared with 21–41% of dual users (i.e., those who smoked flavored tobacco products and also used e-cigarettes) and 35–39% of those who used flavored e-cigarettes. Those who smoked flavored tobacco products endorsed online sellers (14–16%) as the channel they would most likely use to try to obtain the banned products although the difference among the three channels was small. Users of flavored e-cigarettes mainly endorsed local retailers (24–28%), much more favorable than the other two channels. For most users under most scenarios, illicit market sellers would be their least favorable channel compared with the other two. Roughly, fewer users would seek illegal channels under bans restricting more than one type of flavored tobacco product than a ban that restricts one type. For example, about 30% of people who smoked would see illicit channels under ban scenario C compared with about 24–25% under ban scenarios E1C and E2C.

Table 1.

The percentage of respondents who endorsed seeking illicit channels for banned tobacco and nicotine products under various flavor ban scenarios, stratified by people who smoked but did not use e-cigarettes, those who smoked and also used e-cigarettes, and users of flavored e-cigarettes, among adult users of flavored tobacco/nicotine products (USA, 2021–2022)

	Ban scenario	Intention to seek illicit channels for banned products
		Local retailers	Online retailers	Illicit market sellers	Any illicit channels
Relevant people who smoked but did not use e-cigarettes (N=1,473)	C	16.2	16.1	16.3	30.2
	E1C	11.7	14.0	11.3	24.6
	E2C	12.5	14.7	11.4	24.4
Relevant people who smoked and also used e-cigarettes (N=1,080)	E1	37.2	15.1	31.1	40.8
	E2	19.2	25.4	15.0	33.4
	C	18.2	15.9	14.9	23.7
	E1C	13.6	12.6	12.3	20.5
	E2C	15.6	21.5	12.5	29.3
Relevant users of flavored e-cigarettes (N=2,347)	E1	28.4	23.4	20.2	39.0
	E2	25.2	23.3	19.2	38.3
	E1C	25.3	20.7	20.6	35.1
	E2C	24.3	22.9	19.4	36.8

Ban scenarios:

• E1:
  ban on e-cigarettes with any flavors except menthol;
• E2:
  ban on e-cigarettes with any flavors including menthol;
• C:
  ban on menthol cigarettes and flavored cigars, with flavored e-cigarettes remaining available;
• E1C:
  ban on e-cigarettes with any flavors except menthol, menthol cigarettes, and flavored cigars;
• E2C:
  ban on e-cigarettes with any flavors including menthol, menthol cigarettes, and flavored cigars.

Relevant respondents: not all people who smoked and users of flavored e-cigarettes would be impacted by all five ban scenarios. For example, only those who smoked methol cigarettes or flavored cigars and also used flavored e-cigarettes would be impacted by scenarios of E1 and E2.

As shown in Table 2, compared with younger adults, older adults would be less likely to have intentions to seek illicit channels for menthol cigarettes and flavored cigars but be more likely to have intentions for flavored e-cigarettes. Compared with non-Hispanic Whites, non-Hispanic Asians and all others would be less likely to intend to seek illicit channels for menthol cigarettes and flavored cigars and Hispanics would be more likely to intend to seek illicit channels for both banned products. People with higher educational attainment would be less likely to intend to seek illicit channels for menthol cigarettes and flavored cigars but more likely to intend to seek flavored e-cigarettes. Among both groups – those who smoke flavored tobacco products and those who use flavored e-cigarettes – current daily use was associated with a higher likelihood of intending to seek illicit channels post-ban compared with those who currently use the flavored product less than daily (more than three times for people who smoked and twice for users of flavored e-cigarettes). For both groups, concurrent uses of non-flavored tobacco and nicotine products increased the intention of seeking illicit channels. Among users of e-cigarettes, those who had intentions to quit were more likely to intend to seek illicit channels compared with those who had less or no intentions to quit. Compared with those who reported currently smoking only menthol cigarettes, respondents who smoked flavored cigars or both menthol cigarettes and flavored cigars had greater intentions to try illicit channels. Among those who used flavored e-cigarettes, those who used relatively complicated devices (e.g., tanks, vape pens, mods, rechargeable, compared with disposable products), and those who had smoked before, were more likely to intend to seek illicit channels than their counterparts.

Table 2.

Result of logistic regression models: odds and 95% confidence intervals of having intention to seek illicit channels for banned tobacco and nicotine products among users of menthol cigarettes or flavored cigars and users of flavored e-cigarettes, among adult users of flavored tobacco/nicotine products (USA, 2021–2022).

		Users of menthol cigarettes or	Users of flavored e-cigarettes
Gender	Male (ref)	1	1
	Female	0.93(0.84,1.03	1.11(1,1.24)*
Age	18–34 years (ref)	1	1
	35–64 years	0.98(0.89,1.09	0.99(0.89,1.1)
	65 years and above	0.56(0.45,0.7)	2.73(1.53,4.87
Race/ethnicity	Non-Hispanic White (ref)	1	1
	Non-Hispanic Black	1.06(0.92,1.23	1.81(1.48,2.21
	Non-Hispanic Asian and all others	0.69(0.55,0.89	0.82(0.66,1.03
	Hispanic C.	1.33(1.17,1.52	2.12(1.82,2.46
Educational attainment		1.39(1.29,1.51	1.63(1.48,1.79
Household lncome		0.93(0.89,0.97	0.89(0.85,0.94
Urbanization	Urban (ref)	1	1
	Suburban	1.16(1.02,1.33	0.94(0.81,1.08
	Rural J	1.03(0.84,1.26	0.62(0.49,0.78
If uses daily		3.04(2.56,3.61	2.17(1.585,2.56
If intends to quit		1.06(0.96,1.18	2.07(1.84,2.32
if concurrent uses of non-flavored cigarettes or cigar		1.31(1.18,1.46	Not applicable
if concurrent uses of e-cigarettes		0.94(0.85,1.05
Use of menthol cigarettes and flavored cigars	Menthol cigarettes only (ref)	1
	Flavored cigars only	1.27(1.12,1.45
	Both menthol cigarettes and flavored cigars	1.6(1.41,1.81)**
1 Ban scenarios	C (ref)	1
	E1C	0.8(0.71,0.9)**
	E2C	0.81(0.72,0.91
If had smoked before		Not applicable	2.24(1.96,2.57
Concurrently used tobacco or non-flavored e-cigarette			1.22(1.05,1.41
E-cigarette device	Disposable (ref)		1
	Tanks,or vape pens		1.32(1.11,1.57
	Mods		1.78(1.49,2.12
	Rechargeable		2.24(1.87,2.7)
Used flavored tobacco or nicotine	Any flavors except menthol (ref)		1
	Menthol but not other flavors		1.49(1.29,1.71
	Both menthol and other flavors		1(0.87,1.14)
1 Ban scenario	E1 (ref)		1
	E2		0.89(0.76,1.05
	E1C		0.76(0.64,0.9)
	E2C		0.74(0.63,0.87

Notes:

^*for p<0.05,and

^**for p<0.01. Values in parentheses are 95% confidence intervals.

¹Ban scenarios: E1 ban on e-cigarettes with any flavors except menthol; E2 ban on e-cigarettes with any flavors including menthol; C ban on menthol cigarettes and flavored cigars, with flavored e-cigarettes remaining available; E1C ban on e-cigarettes with any flavors except menthol, menthol cigarettes and flavored cigars; and E2C ban on e-cigarettes with any flavors including menthol, menthol cigarettes and flavored cigars.

In general, respondents’ likelihood of intending to seek illicit channels decreased when the ban scenario included more banned products. Among those who smoked flavored tobacco products, scenarios E1C and E2C were associated with 80% and 81% of the intention to seek illicit channels compared with scenario C, respectively. For users of flavored e-cigarettes, scenarios E2, E1C, and E2C were associated with 89%, 76%, and 74% of the intention to seek illicit channels compared with scenario E1.

Discussion

Overall, among all respondents, about 20–40% of those who used flavored tobacco and nicotine products had the intention to seek illicit channels to obtain flavored products if they were banned. These levels are roughly consistent with one study that estimated about one-quarter of users might intend to seek illicit channels if menthol cigarettes were banned [38]. Our findings suggest that a significant portion of users of flavored products would be open to using illicit products if the flavored options were unavailable through legal channels. Nonetheless, the intention to seek illegal channels may differ from actuality given unanticipated inconvenience, cost, lack of availability, or legal or social consequences [21]. Participants may express a higher intention to try to buy banned products from local stores and online sellers because of perceived convenience and accessibility. However, these products might not be readily available due to stricter enforcement and scrutiny. On the other hand, participants may be less inclined to purchase banned items from street sellers because they may perceive greater associated risks.

Among those who smoked flavored tobacco products, intending to seek illicit channels for banned products decreased when flavored e-cigarettes were added to the ban although introducing a flavored e-cigarettes ban may not directly impact them because the majority of respondents did not use e-cigarettes concurrently. Similarly, those who used flavored e-cigarettes would be less likely to intend to seek illicit channels when menthol cigarettes and flavored cigars were added to a ban on flavored e-cigarettes, although those who used flavored e-cigarettes in this study did not use either menthol cigarettes or flavored cigars at all. Overall these findings indicate that users of flavored tobacco and nicotine products would be less likely to intend to seek illegal channels under a ban scenario with broader coverage (i.e., when fewer types of flavored tobacco-nicotine products remain legally available). This finding is surprising given the expectation that users of banned flavored products would be less likely to seek out illicit flavored products if they could still legally secure flavored versions of other types of flavored tobacco products. But perhaps users of banned flavored products are more likely to see the ban as fair and worthy of respect if it applies broadly (and does not just target their preferred type of flavored product) or are more likely to perceive flavored products as bad to use or otherwise undesirable if they are more broadly banned [39]. Another possible explanation is that respondents perceived a comprehensive flavor ban as more likely to be ubiquitous and strictly enforced, making any effort to try to find illicit products less promising or worth considering.

The intention to seek illicit channels varied significantly according to some demographics. Older users would be less likely to intend to seek illicit channels for menthol cigarettes and flavored cigars than younger users but older users of e-cigarettes would be more likely to do so for flavored e-cigarettes than younger counterparts. This may be because older adults tend to smoke long time, with heightened awareness of the risks associated with smoking, and may use e-cigarettes as aids in smoking cessation. The finding related to the variation by educational attainment among those who used flavored e-cigarettes and among people who smoked is consistent with a previous study that reported individuals with higher levels of education have higher odds of engaging in illicit purchases compared with their counterparts [31]. Our result about the variation of the intention of seeking illicit channels among race/ethnicity is in line with the finding from the 2021 National Survey on Drug Use and Health (NSDUH) that Asian people were much less likely to have used illicit drugs compared with people in most other racial or ethnic groups [40]. At the same time, our result is different from the 2021 NSDUH data which reported that Hispanics were less likely to have used illicit drugs compared with Whites and Blacks. It should be noted that the FDA’s ban on flavored tobacco and nicotine products focuses on restrictions for retailers rather than consumers, and it will be not illicit for consumers to seek banned products from illicit channels. Thus, the results above ‘seek illicit channels’ should not be interpreted as engaging in illicit activities.

Consistent with previous studies [11, 31],“heavy” users of tobacco or nicotine products – i.e., those who either used some products daily and those who used complicated e-cigarette devices would be more likely to intend to seek illicit channels. Among users of flavored e-cigarettes, those who had smoked before were more likely to intend to seek illicit channels, and this may be explained by their nicotine dependence and motivations for using flavored e-cigarettes as an alternative to smoking [41, 42]. These respondents, probably more dependent users, were found to be less likely to quit using any tobacco and nicotine products than their counterparts [33, 43]. Interestingly, our results indicated that among users of flavored e-cigarettes, those who had the intention to quit the use were more likely to seek illicit channels if these products were banned, although the intention to quit tobacco and nicotine use may be associated with a heightened awareness of the health risks and the desire to make a change. Further studies are needed to explain this counterintuitive finding.

Our results indicated that the intention of trying illicit channels was associated with the specific flavors that people used – e.g., those who smoked flavored cigars would be more likely to try illicit channels compared with those who smoked menthol cigarettes, and those who used menthol e-cigarettes would be more likely to intend to seek illicit channels compared with those who used e-cigarette of flavors other than menthol, respectively. One previous study also reported that those who smoked menthol cigarettes were less likely to intend to seek illicit market sellers compared with those who smoked other flavored products [33]. This conclusion may be explained by initial evidence that different flavors are associated with different levels of addiction [44]. Further studies are needed to understand the underlying mechanisms.

This study had several limitations. First, our respondents were drawn from a convenience sample and were not representative of all users of tobacco and nicotine products in the US. Although the addition of weights may alleviate this limitation. Second, it should be noted that people’s intentions for seeking illicit channels may be different than their behaviors in reality. For example, respondents may attempt to transition to non-flavored alternative products or explore other substitutes, or they may opt to cease the use of any tobacco and nicotine products [45, 46]. In this study, we provided several illicit channels for respondents to choose from, and thus respondents may have been misled to believe that all of these channels would be available to them. However, some of the illicit channels may not exist or may be hard to access in an actual ban scenario. For example, for an illicit market to exist, large-scale cross-border smuggling or illegal domestic manufacturing, as well as a supply network would be necessary. Such conditions may not be true in the US [21]. Without a doubt, the availability of some illicit channels such as local retailers who also illicitly sell banned products and online sellers who violate the ban is contingent on how strictly the ban is implemented and enforced.

The study has implications for interventions. The identification of the groups who are at high risk for seeking illicit channels for banned tobacco and nicotine products can inform the design of tailored programs to address the issue. To better prevent the development of illicit tobacco and nicotine product markets and related illicit activities, we will need to not only examine the demand from potential users but also investigate possible supply channels. Considering that half of the users of flavored products would not reject using illicit banned products, it will be crucially important to constrict the supply channels and develop targeted public education messaging to discourage illicit use and encourage cessation among those subgroups most likely to try to secure illicit products.

Highlights.

1. 24–30% of people who smoked would seek illicit channels for banned flavored products

2. 21–41% of dual users and 35–39% of e-cigarette users would seek illicit channels

3. Heavy users are more likely to seek illegal channels

4. Comprehensive bans may decrease users’ likelihood of trying illegal channels