Miscommunication about the US federal Tobacco 21 law: a content analysis of Twitter discussions

Abstract

Objective

Tobacco 21 is a law that sets the minimum legal sales age of tobacco products to 21. On 20 December 2019, the USA passed a federal Tobacco 21 law. The objective of this study is to explore Twitter discussions about the federal Tobacco 21 law in the USA leading up to enacted.

Methods

Twitter messages about Tobacco 21 posted between September and December 2019 were collected via RITHM software. A 2% sample of all collected tweets were double coded by independent coders using a content analysis approach.

Results

Findings included three content categories of tweets (news, youth and young adults and methods of avoiding the law) with eight subcodes. Most news tweets incorrectly described the law as a purchase law (54.7%). However, Tobacco 21 is in fact a sales law—it only includes penalties for tobacco retailers who sell to underage purchasers. About one-fourth (27%) of the tweets involved youth and young adults, with some claiming the law would reduce youth smoking and others doubting its ability to limit youth access to tobacco products. Few tweets (2.5%) mentioned methods of circumventing the policy, such as having an older peer purchase tobacco.

Conclusions

As several countries explore raising their age of sale of tobacco laws to 21, they should couple policy enactment with clear and accurate communication about the law. Compliance agencies at all levels (eg, local, regional, national) can use social media to identify policy loopholes and support vulnerable populations throughout the policy implementation process.

INTRODUCTION

On 20 December 2019, the US president signed a bill entitled ‘Further Consolidated Appropriations Act, 2020’ that included a provision that raised the minimum legal sales age (MLSA) of tobacco products (including e-cigarettes) from 18 to 21 years within the Federal Food, Drug, and Cosmetic Act. With tobacco as the leading cause of premature death and disease worldwide and 85% of regular users using these products before 21 years of age,¹ the federal law has the potential to decrease morbidity and mortality in the USA. Moreover, several European countries (eg, Scotland, Ireland, Finland, France, the Netherlands) have discussed raising their age-of-sale laws to 21 as a part of the smoke-free generation movement, which supports the WHO’s ultimate goal of eliminating the sale and use of all tobacco products.² However, there has been confusion between MLSA laws and purchase, use or possession (PUP) laws that impose penalties on the retailer and the under-age purchaser, respectively. If MLSA laws are not interpreted correctly or implemented as intended, or if loopholes exist that allow industries and users to circumvent these policies, they may not have their intended affect.

The US federal Tobacco 21 law regulates the MLSA and does not impose a penalty for under-age purchasers. PUP provisions can be found in some state and local Tobacco 21 laws in the USA³; however, the federal law only penalises tobacco retailers who sell tobacco products.⁴ Although some have incorrectly labelled MLSA laws as purchase laws, the distinction between the two is important. For example, one study regarding New York City’s Tobacco 21 law incorrectly defined it as a minimum legal purchase age law,⁵ despite the fact that it did not include a PUP provision. Further, while there is international concern for the effectiveness of Tobacco 21 laws to decrease adolescent smoking,⁶ few studies have recognised the differences in policy components.^{3 7} For example, including PUP provisions may undermine enforcement because it redirects attention from tobacco retailers to under-age purchasers.^{8 9}

Criticisms of Tobacco 21 laws have also included the feasibility to enforce these laws. Tobacco industry advocacy groups have propagated messages that these laws are not effective because proxy buyers—those older than 21 years who purchase tobacco products for their under-age peers—create avenues for those under 21 to circumvent the policy.^{1 10} However, for nations focused on decreasing adolescent smoking, research suggests that these policies will be most effective among 15–17 year-olds, who have fewer peers aged 21+ years than they do peers aged 18–20 years.¹ Moreover, some suggest that more evidence regarding the efficacy of these laws is needed to encourage European countries to expand their age of sale to 21 years.⁶

Social media data are a valuable source of organic information that can help to understand the use of emerging tobacco products^11–14 and public response to tobacco control policies.^15–17 Such data can help identify potential barriers to implementation or policy loopholes that exist. For example, even after flavoured cigarettes were prohibited in the USA, tobacco companies circumvented this policy by creating a flavoured cigar option that was not regulated, and online promotion of these illegal flavoured cigars was still identified after they were outlawed in 2009.¹⁸ Another study found those under 18 years of age were able to purchase cigarettes online in 2014, indicating that online tobacco sales were largely unregulated and provided a loophole for youth access of tobacco products.¹⁹ Similarly, actions described by social media users before and after the enactment of the federal Tobacco 21 law can help researchers understand the impact it may have on youth and young adults’ access to and use of tobacco. However, to date, no study has examined content about Tobacco 21 from social media sources. This is a gap in literature because information provided about this law on social media could influence public awareness, understanding and perceptions about Tobacco 21. Twitter is an ideal place to search for misinformation about Tobacco 21 for several reasons: past research has found Twitter users engage in discussions about tobacco control policies^{15 17}; it is an open platform that allows public information to be shared regardless of connections (eg, friends); and it is used by 32% of US youth.²⁰ Further, research suggests that Twitter has been instrumental in setting agendas for traditional news coverage,²¹ particularly about e-cigarettes,²² due to its ability to share information instantaneously and allow users to engage with discussions about news. Therefore, the objective of this study is to explore Twitter discussions about the federal Tobacco 21 law, leading up to its enactment.

METHODS

Data collection

Data were collected using publicly available ‘RITHM’ software²³ that allowed for retrieval of real-time Twitter messages (ie, tweets) posted between 1 September 2019 and 31 December 2019, the final 3 months before the federal Tobacco 21 law was enacted. Using a broad primary search along with a more focused filtered search, key search terms were used to identify content related to the federal Tobacco 21 law. The primary search terms included terms within tweets related to tobacco and nicotine products (ie, vape(s), vaper(s), vaping, vaped, ecig(s), e-cig(s), cigarette(s), cig(s), nicotine, tobacco). Next, filters were used to narrow the search to tweets specific to Tobacco 21 (ie, 18, 21, age, buy, mcconnell, mitch, purchase, t21, tobacco21) based on obvious important key terms and the names of legislators who championed the first Tobacco 21 bill. Keywords also included hashtags (ie, ‘#’ prefix on text) and capitalisation (eg, ‘t21’ also captured ‘#T21’) to ensure a broad capture of relevant tweets. The initial search identified 615 574 tweets. After removing redundant tweets or ‘retweets’ (ie, sharing others’ tweets), 231 447 tweets remained. Finally, a random subsample of 2% of tweets (n=4628 tweets) was identified for human annotation. Sample sizes and random selection of tweets employed methods used in prior published work that ensured adequate representation of tweets.^{14 23 24}

Codebook development

The codebook was developed via a systematic process that allowed the research team to identify important topics discussed on Twitter about the federal Tobacco 21 law. Three researchers first developed an initial codebook based on prior work that identified relevance, sentiment and news.¹² Next, two coders assessed 500 tweets per week for 9 weeks. After all data were coded, original tweets irrelevant to Tobacco 21 (n=3328) were removed, and the research team returned to the codebook to add, delete, expand, separate and/or collapse codes that represented discussions captured in the tweets.^{25 26} Using this empirically grounded approach, in conjunction with prior research,³ content was analysed at a tweet level to develop content categories with subcodes within these categories. Thus, the codebook was expanded, and relevant tweets were re-examined within coding categories to identify exemplar content and to synthesise contextual nuance from the data.

Initial codes indicated if the tweet referred to a Tobacco 21 law, and if so, coders indicated if the tweet was RELEVANT to the federal law or a state, local or business policy. Those that did not discuss the federal Tobacco 21 law were removed from the analysis. The initial codebook included codes for NEWS (news articles about Tobacco 21) and COMMERCIAL (posts that appeared to be promotional posts about tobacco products (eg, WARNING: This product contains nicotine, which is an addictive chemical. Must be 21+ to RT or purchase)). In previous social media-related studies, Twitter messages made by news sources were removed¹²; however, for the current study, after all tweets were reviewed, the authors noticed language differences between how the news-related tweets described the law (ie, sales law or purchase laws). Based on the current literature,³ three subcodes were developed to indicate if the news-related post mentioned the purchase age, sales age or neither, if the tweet did not include either the purchase or sales age of tobacco products or simply referred to the law as Tobacco 21. Tweets coded as COMMERCIAL (eg, product promotion) were removed from analysis because they only mentioned Tobacco 21 as a disclaimer and did not provide a meaningful discussion.

Based on prior research,^{3 12 27} the research team developed two additional codes (each with three subcodes) following the review of all tweets: ‘youth and young adults’ and ‘methods of avoiding the law’. These categories included: YOUTH AND YOUNG ADULTS and AVOID THE LAW. For operational definitions and examples of each code and subcode, see table 1.

Table 1.

Operational definitions and example tweets for each code (n=1109)

Code	Definition	Example tweets	n	%
Subcode
News	A news article about Tobacco 21.	The Tobacco-Free Youth Act would raise the age to buy tobacco products, including e-cigarettes, from 18 to 21. (link to news article)	190	17.1
Purchase age	News post that uses the term purchase or buy when describing the law.	Congress Raises the Legal Age to Buy Tobacco and Vaping Products to 21. (link to news article)	104	9.4
Sales age	News post that uses the term sales (refers to retail sales) when describing the law.	Congress Approves Raising Age to 21 for E-Cigarette and Tobacco Sales. (link to news article)	36	3.3
Neither	News post that does not include either purchase or sales terms or simply referred to the law as Tobacco 21.	Congress negotiating deal to raise tobacco age. (link to new article)	50	4.5
Youth and young adults	Mentions youth or young adults.	I think the smoking age needs to stay 19, teen smoking is not the government’s problem, it’s the parents if the teens smoking and vaping.	309	27.9
High school/youth	Mentions kids, high school, minors and/or youth.	The best answer to youth access. Would be to take vapes out of corner stores and gas stations. And slap MASSIVE fines on those selling to minors!! Sure T21 helps 2. But geez the min wage PT clerk cares little who they sell to!! #NotaBot #WeVapeWeVote #VapingSavesLives.	100	9.0
18–20 year-olds	Mentions those between 18 and 20 years of age.	It would suck to be 18–20 and have a cig addiction. Yesterday they could just buy them. Now they gotta search for the cig plug. Wait 8 hours for him to reup and then he out by the time they get to him. smh. I got the cigs though 1 for 10. 2 for 20. 3 and a half for 30.	216	19.5
Issues	Mentions issues or frustration with the law.	HAHAH u freshman in high school are about to be tweaking out.	22	2.0
Avoid the law	Describes a way for those under 21 to gain access to tobacco.	bro how tf am I gonna buy woods now y’all doing too much.	74	6.7
Will sell to those under age	Explains how a clerk/tobacco retailer will sale tobacco products to those under 21.	Bro you have to be 21 now to buy tobacco. Shits dumb. I remember Trynna find an 18 year old to buy me wraps, dip, cigs. Shits fucking rough. Ya’ll can just come to my gas station while I’m working, just don’t get me fired.	11	1.0
Looking for a buyer	Tweet is looking for someone else to buy tobacco products for them.	hey ems, can u buy me some tobacco products bc I’m not 21 yet.	28	2.5
Plan to buy while under age	Explains how the person will buy tobacco (with or without a fake ID) when under 21.	I bought my gas station attendant baby clothes when he told me his gf was expecting yall ain’t about to stop shit.	10	0.9

Coding and analysis

Trained coders used a qualitative content analysis approach to independently annotate the initial 4628 tweets. After refining the codebook, two independently working coders annotated the 1300 tweets relevant to the Tobacco 21 law over 4 weeks. Coders were provided with the original tweet text, text of the original tweet (when quoting another tweet) and the link to the online version of each tweet. Coders were allowed to follow links to the tweet and to news articles (when applicable) to review the tweet and text further, especially when the context such as embedded images, emojis and videos was provided. The two coders met with the principal investigator (PDD) weekly to examine the coding reliability and assist with interpretations of operational definitions. Each week and following annotation of all data, inter-rater reliability was measured via Cohen’s κ.²⁸ Acceptable agreement (κ>0.60)²⁹ was found after the first week of coding; κ values ranged from 0.673 to 1.00 for all final codes. Coding discrepancies were adjudicated under the supervision of the principal investigator and consensus was achieved for all codes. Frequencies of final codes were reported. Tweets that referenced a state, local or business policy were removed from the analysis to ensure discussion was about the federal Tobacco 21 policy only.

RESULTS

News

Of the 1300 tweets relevant to Tobacco 21, a total of 1109 (85.4%) were about the federal law, and of those, 190 (17.1%) included a news report. Among these news tweets, 104 (54.7% of all news tweets) incorrectly stated that it raised the buying age or the age at which someone could purchase tobacco. Thirty-six (19.0%) tweets correctly stated that the MLSA law restricts the sale of tobacco, and 50 (26.3%) generally stated that the law restricted the tobacco age or broadly referenced Tobacco 21. See figure 1 for frequencies and example tweets for each subcode.

Figure 1.

Descriptions of news posts about the federal Tobacco 21 law on Twitter (n=190). FDA, Food and Drug Administration.

Youth and young adults

When the federal Tobacco 21 law was mentioned, 309 (27.9%) of the 1109 tweets discussed use among youth (those under 18 years of age) or young adults (those between 18 and 24 years of age). Overall, 100 (9.0%) tweets referred to high school students, youth or kids. These tweets included support for the law, such as one tweet that described the current trend of JUUL use among high school students: ‘Back in the day kids would start smoking cigarettes around 14 or 15 if not younger. 18 year olds today are buying juuls for younger kids in there high school. If u wanna stop underage vaping change the age limit to 21’. Other tweets criticised the law or were sceptical that it would work, such as: ‘How? The age went from 18 to 21. Kids weren’t allowed to purchase cigarettes before this law passed. How is this a step in any direction?’. Additionally, 216 (19.5%) of the tweets referred to individuals 18–20 years of age. For example, one aged 20 years posted, ‘I turned 18 to be able to buy cigs & wraps for 2 yrs all for the law to change & now I feel 17 all over again since you have to be 21 to buy cigs now’. Overall, 22 (2.0%) tweets described challenges that young adults would face due to the change in the law. For example, one tweet read, ‘just had a breakdown in the 7–11 parking lot after finding out the national tobacco age is now 21’.

Methods of avoiding the law

Seventy-four (6.7%) tweets discussed methods of evading enforcement of the Tobacco 21 law. There were 11 (1.0%) tweets offered to buy tobacco products for or sell tobacco products to those below 21. For example, one tweet read, ‘If you’re 18–20 and need tobacco products, hit me up. I live right across the street from super smoke shop’. Additionally, 28 (2.5%) tweets were posted by those under the age of 21 expressed needing someone to buy tobacco products for them. One user posted, ‘I never thought I’d be 20 years old and sending a friend to go buy cigarettes for me, but here we are’. Finally, 10 (1.0%) tweets explained that they would purchase tobacco products themselves from retailers or online despite being less than 21 years of age. Some suggested that they or others would use a fake identification, while others believed they could purchase tobacco themselves without a fake identification. For example, one person stated that they had evaded the law when the MLSA was 18 years of age, ‘Just buy them in bulk on amazon. That’s what I was doing when I started smoking at 16’.

DISCUSSION

We found that most news-related Twitter posts about the US federal Tobacco 21 policy incorrectly described it as a purchase policy (54.7%), when in fact it restricts the sale of tobacco. This information is important because the federal law does not include a penalty for under-age purchasers; however, there is a graduated penalty for retailers if they sell to an under-age purchaser during a compliance check.⁴ We also found that some Twitter users discussed methods of avoiding the current Tobacco 21 law by either buying tobacco for someone who was under 21 years, asking someone else to buy tobacco products for them or attempting to purchase tobacco products with or without a fake identification. These findings provide valuable information to the tobacco health policy community because incorrect framing of this policy may change the community or media sentiment towards the law, which could undermine public support needed to pass similar laws in other countries. For example, European countries and other nations that may be considering raising their legal sales age of tobacco to 21 years may find it useful to consider these miscommunications and potential loopholes as they design, implement and communicate policy changes.⁶

PUP laws have had mixed results around efficacy of the laws to reduce youth purchase and use of tobacco products,^30–32 resulting in controversy.³³ PUP laws have been debated for more than 20 years, with recognition that they redirect tobacco control from the retailers and industry who profit from the sale of tobacco to the under-age purchaser.⁹ More recently, tobacco control advocates have pointed out that PUP laws may create health disparities by disproportionally putting vulnerable populations at risk of racial discrimination and police brutality by allowing governing authorities opportunities to profile African American youth and young adults.³⁴ For example, Eric Garner was accused of illegally selling cigarettes before being killed by a white police officer using a restraining chokehold.³⁵

Some Twitter users discussed issues young audiences would face, such as frustration that they could no longer buy tobacco products or experiencing withdrawal symptoms from novel tobacco products. Some of these tweets described how 18–20 year-olds who had become addicted to nicotine may have difficultly quitting overnight, as the federal law when into effect immediately. Other research has highlighted the addictive nature of emerging tobacco products,^{27 36 37} and some have begun creating cessation programmes and health communication campaigns to help young audiences quit using all tobacco products, including e-cigarettes.^38–40 This discussion further supports the need to eliminate PUP laws that penalise youth and young adults struggling with addiction that can be attributed to tobacco industry marketing campaigns⁴¹ or to replace monetary penalties with educational or cessation-based programmes. Moreover, some states’ definition of tobacco within their MLSA exempts products that have been authorised for sale by the Federal Food, Drug, and Cosmetic Act (eg, nicotine replacement therapy).³ More research is needed to determine the safety and efficacy of nicotine replacement therapy for youth and young adults who are addicted to novel tobacco products. If found to be an effective method, such considerations could guide future tobacco control policies within state tobacco product definitions.

Our findings include some comments by Twitter users about methods of avoiding or circumventing the federal Tobacco 21 law. Some users described having an older sibling or friend purchase the product for them, and others offered to purchase tobacco products for those under 21 for an additional fee. Such claims have been propagated by the tobacco industry as justification for not passing youth access laws^{10 42}; however, decreasing the availability and accessibility of tobacco via Tobacco 21 laws has been found to effectively reduce the use^{43 44} and sales of tobacco products to those under 21 years of age,^45–47 even among popular cigarette brands.⁴⁸ Moreover, only 2.5% of our sample of tweets about Tobacco 21 included requests for other Twitter users to buy tobacco for someone under 21 and only 1.0% of the tweets claimed that the person would buy tobacco for their under-age peers. While it is probable that others who purchase for their under-age peers did not post such claims on social media, this may occur less often than the tobacco industry would like for policymakers to believe.

Some researchers recommend that the sale of tobacco should be limited to government-controlled outlets to offer policy proportionality, regulation of a substance based on the level of harm it presents, as seen in other countries such as Hungary.⁴⁹ Such policies would remove the need for sales laws and the need for government agencies to develop and implement inspections or ‘sting operations’ with an under-age decoy to enforce the MLSA policy for the retailers. However, this could result in a need for PUP laws, because MLSA laws that include penalties for tobacco retailers would no longer exist. Although such change would decrease confusion between sales and PUP laws, it would also warrant caution when implemented. If such changes were made to decrease tobacco exposure, PUP laws should include reasonable educational penalties instead of monetary fines or penalties that could marginalise already vulnerable populations (eg, African American communities).^{10 34} Thus, in the attempt to end all tobacco use, complementary control strategies that impact each other should be carefully considered.

Limitations

Although our findings include organic discussions about Tobacco 21 via a publicly available forum, these data should not be misinterpreted as coming from a representative sample of the USA. Also, some search terms may have limited utility over time. For example, because Mitch McConnell championed the first Tobacco 21 bill presented earlier in the year, we used ‘McConnell’ as a search term. For those seeking to replicate our procedures, it will be important to hone the list of search terms in that future context.

CONCLUSIONS

Twitter posts about Tobacco 21 may have misled Twitter users to believe that the policy would penalise those under the age of 21 from attempting to purchase tobacco products. If they do not know that the federal Tobacco 21 law is a sales law, Twitter users may not be aware that tobacco retailers are the responsible party for implementation of tobacco control policies in their communities. Further, Twitter discussions about youth and young adults discussed frustrations young tobacco users may experience, including their struggle with addiction. While more research is needed to determine how effective the federal law is to reduce youth and young adult initiation and use of tobacco products, those enforcing the law should consider the loopholes that exist and how to combat such issues without marginalising vulnerable populations. Further, public health campaigns can help alleviate Twitter users’ confusion about and rapid proliferation of misinformation about Tobacco 21 and other tobacco control policies. Ending online sales and transferring tobacco sales to government-controlled outlets could help achieve downstream goals of the tobacco control; however, it would require review of PUP laws that penalise victims of tobacco marketing.

What this paper adds.

• Most news-related tweets incorrectly described the federal minimum legal sales age law as a purchase law—a law that penalises the youth purchaser.

• Twitter provided valuable data regarding the public debate about the effectiveness of Tobacco 21 to reduce youth access to and use of tobacco products.

• Twitter can help identify methods that those under 21 years use to circumvent the law.

Data availability statement

Data are available from the corresponding author, PDD, upon reasonable request.