Article Text

Trends in annual sales and current use of cigarettes, cigars, roll-your-own tobacco, pipes, and smokeless tobacco among US adults, 2002–2012
  1. Israel T Agaku,
  2. Hillel R Alpert
  1. Department of Social and Behavioral Sciences, Center for Global Tobacco Control, Harvard School of Public Health, Boston, Massachusetts, USA
  1. Correspondence to Dr Israel T Agaku, Department of Social and Behavioral Sciences, Center for Global Tobacco Control, Harvard School of Public Health, Boston, MA 02115, USA; iagaku{at}post.harvard.edu

Abstract

Introduction Regulatory imbalances exist in the treatment of cigarettes and non-cigarette tobacco products in the USA. We assessed whether declines in cigarette consumption during 2002–2012 were offset by increased use of non-cigarette tobacco products—cigars, pipes, roll-your-own (RYO) and smokeless tobacco.

Methods Industry-reported taxable removals (actual sales) were converted into packs for cigarettes and cigarette pack equivalents (CPEs) for loose tobacco (RYO plus pipe tobacco) and moist snuff. Cigars were not converted to CPEs because of their heterogeneity in size/tobacco content. Per capita sales were calculated for the US adult population aged ≥18 years based on the US Census Bureau data. Self-reported data on current (past 30-day) tobacco use among US adults aged ≥18 years were from the National Survey on Drug Use and Health (NSDUH). Joinpoint and logistic regression were used to assess linear trends during 2002–2012.

Results During 2002–2012, cigarette sales declined from 96.91 to 59.85 cigarette packs per capita; increases occurred for sale of cigars (30.51–57.42 cigars per capita), loose tobacco (2.50–5.63 CPEs per capita) and moist snuff (10.64–14.58 CPEs per capita; all p<0.05 for trend). Self-reported current cigarette smoking declined during 2002–2012 (27.4–23.6%); increases were noted for current RYO (2.6–3.6%) and smokeless tobacco use (3.5–3.7%; all p<0.05 for trend).

Conclusions The increase in non-cigarette tobacco consumption is a public health concern because all tobacco products are harmful. Eliminating imbalances in tax structure and regulations between cigarettes and non-cigarette tobacco products may help reduce aggregate tobacco consumption.

  • Disparities
  • Non-cigarette tobacco products
  • Public policy
  • Socioeconomic status
  • Surveillance and monitoring

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Tobacco use remains the leading cause of preventable death and disease in the USA.1 The 50th Anniversary Surgeon General's Report on the Health Consequences of Smoking indicated that since 1964, approximately 20 million premature deaths have occurred that are attributable to cigarette smoking.2 Although cigarette smoking has declined significantly within the past five decades, the overall prevalence is too high and large disparities in tobacco use remain, with a higher prevalence noted among racial minorities, lesbian, gay, bisexual or transgender individuals, persons with disabilities, and those of low socioeconomic status.1 ,2 In addition, novel and modified tobacco and nicotine delivery products continue to multiply in the USA, and some of these have been rapidly increasing in popularity and use.3–5

The Healthy People 2020 national objectives of the US Department of Health and Human Services include a reduction of both cigarette and non-cigarette tobacco product use, namely the proportions of US adults who smoke cigarettes to ≤12%, cigars to ≤0.2% and smokeless tobacco to ≤0.3%.6 However, regulatory imbalances exist in the treatment of cigarettes and some non-cigarette tobacco products in the USA.7 Specifically, under two tobacco control laws that became effective in 2009—the Children's Health Insurance Program Reauthorization Act (CHIPRA) and the Family Smoking Prevention and Tobacco Control Act (FSPTCA)8 ,9—cigarettes are subject to stricter tax and other regulations than some other tobacco products.

Much larger increases in federal excise tax rates were implemented by CHIPRA on cigarettes (from $0.39 to $1.0066 per pack), small cigars (from $0.037 to $1.0066 per pack) and roll-your-own (RYO) tobacco (from $0.045 to $1.0066 per cigarette pack equivalent (CPE)), in comparison to moist snuff (from $0.044 to $0.113 per can), large cigars (from $0.049 to a maximum of $0.4026 per large cigar) and pipe tobacco (from $0.0069 to $0.177/oz).8 The tobacco industry responded to this multitiered federal tobacco tax policy by relabeling or changing the formulation of certain products.10 For example, RYO tobacco was packaged and labelled as pipe tobacco to avoid stricter tax policies and regulation, while small cigars were reformulated and reclassified as large cigars.10 ,11 These evasive marketing practices by tobacco manufacturers were possible because for practical purposes the products are taxed and regulated according to the label on the packaging.11 ,12 Such tax avoidance strategies by tobacco manufacturers have the potential to promote uptake and continued tobacco use among price-sensitive populations by providing cheaper alternative tobacco products in place of cigarettes.11

Similarly, the regulation of tobacco products by the Food and Drug Administration (FDA) under FSPTCA treats cigarettes more strictly than several other products, including the banning of the use of characterising flavours and misleading descriptors such as ‘light’ or ‘low tar’.9 ,13 The FSPTCA granted FDA immediate regulatory authority over certain tobacco products, including cigarettes, RYO and smokeless tobacco, but did not specify authority over pipe tobacco, hookahs, electronic cigarettes and small and large cigars, which are thus not subject to the same restrictions on characterising flavours, sales or distribution.9 ,13 In April 2014, FDA indicated its intent to issue a proposed rule that would deem products meeting the statutory definition of “tobacco product” to be subject to FDA's regulation.14 FDA is reviewing the comments received on this proposed rule.

The objectives of this study were to determine whether declines in cigarette smoking during 2002–2012 may have been offset by an increased use of other non-cigarette tobacco products in the USA.15 We assessed trends in tobacco consumption during 2002–2012 using two standard approaches: (1) estimating US per capita sales of tobacco products based on industry-reported tobacco excise tax data and (2) estimating prevalence of current (past 30-day) tobacco product use based on self-reported tobacco use behaviour from a national survey of US adults. These two approaches are complementary measures of tobacco consumption with the caveat that not all tobacco products sold may be actually used by end consumers, and also that multiple tobacco products could be purchased and used by a single consumer. Nonetheless, a nuanced assessment of industry-reported sales data for various tobacco products provides insight into tobacco manufacturers’ strategies in developing and marketing their products to avoid or circumvent regulations. Likewise, analyses of self-reported tobacco use behaviour from survey data can help identify overall trends as well as disparities in tobacco use among various population subgroups. For our analyses, we examined five tobacco product types: cigarettes, cigars, RYO, pipe and smokeless tobacco products.

Methods

Data sources

Per capita tobacco sales

Taxable removals (actual sales) of cigarettes, small cigars, large cigars, RYO and pipe tobacco were obtained from the Alcohol and Tobacco Tax and Trade Bureau, and for moist snuff from the US Department of Agriculture during 2002–2012, based on monthly and quarterly filings by the tobacco companies, respectively.16 ,17 Sales were adjusted for imports and exports for each product type, except for moist snuff for which these data were unavailable. In view of evidence that tobacco manufacturers reformulated and marketed small cigars as large cigars and RYO tobacco as pipe tobacco after the 2009 Federal tax change,10 ,11 we analysed small and large cigars together (as cigars); and RYO and pipe tobacco together (as loose tobacco) in order to compute more accurate and realistic estimates of aggregate tobacco sales for each product type.

Self-reported current tobacco use

Self-reported tobacco use data were obtained from The National Survey on Drug Use and Health (NSDUH), an annual nationwide survey of the civilian, non-institutionalised US population aged ≥12 years.18 Owing to improvements made to NSDUH in 2002 onwards, which limited comparability with data collected in 2001 or earlier, we restricted our analyses to 2002–2012 data. Questionnaires were administered completely in-person using computer-assisted interviews; overall response rates ranged from 73.0% in 2012 to 79.0% in 2002. For this study, we restricted all analyses to persons aged ≥18 years. The final analytical sample of persons aged ≥18 years ranged from n=36, 370 in 2002 to n=39 259 in 2010.

The past 30-day (current) use of cigarettes, cigars, RYO, smokeless tobacco (chewing tobacco/snuff) and pipes was self-reported. Sociodemographic characteristics were sex (male or female), age (18–25, 26–34, 35–49 or ≥50 years), race/ethnicity (non-Hispanic Caucasian, non-Hispanic African-American, Hispanic or non-Hispanic other), annual household income (≤$19 999, $20 000−$49 999, $50 000−$74 999, or ≥$75 000) and education (<high school, high school diploma or >high school).

Analyses

Per capita tobacco sales

The numbers of cigarettes sold were divided by 20 to obtain cigarette pack sales. CPEs were estimated for loose tobacco (RYO plus pipe tobacco) based on weight, 14.6 g loose tobacco per cigarette pack, supported by reports of nicotine ratings filed by manufacturers with the Massachusetts Department of Public Health under Massachusetts regulation 105 CMR 660.000. Moist snuff CPEs assumed the equivalence of a 1.2 oz tin to 2.5 packs of cigarettes based on daily consumption.19 ,20 Cigar sales were not converted to CPEs because of variations in size and tobacco content by cigar type. Adult per capita sales was estimated by dividing total sales of cigarette packs, CPEs of loose tobacco (RYO plus pipe tobacco) and moist snuff, or total cigar unit sales by the number of persons aged ≥18 years in the USA each year using data from the US Census Bureau.21

To test the statistical significance of discrete changes in trends in per capita sales during the study period, joinpoint regression analyses were performed using NCI's Joinpoint V.4.0.1 software.22 Annual percentage changes (APCs) with 95% CIs were computed for each trend line segment and average APCs (AAPCs) were computed to summarise the temporal trends over the entire time period for each product type. APC is a measure of linear change which assumes that the prevalence rates change at a constant percentage of the rate of the previous year. However, since it is reasonable to expect some non-linear trend, we assessed when and how often the APC changed over the entire study period for each tobacco product type (ie, the number of joinpoints; zero joinpoints indicate a straight line).

Self-reported current tobacco use

All analyses were restricted to adults aged ≥18 years, and data were weighted to yield nationally representative estimates. Prevalence of current use of each product type was calculated overall and further stratified by sex, age, race/ethnicity, annual household income and education. Estimates with relative SEs ≥40% were deemed statistically unreliable. Adjusted linear trends in prevalence during 2002–2012 were calculated in a binary logistic regression model adjusting for age, sex and race/ethnicity (p<0.05). For all tobacco products assessed, prevalence estimates are presented within the manuscript for only the years 2002 and 2012 for brevity; however, detailed tables with data for all years during 2002–2012 for each tobacco product can be found in online supplementary tables S1–S5.

Results

Trends in tobacco sales during 2002–2012

Cigarettes

The total number of cigarettes sold declined significantly from 395.24 billion cigarettes in 2002 to 279.84 in 2012 (p<0.05; table 1). Per capita sales declined from 96.91 cigarette packs per capita in 2002 to 79.59 in 2007 (APC=−3.6, 95% CI −4.8 to −2.4), further declined more steeply to 64.05 cigarette packs per capita in 2010 (APC=−7.5; 95% CI −12.5 to −2.2) and then to 59.85 cigarette packs per capita in 2012 (APC=−3.2; 95% CI −8.4 to 2.3; tables 1 and 2). AAPC in per capita sales of cigarette packs during 2002–2012 was −4.7 (95% CI −5.9 to −3.4).

Table 1

Annual taxable removals and per capita sales* of cigarette and non-cigarette tobacco products, USA, 2002–2012

Table 2

APC and AAPC in per capita sales* of cigarette and non-cigarette tobacco products, USA, 2002–2012

Cigars

The total number of cigars sold increased significantly from 5.95 billion cigars in 2002 to 10.14 in 2012 (p<0.05). Per capita sales of cigars increased significantly from 30.51 cigars per capita in 2002 to 58.76 in 2010 (APC=8.5; 95% CI 7.3 to 9.6) before a single joinpoint was observed and the trend levelled through 2012. AAPC in per capita sales of cigars during 2002–2012 was 6.4 (95% CI 4.6 to 8.3).

Analyses of total cigar market share by volume (large and small cigars) during 2002–2012 indicated that the market share for large cigars declined during 2002–2008 (62.2–46.1%), and then increased steeply during 2009–2012 to reach 93.1% during 2012. Figure 1 shows the taxable removals (actual sales by number) of small and large cigars in the USA during 2002–2012.

Figure 1

Taxable removals (actual sales by number) of small and large cigars in the USA, 2002–2012. Data obtained from the Alcohol and Tobacco Tax and Trade Bureau based on monthly filings by the tobacco companies. Sales were adjusted for imports and exports.

Loose tobacco (RYO plus pipe tobacco)

The total amount of loose tobacco sold (by weight) increased significantly from 15.19 million pounds in 2002 to 40.68 in 2012 (p<0.05). Per capita sales of loose tobacco CPEs increased from 2.50 CPEs of loose tobacco per capita in 2002 to 3.35 in 2009 (APC=5.7; 95% CI 3.4 to 8.1) and then increased very steeply to 5.63 in 2012 (APC=17.1; 95% CI 7.8 to 27.3). AAPC in per capita sales of loose tobacco CPEs during 2002–2012 was 9.0 (95% CI 6.5 to 11.6).

Analyses of total loose tobacco market share by volume (RYO and pipe tobacco) indicated that the market share for pipe tobacco declined during 2002–2008 (30.4–13.6%), and then increased steeply during 2009–2012 to reach 89.6% in 2012. Figure 2 shows the taxable removals (actual sales by weight) of RYO and pipe tobacco in the USA during 2002–2012.

Figure 2

Taxable removals (actual sales by weight) of roll-your-own and pipe tobacco in the USA, 2002–2012. Data obtained from the Alcohol and Tobacco Tax and Trade Bureau based on monthly filings by the tobacco companies. Sales were adjusted for imports and exports.

Moist snuff

The total amount of moist snuff sold (by weight) increased significantly from 68.55 million pounds in 2002 to 104.41 in 2012 (p<0.05). Per capita sales of moist snuff CPEs increased from 10.64 CPEs of moist snuff per capita in 2002 to 14.58 in 2012 (APC=3.2; 95% CI 2.9 to 3.5). AAPC in per capita sales of moist snuff CPEs during 2002–2012 was the same as APC since no joinpoint was observed.

Trends in self-reported tobacco use during 2002–2012

Cigarettes

Current cigarette smoking prevalence among all US adults declined significantly between 2002 and 2012 (27.4–23.6%; p<0.001 for linear trend; table 3). Significant declines in current cigarette smoking prevalence during 2002–2012 were observed among both sexes, among the 18–25-year-olds and 35–49-year-olds, all race/ethnicity groups, persons with annual household income $50 000–$74 999 and ≥$75 000 and those with >high school education (table 3).

Table 3

Prevalence trends* in self-reported current (past 30-day) use of cigarettes and non-cigarette tobacco products among US adults aged ≥18 years, National Survey on Drug Use and Health, 2002–2012

Cigars

The overall prevalence of current cigar smoking among US adults did not change significantly between 2002 (5.4%) and 2012 (5.6%; table 3). Significant increases were, however, observed among females (1.6–2.1%), ≥50-year-olds (2.4–3.5%) and persons with annual household income ≤$19 999 (5.7–6.9%; all p<0.001 for linear trend). Significant declines were observed among 18–25-year-olds (11.0–10.9%), 35–49-year-olds (5.8–4.9%) and persons with annual household income ≥$75 000 (6.3–5.4%; all p<0.001 for linear trend).

RYO tobacco

During 2002–2012, the prevalence of RYO smoking among all US adults increased significantly (2.6–3.6%; p<0.001 for linear trend; table 3). Significant increases were observed among males and females, all age, annual household income and education groups, as well as among non-Hispanic Caucasian.

Pipe tobacco

The overall prevalence of current pipe tobacco smoking among US adults did not change significantly between 2002 (0.8%) and 2012 (1.0%). Significant increases were, however, observed among 18–25-year-olds (1.1–1.8%), 26–34-year-olds (0.5–1.1%) and persons with annual household income ≤$19 999 (0.9–1.6%; all p<0.05 for linear trend).

Smokeless tobacco

A significant increase was observed in the overall prevalence of smokeless tobacco during 2002–2012 (3.5–3.7%; p<0.001 for linear trend; table 3). A significant increase was observed among males (6.7–7.1%; p<0.001 for linear trend), but not females. Significant increases were observed among 18–25-year-olds and 35–49-year-olds, non-Hispanic Caucasian, Hispanics, persons with annual household income $50 000−$74 999 or ≥$75 000, and persons with high school or >high school education.

Discussion

Analyses of industry-reported taxable removals showed a significant decline in sales of cigarettes during 2002–2012, whereas significant increases were noted for loose tobacco (RYO plus pipe tobacco), cigars and moist snuff. Self-reported tobacco use data during 2002–2012 also showed declines in current cigarette smoking prevalence, but increases in current use of RYO and smokeless tobacco. A significant joinpoint was observed during 2009–2010 for per capita sales of cigarettes, loose tobacco and cigars, following which an accelerated decrease for cigarettes, an accelerated increase for loose tobacco and a plateauing for cigars were observed with respect to per capita sales. Tobacco manufacturers’ responses to the imbalances in tobacco product tax structure and regulations might account for the dramatic shifts observed in 2009/2010 with regard to the sales of cigarette and other non-cigarette products.15

In response to tax disparities between RYO tobacco and pipe tobacco, and between small cigars and large cigars, tobacco manufacturers began to change the product specification.10 Our results showed that during 2002–2012, the net total sales of loose tobacco (RYO tobacco plus pipe tobacco) increased significantly; however, the market share for RYO tobacco declined dramatically from 2009 onwards while pipe tobacco sales increased dramatically. Interestingly, self-reported data during 2002–2012 showed a significant increase in current RYO smoking but no change in current pipe smoking. Taken together, this may suggest that smokers are increasingly buying loose tobacco labelled as pipe tobacco for use as RYO as a tax avoidance strategy, considering that pipe tobacco is taxed at less than 20% the rate of RYO tobacco.11 ,23 ,24 Indeed, research shows that after the 2009 federal tax change, tobacco manufacturers have marketed pipe tobacco as ‘dual use tobacco’ or ‘multiuse-tobacco’ (ie, for use as pipe tobacco or RYO).11 Our results also showed that while a net increase was seen in the total number of cigars sold (large and small) during 2002–2012, the market share for small cigars (by volume) declined dramatically from 2009 onwards while that of large cigars increased, most likely because of the differential taxation for large and small cigars under the 2009 Federal tax change. We could not, however, analyse self-reported current use of small and large cigars separately because of lack of data by cigar type in NSDUH. The overall trends in self-reported cigar use during 2002–2012, however, remained stable.

We found an increase in moist snuff consumption during 2002–2012. Smoke-free laws, which have proliferated across the USA in the past decade, apply only to combusted tobacco products and sparsely to electronic cigarettes (e-cigarettes), but not to smokeless tobacco products.25 Indeed, the tobacco industry often advertises non-combustible tobacco products as desirable alternative sources of nicotine in places where smoking is prohibited, which could undermine the effectiveness of smoke-free laws to help reduce smokers’ addiction while protecting non-smokers.26 ,27

These findings underscore the need for a comprehensive and integrated framework for tobacco control to reduce aggregate tobacco use.28 The FDA's proposed deeming regulations under consideration are intended to broaden FDA's regulatory oversight to other tobacco products including e-cigarettes, cigars, pipes and water pipes/hookahs for the protection of public health.14 However, as proposed, the deeming regulation still leaves several gaps related to other non-cigarette tobacco products. For example, it does not deal with tax rates or flavours, nor does it contain the same marketing and sales restrictions that are present for cigarettes. Intensified efforts are thus needed to address the issue of regulatory imbalances between cigarettes and other tobacco products so as to close the existing gaps in nicotine and tobacco product regulation which allow tobacco manufacturers to increase the appeal and marketing of non-cigarette tobacco products.23 Tobacco control policies implemented at state, local and tribal government levels can have a more immediate impact in levelling legislative imbalances between cigarettes and other tobacco products with regard to taxation, particularly considering that tax rates vary at the state and local level.29 Notably, some jurisdictions currently have restrictions on the sale of flavoured tobacco products, including those currently not under FDA mandate such as hookahs.30 ,31 Encouraging strong laws at the local level can help protect public health and serve as a template for stronger federal/state legislation.

This study's strength was the use of nationally representative data to analyse trends in tobacco consumption including industry-reported tobacco sales data as well as self-reports of current tobacco use behaviour among US adults, for different tobacco products, during 2002–2012. These data provide insight regarding the tobacco industry response to regulatory imbalances in the treatment of different tobacco products. However, these findings should be interpreted within the context of the following limitations. First, per capita sales data may not necessarily translate into actual consumption because not all tobacco products purchased may be actually used by the consumer due to loss, damage or quitting tobacco use. Second, the measure for per capita tobacco sales only accounted for products taxed for legal sale in the USA and did not account for illicit sales, such as those smuggled into or out of the country, or for untaxed tobacco products that are produced or sold on American Indian sovereign lands. Third, differential time frames in analysing industry-reported tobacco sales (past year) and self-reported current tobacco use (past 30 days) may not capture identical trends accurately. Fourth, our definition of smokeless tobacco assessed within the context of per capita sales (ie, moist snuff) was narrower relative to self-reported use (ie, snuff and/or chewing tobacco). However, moist snuff is the dominant smokeless tobacco product in the USA, accounting for 85.6% of all smokeless tobacco sales in 2011, whereas dry snuff and chewing tobacco use are relatively low and declining.32 Fifth, these analyses do not include consumption of emerging tobacco products such as hookahs, electronic cigarettes or snus because of lack of data.

Conclusion

This study demonstrates that although cigarette per capita sale and prevalence of use declined significantly during 2002–2012, significant increases were observed for other non-cigarette tobacco products. A switch from cigarettes to lower-taxed tobacco products blunts the effect of increasing prices, one of the most effective ways to reduce smoking and prevent youth smoking initiation. Eliminating imbalances in tax structure and regulations between cigarettes and alternative tobacco products may help reduce US aggregate tobacco consumption as well as tobacco use disparities.

What this paper adds

  • Imbalances in tax structure and regulations exist between cigarettes and non-cigarette tobacco products. We assessed whether declines in US cigarette consumption during 2002–2012 may have been offset by increased use of non-cigarette tobacco products.

  • During 2002–2012, US per capita sales of cigarettes declined significantly, while per capita sales of cigars, moist snuff and loose tobacco (roll-your-own plus pipe tobacco) increased significantly.

  • Self-reported tobacco use data also showed declines in current cigarette smoking prevalence during 2002–2012, whereas increases were observed in current use of smokeless tobacco and roll-your-own tobacco.

  • These findings underscore the need for an integrated approach to product regulation and taxation for all tobacco products to reduce aggregate tobacco use.

Acknowledgments

ITA initiated the reported research while being affiliated with the Center for Global Tobacco Control at Harvard University. He is currently affiliated with the Centers for Disease Control and Prevention's Office on Smoking and Health. The research in this report was completed and submitted outside of the official duties of his current position.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors ITA conceived the study, contributed to the data analyses and interpretation, and wrote the first draft of the manuscript. HRA contributed to the analysis and interpretation of the data and was involved with extensive writing and editing of the manuscript. Both the authors approved the final manuscript.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All data used in the current study are publicly available secondary data. No additional unpublished data are being withheld by the authors.