Background Smokers use cigarette expenditure minimising strategies (CEMS) to alleviate the effect of tax increases on their cigarette expenses. We examined changes in smokers’ CEMS use before and after a 2013 Minnesota $1.75 cigarette tax increase.
Methods Data were from representative samples of smokers who participated in the Minnesota Adult Tobacco Survey 2010 (n=948) and 2014 (n=1229). Participants indicated CEMS used in the past year from a list. Weighted multiple logistic regressions were used to examine changes in prevalence of each CEMS use over time adjusting for demographics and cigarette consumption. Characteristics associated with CEMS use in 2014 were examined.
Results Between 2010 and 2014, more smokers tried to save money on cigarettes by rolling their own cigarettes (from 19% to 29%), using other tobacco products (from 13% to 25%), and buying cigarettes from cheaper places (from 48% to 55%). Yet, fewer smokers used coupons/promotions (from 63% to 50%) and bought cigarettes by the carton (from 39% to 32%). These changes varied somewhat by race/ethnicity and education, for example, more smokers with <high school education used discount brands over time than more educated smokers. CEMS use in 2014 varied by demographics, for example, smokers with lower education were more likely than those with higher education to purchase discount brands, roll their own cigarettes, use coupons/promotions and cut back on smoking (p<0.05).
Conclusions Socially disadvantaged smokers were most likely to use CEMS and continue smoking after a cigarette tax increase. Regulations that would reduce CEMS use could boost the effectiveness of cigarette tax increases.
- Cigarette tax
- tax avoidance
Statistics from Altmetric.com
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.
Tobacco use accounts for six million deaths worldwide annually, and is projected to claim eight million lives globally, which is 10% of all deaths, by 2030.1 Tobacco taxes are the most effective tobacco control strategy included in the Framework Convention for Tobacco Control.2 Two systematic reviews also indicated that tobacco tax increases could reduce tobacco use disparities.3 4 However, previous research has indicated that smokers engage in cigarette expenditure minimising strategies (CEMS) to reduce their cigarette expenditures in response to increasing cigarette prices. These strategies often include: buying a cheaper brand of cigarettes, using discount coupons/promotions, buying from a cheaper place, rolling own cigarettes, using another form of tobacco, buying by the carton instead of pack, cutting down on consumption and sharing fewer cigarettes with other smokers.5 6 For example, data from the 2006–2007 US Current Population Survey Tobacco Use Supplement showed that 27.6% of adult smokers purchased cigarettes by the carton, and 3.2% purchased cigarettes from a different state with lower prices to save money on cigarettes.7 Previously, we examined the use of CEMS in a cohort of smokers from a US Midwest state immediately after the 2009 federal cigarette tax increase (US$0.61) and found that smokers engaged in a variety of CEMS (eg, bought from a cheaper place, used coupons/promotions, purchased cigarettes by the carton instead of pack, bought a cheaper brand, rolled their own cigarettes, and used another form of tobacco).5
However, the use of CEMS may hinder cigarette smoking cessation, especially among lower income smokers. Based on our previous work, those who purchased cigarettes by the carton to save money were less likely to make a quit attempt.5 We also found that any use of CEMS and the number of CEMS used was negatively associated with subsequent reduction in the number of days smoked in the past 30 days, after accounting for demographic and smoking-related social environment.5 Moreover, we found that lower income smokers were more likely to use CEMS and to use a higher number of CEMS.5 Therefore, the negative association between CEMS and cigarette smoking reduction and cessation may limit the effect of cigarette tax increases to reduce socioeconomic disparities in cigarette smoking.
The current literature on this topic is limited in the following ways. First, data are generally not available for some CEMS (eg, sharing fewer cigarettes with others to save money). Second, most studies are limited to the capture of CEMS use at one point in time. Therefore, previous studies were unable to assess how the use of CEMS changes before and after a cigarette tax increase. The current study aimed to overcome these limitations by providing data on an expanded list of CEMS, and to assess changes in CEMS use before and after a cigarette tax increase using serial cross-sectional surveys. In 2013, Minnesota, a US Midwest state, implemented a US$1.75 increase in the excise and sales taxes on cigarettes.8 Using the surveillance data collected through the 2010 and 2014 Minnesota Adult Tobacco Surveys (MATS), we assessed how the use of CEMS changed over time. We further assessed if these changes differed by the sociodemographics of the smokers. Finally, we examined the characteristics associated with CEMS after the recent larger cigarette tax increase (US$1.75) to explore if these characteristics differed from those after a smaller (US$0.61) tax increase in 2009.
Data were from the 2010 and 2014 rounds of the MATS. MATS is a statewide, cross-sectional telephone survey designed to assess tobacco use behaviour and attitudes among Minnesota adults aged 18 years or older. The survey design in both 2010 and 2014 included random-digit dialling sampling from two telephone frames—cell phone and landline telephone numbers. Sampling involved a two-stage process with a household screening questionnaire to identify households, and then sampling within the households. Cell phone sample numbers were screened and not included if they did not rely exclusively on their cell phone for voice communication. In 2014, the design additionally included geographic stratification to sample within eight regions around the state. The final analytic sample in 2010 was 7057 and 9304 in 2014. (See MATS Methodology Report for a more complete description of methods; http://www.mnadulttobaccosurvey.org.) MATS includes survey weights that account for sampling to ensure statewide representativeness. The current analysis included past 30-day smokers who participated in the MATS 2010 (n=948) and MATS 2014 (n=1229).
In the 2010 survey, participants were asked, ‘In the past year have you done any of the following things to try and save money on cigarettes (yes/no)?’ A list of six CEMS were queried: bought a cheaper brand of cigarettes; rolled your own cigarettes; used another form of tobacco other than cigarettes; used coupons, rebates, buy 1 get 1 free or other special promotions; purchased cartons instead of individual packs; and found less expensive places to buy cigarettes. The same question was included in the 2014 survey with two additional strategies added based on a study involving in-depth interviews with smokers: smoked fewer cigarettes and shared fewer cigarettes with others.6
Collected demographic information included age, gender, race/ethnicity, annual household income and education attainment. Because of the small number of participants who were racial/ethnic minorities (reflective of the Minnesota demographics), race/ethnicity was categorised as white versus others (including Black, Asian, Pacific Islander, American Indian, Hispanic, other/mixed). Cigarette smoking behaviours included number of cigarettes smoked and menthol brand. Participants were asked to report the number of cigarettes smoked per day, and responses were categorised into 1–10, 11–20 and >20 cigarettes per day. Participants reported whether their usual cigarette brand was menthol or non-menthol. We included usually smoking menthol cigarettes as a covariate because previous research has found that menthol cigarette smokers were less likely than non-menthol cigarette smokers to quit smoking among race/ethnic minority groups.9
Survey weights were included to account for multistage sampling and to be representative of Minnesota. First, we examined the changes in use of each of the six CEMS measured both in the 2010 and 2014 surveys using weighted multivariable logistic regression models, including survey year (2014 vs 2010), age, gender, race/ethnicity, annual household income, education attainment, cigarettes per day and smoking menthol cigarettes. Second, we tested the interactions between year and demographics on the use of each of these six CEMS, and presented stratified analysis by demographics that showed significant interactions with year. To test the interactions in each of these models, interaction terms between survey year and demographic variables, together with survey year, age, gender, race/ethnicity, annual household income, education attainment, cigarettes per day and usually smoking menthol cigarettes were included. Finally, we examined the characteristics associated with using each of the eight CEMS in 2014 using weighted multivariable logistic regression models adjusting for demographics, cigarettes per day and usually smoking menthol cigarettes. All analyses were conducted in SAS version 9.4, using PROC SURVEYFREQ and PROC SURVEYLOGISTIC.
Table 1 presents the weighted demographics and smoking behaviours of the 2010 (before the cigarette tax increase) and 2014 (after the cigarette tax increase) samples. We observed significant changes in the use of five out of the six CEMS (table 2). Between 2010 and 2014, an increasing proportion of Minnesota past 30-day smokers rolled their own cigarettes, used another form of tobacco, and found less expensive places to buy cigarettes (p<0.05). Finding cheaper places to buy cigarettes became the most used strategy to save money. In contrast, a decreasing proportion of smokers used coupons/promotions or purchased cartons instead of packs of cigarettes (p<0.05). Despite the decrease from 2010, using coupons/promotions remained the second most used strategy.
Table 3 summarises the significant differences in changes in use of CEMS by demographics before and after the cigarette tax increase. Buying a cheaper brand of cigarettes to save money was increasingly prevalent among white smokers compared with smokers of other races/ethnicities (interaction p=0.0272). This strategy was also more commonly used by smokers with less than high school education and those with college education and beyond, but not among smokers with high school/GED or some college education (interaction p=0.0036). Use of roll-your-own cigarettes was increasingly prevalent among both male and female smokers, but the increase was more substantial among female smokers (interaction p=0.0114). This strategy was also increasing common among white smokers than smokers of other racial/ethnic group (interaction p=0.0301). Use of coupons/promotions was reduced more substantially among smokers of other races/ethnicity than white smokers (interaction p=0.0271).
In addition, 69.8% smokers reported smoking fewer cigarettes, and 50.4% reported sharing fewer cigarettes with others to save money on cigarettes in 2014. Use of specific CEMS after the cigarette tax increase differed by demographics and smoking behaviours (table 4). Compared with smokers 65 years or older, younger smokers were more likely to have used the following strategies to save money on cigarettes: used another form of tobacco, used coupons/promotions, bought from cheaper places, cut back on smoking and shared fewer cigarettes with others (p<0.05). In contrast, younger smokers were less likely to have purchased cartons instead of packs to save money on cigarettes (p<0.05). Female smokers were more likely than male smokers to have bought cheaper brands, purchased cartons instead of packs, bought from cheaper places and share fewer cigarettes, but less likely to have used another form of tobacco (p<0.05). Smokers of races/ethnicities other than white were less likely than white smokers to have used coupons/promotion (p<0.05). Lower household income was associated with having bought a cheaper brand, rolled own cigarettes, used coupons/promotions, cut back on smoking and shared fewer cigarettes with others. Lower education was associated with having bought a cheaper brand, rolled own cigarettes, used coupons/promotions, bought from cheaper places, cut back on smoking and shared fewer cigarettes with others (p<0.05). Finally, smokers who smoked more than 10 cigarettes per day were more likely than those who smoked 10 or less cigarettes per day to have used coupons/promotions, purchased cartons instead of packs and bought from cheaper places (p<0.05). Usually smoking menthol cigarettes was not associated with CEMS use (p>0.05).
In 2014, Minnesota implemented a substantial cigarette tax increase, which increased the proportion of cigarette per pack retail price attributed to cigarette taxes from 44.9% in 2013% to 56.9% in 2014.10 We observed significant changes in CEMS use before and after the tax increase. Our previous work found that increases in cigarette prices exceeded the amount of the tax increase, and neighbourhood characteristics were not associated with the magnitudes of the price increases,11 suggesting that the tax increase was fully passed onto consumers. Therefore, we believe most, if not all, of the changes in CEMS use can be attributed to the state cigarette tax increase as there were no other significant statewide policy changes.
Our findings provide mixed information about the potential impact of a cigarette tax increase as a tobacco control strategy. On the one hand, substantial cigarette tax increases seem to promote self-reported reduction in cigarette consumption, although since the data were from two serial cross-sectional studies, we could not determine the magnitudes of such reduction within individual smokers. They also appear to reduce the availability of cigarettes through social sources in the population as smokers reported sharing fewer cigarettes with others to save money. This is important because social sources are a major source of cigarettes for smokers who are still developing their smoking behaviours.12 13 The reduction in discount coupons use does not seem to be due to availability of these coupons, since the tobacco industry spent US$236 million in 2010 and US$248 million in 2013 on this marketing strategy.14 A potential explanation for the decreases in using some CEMS was that smokers realised these strategies were not very effective over time to save money, as reported by a previous qualitative study that interviewed smokers before and after a cigarette tax increase.6 A previous US study also found that discount coupons only provide US$0.50 savings on average,15 which is less than a third of the amount of the tax increase (US$1.75). On the other hand, cigarette tax increases may prompt uptakes in other CEMS (eg, rolling own cigarettes, use another form of tobacco, finding less expensive places to buy cigarettes).
The role of a cigarette tax increase in reducing racial/ethnic and socioeconomic disparities in cigarette smoking was not clarified in this study. On the one hand, a substantial cigarette tax increase did not result in an increase in prevalence of purchasing a cheaper brand of cigarettes and rolling own cigarettes among racial/ethnic minority smokers (but increased among white smokers), despite previous studies showing that cheaper brands of cigarettes are prevalent in the USA.16 17 It also discouraged racial/ethnic minority smokers from using discount coupons/promotion more than white smokers. Since previous research showed that redemption of cigarette discount coupons was associated with lower likelihood of smoking cessation,18 a substantial cigarette tax increase could reduce racial/ethnic disparities in smoking through less use of CEMS. However, the tax increase did prompt more smokers with less than high school education to purchase cheaper brands of cigarettes. Additionally, after the tax increase, smokers of lower socioeconomic status were still more likely to engage in CEMS compared with higher socioeconomic smokers. Furthermore, while data showed that lower income (not education) was associated with CEMS use in response to the 2009 US$0.61 cigarette tax increase,5 lower income and lower education were both associated with CEMS use in response to the 2013 US$1.75 cigarette tax increase. Previous studies found that CEMS use was associated with lower likelihood of smoking cessation5 18; and therefore, if a substantial tax increase prompts increases in CEMS use among smokers of lower socioeconomic status, it may increase socioeconomic disparities among those who continue to smoke.19 20 Future longitudinal studies are needed to test this hypothesis. It is unclear why the increase in the prevalence of roll-your-own cigarettes was higher among female than male smokers, particularly given the historically lower prevalence of roll-your-own cigarettes use among women.21 Given female smokers were also more likely than male smokers to use a few other CEMS after the tax increase, this may be reflective of the higher price sensitivity among female than male smokers.22
While cigarette tax increases are effective in reducing the prevalence of smoking, a recent study found that only 14 US states and the District of Columbia raised their cigarette excise taxes between 2010 and 2014, compared with 46 US states and the District of Columbia during the period 2000–2009.23 The increase was much more modest in the latter period (US$0.20) than the former period (US$0.92). Cigarette tax increases remain an important public policy goal because they help smokers to commit to smoking cessation20 24; however, policies that impact CEMS use are also important as complementary policies to uphold the effect of cigarette tax increases. For example, the WHO Framework Convention on Tobacco Control Article 13 stated that prohibiting price promotions (eg, discount coupons) is part of the comprehensive marketing restrictions.25 Cities in the USA (eg, New York and Providence) have successfully implemented regulations to prohibit the redemptions of tobacco discount coupons. Additionally, increasing taxes for other tobacco products in conjunction with cigarette tax increases may prevent smokers from using roll-your-own cigarettes and other tobacco products to mitigate the effect of cigarette tax increases. The Children’s Health Insurance Program Reauthorization Act of 2009, for example, increased taxes on all tobacco products, including cigarettes, cigars, pipe tobacco and roll-your-own tobacco.26
Our study findings have limited generalisability to US states with higher racial/ethnic diversity or to other countries because the samples came from a single US Midwest state (Minnesota). However, we believe the implications of our findings are still applicable to other US states and even other countries since many WHO member countries still allow tobacco price promotions and have cigarettes at different price tiers.25 Also, respondents may not be able to accurately recall their use of CEMS in the past 12 months. We did not collect information on two CEMS (ie, smoke fewer cigarettes and share fewer cigarettes with others) in 2010, and therefore cannot assess how a cigarette tax increase influenced these CEMS.
In conclusion, the use of CEMS among smokers differed before and after a substantial cigarette tax increase in 2013. We also observed reductions in use of some CEMS and increases in use of some other CEMS after the tax increase. Changes in CEMS before and after the tax increase also varied somewhat by demographics. A combination of tax and non-tax policies to maintain high cigarette prices may be useful in upholding the effectiveness of cigarette tax increases on reducing the prevalence of smoking and its disparities.
What this paper adds
Tobacco taxes are the most effective tobacco control strategy. However, previous studies showed that smokers engaged in cigarette expenditure minimising strategies (CEMS) to save money on cigarettes and continue smoking. Previous studies also found that engaging in CEMS were also associated with lower likelihood of smoking cessation.
However, data were not previously available to examine how CEMS use changes before and after cigarette tax increases.
We found that, concurrent with a 2013 cigarette tax increase of US$1.75 in Minnesota, USA, more smokers tried to save money on cigarettes by rolling their own cigarettes, using other tobacco products and buying cigarettes from cheaper places. Meanwhile, fewer smokers used coupons/promotions and bought cigarettes by the carton. Changes in CEMS use varied somewhat by demographics.
Contributors KC and RGB designed and conducted the analysis in this manuscript, drafted and revised the manuscript, provided final approval of the version to be published and agreed to be accountable for all aspects of the work.
Competing interests None declared.
Ethics approval Westat Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.