The Impact of Tobacco Tax Cuts on Smoking Initiation Among Canadian Young Adults

Transcription

1 The Impact of Tobacco Tax Cuts on Smoking Initiation Among Canadian Young Adults Bo Zhang, MPH, Joanna Cohen, PhD, Roberta Ferrence, PhD, Jürgen Rehm, PhD Background: Methods: Results: Conclusions: Although the majority of smokers initiate smoking during their teenage years, significant rates of initiation occur among young adults. Adolescents are more price sensitive than adults, but little is known about the impact of tobacco taxation on smoking initiation among young adults. Using a longitudinal design, this study examined the impact of decreased cigarette price, resulting from tobacco tax cuts, on smoking initiation among Canadian young adults aged 20 to 24 years. Using Statistics Canada s National Population Health Survey longitudinal file, this study examined young adults who did not smoke at baseline in (n 636, representing over 1 million young adults) and who were reassessed at follow-up ( ). Multivariable logistic regression analysis using bootstrap weights was conducted to estimate the impact of decreased cigarette price on smoking initiation. The analysis controlled for the potential confounding effect of sociodemographic and tobacco control variables. Sensitivity analyses were conducted. Price elasticity was estimated. Analyses were conducted in 2003 and Approximately 10% of young adults had initiated smoking at follow-up. Decreased cigarette price was significantly associated with higher smoking initiation (adjusted odds ratio per $1 decrease for a carton of cigarettes 1.15, 95% confidence interval [CI] , p 0.042). Sensitivity analyses showed similar results. Price elasticity was 3.36 (95% CI ). Young adults are sensitive to cigarette prices. Reductions in cigarette prices will lead to increased smoking initiation among this group. Tobacco taxation should be an effective strategy to reduce smoking initiation among young adults. (Am J Prev Med 2006;30(6): ) 2006 American Journal of Preventive Medicine Introduction Cigarette smoking is the major preventable cause of morbidity and mortality. 1 5 The importance of taxation has long been recognized as part of a comprehensive approach to reduce the burden of illness due to tobacco use Because, historically, a high proportion of smokers took their first cigarette during their teenage years, most previous studies have focused on preventing smoking in adolescents. 14,15 However, recent studies have shown that young adults are not immune to smoking initiation. 16,17 With restrictions on sales and marketing to adolescents, the tobacco industry has shifted its focus to young adults Although overall smoking is From the Ontario Tobacco Research Unit (Zhang, Cohen, Ferrence), and Department of Public Health Sciences (Zhang, Cohen, Ferrence, Rehm), University of Toronto, Toronto, Canada; Centre for Addiction and Mental Health (Ferrence, Rehm), Toronto, Canada; and Addiction Research Institute (Rehm), Zurich, Switzerland Address correspondence and reprint requests to: Bo Zhang, MPH, Ontario Tobacco Research Unit, 33 Russell Street, Room T511, Toronto, Ontario, Canada M5S 2S1. bo_zhang@camh.net. decreasing, U.S. studies have shown that smoking initiation in young adults is increasing. 16,23 26 In Canada, young adults (aged 20 to 24) have the highest smoking rate of any age group. 27 Only a few studies have examined the impact of cigarette taxation on smoking initiation. Some of these found no effect Other studies have reported that higher cigarette prices are negatively associated with smokinginitation. 12,32 34 Onlythreestudiesofsmoking initiation used a longitudinal design, and these had inconsistent results. No study has used a longitudinal design to estimate the impact of cigarette price on smoking initiation among young adults. In February 1994, the Canadian federal government and five provinces reduced cigarette and tobacco taxes in response to pressure to curb increased smuggling of exported Canadian cigarettes back into Canada. As a result, five provinces had relatively high cigarette prices, and five had prices about 50% lower. This natural experiment, involving a large decline in real price, provided an opportunity to examine the impact of cigarette prices on smoking initiation in young adults using a longitudinal study design. 474 Am J Prev Med 2006;30(6) /06/$ see front matter 2006 American Journal of Preventive Medicine Published by Elsevier Inc. doi: /j.amepre

2 Methods Survey Data This study used data from Cycle 1 ( ) and Cycle 2 ( ) of Statistics Canada s National Population Health Survey (NPHS) longitudinal file. A stratified two-stage sampling design was used to select household residents across Canada. 35 One knowledgeable person provided general demographic, socioeconomic, and health information about each household member. One randomly selected individual from the participating household provided in-depth information about his or her own health for the health component. Data were weighted to reflect the sample design, adjustment for nonresponse, and poststratification. Respondents Of the 17,276 randomly selected individuals from Cycle 1 who were followed in Cycle 2, 16,168 (93.6%) responded, 36 including 1088 young adults aged 20 to 24. Of these, 452 respondents were excluded who had smoked at baseline or had not provided information on smoking. The sample for this study comprised the 636 young adults aged 20 to 24 who did not smoke at baseline ( ) and were followed up in Outcome Measure Those who did not smoke at baseline ( ) but smoked (daily or occasionally) at the follow-up interview ( ), were considered to be new smokers. The odds of becoming a new smoker were used to determine the impact of cigarette price on smoking initiation among young adults. Independent Variables Cigarette price change was the primary independent variable. The retail price of a carton of cigarettes (200 cigarettes) at baseline and at follow-up was measured in constant (1986) dollars. Statistics Canada regularly monitored the retail price of cigarettes in 26 major Canadian cities until From 1994 onward, only cigarette price indices were reported. To calculate quarterly cigarette prices from 1994 to 1997, the December 1994 retail price and changes in provincial cigarette price indices were used. Provincial consumer price indices were then applied to calculate quarterly constant (1986) dollar cigarette prices for cigarettes. Respondents who did not live in one of these cities were assigned the average price of cigarettes for their province. The quarterly prices were assigned to respondents during the quarter that they were surveyed. Cigarette price change was determined for each respondent by subtracting cigarette price at follow-up from the price of cigarettes at baseline. Control Variables Many individual and policy factors are associated with smoking. 6,11,37,38 Toestimate the independent impact of cigarette price on smoking initiation, we simultaneously controlled for individual characteristics (age, gender, educational attainment, income adequacy, and marital status) and policy variables (smoke-free bylaws for restaurants, enforcement, signage, and tobacco-control expenditures). Information on municipal smoke-free bylaws was obtained from Smoking By-laws in Canada Smoke-free bylaws for restaurants, enforcement of restrictions, and signage requirements were coded as yes or no. Public tobacco-control expenditures per capita were collected from ten provincial health departments during the fiscal year 1994 to 1995, 40 calculated as amount spent on tobacco control (dollars per capita). To control for unobserved variation among provinces that may influence the impact of cigarette price on smoking initiation, 33 indicator variables for provinces were included in the model. Linkage of Data Sets The NPHS longitudinal file was merged with the price and policy data file based on standard geographic classification codes. These are province-unique five-digit numbers, where the first two digits represent the census division code, and the remaining three digits represent the census subdivision code. Sixty-five percent (n 415) of the young adult respondents had corresponding records in the price and policy data file. In the geographic areas where no policy data information was available, further investigation revealed that smoke-free bylaws had not been implemented in 80% of these cases. 11 Asa result, missing bylaw variables were coded as zero, indicating no smoke-free bylaws. Although minor misclassification error may have been introduced, it is believed that the advantage of increasing the total usable sample in this way outweighed the disadvantages of excluding these respondents. Analysis All analyses were conducted using SAS/STAT software, release 8.2 (SAS Institute Inc., Cary NC, 2001). Relationships between cigarette price change and smoking initiation by sociodemographic characteristics, and by tobacco-control policy variables were examined with bivariate analysis. Multivariable logistic regression analysis was used to estimate the influences of cigarette price change on smoking initiation, while controlling for sociodemographic characteristics, tobacco-control policies, and variations among provinces. All analyses were conducted at the Statistics Canada Research Data Centre (RDC) at the University of Toronto, where all output was reviewed by the RDC manager to ensure no breaches of confidentiality. No survey design information on strata or primary sampling unit was available. In taking into account the complex sample design underlying the survey, the bootstrap method was used to calculate exact variances. 41 All analyses were conducted between 2003 and The odds ratio (OR) for cigarette price change from the logistic regression model can be interpreted as the percent change in the odds of smoking initiation for every $1 decrease in the constant dollar price of a carton of cigarettes from baseline to follow-up. Price elasticity was estimated based on the OR and its 95% confidence interval (CI) for percentage price change baseline price follow-up price /baseline price 100 for each participant from the multivariable logistic regression model, controlled for potential confounders. For instance, if June 2006 Am J Prev Med 2006;30(6) 475

3 Table 1. Smoking prevalence, inflation-adjusted prices, and smoking initiation at baseline and follow-up among young adults, by province and tax cut status, NPHS longitudinal study, Canada, 1994/ /1997 Smoking prevalence at baseline (%) a Smoking prevalence at follow-up (%) a Province New Brunswick Nova Scotia Prince Edward Island Quebec Ontario Newfoundland Manitoba Saskatchewan Alberta British Columbia Tax-cut status Tax-cut provinces Non tax-cut provinces a Point estimates were calculated using sample weights. NPHS, National Population Health Survey. Price of 200 cigarettes at baseline Mean ($) a Price of 200 cigarettes at follow-up Mean ($) a Smoking initiation from baseline to follow-up(%) a the OR is (95% CI ), it means that for each 1% price decrease, the smoking initiation rate increases by 3.4% (a small OR approximates relative risk). Thus, the price elasticity is 3.4 (95% CI ). Sensitivity analyses were conducted for respondents with primary policy data, respondents who were not prone to obtain smuggled cigarettes (those who did not live within 40 km of the border of an adjacent province with cigarettes at least $5 cheaper per carton), and respondents who remained in the original province at follow-up. CI , p 0.042) (Table 3). The sensitivity analyses showed similar results. When controlling for individual characteristics, tobacco-control policies, and unobserved provincial variations, the price elasticity was 3.36 (95% CI , p 0.045) based on the OR of percentage price change and its 95% CI. This means that for each 1% price decrease, the smoking initiation rate increases by 3.36%. Results There were 361 respondents in the five tax-cut provinces (New Brunswick, Nova Scotia, Prince Edward Island, Quebec, and Ontario: weighted percentage 66%) and 275 in the five non tax-cut provinces (Newfoundland, Manitoba, Saskatchewan, Alberta, and British Columbia: weighted percentage 34%). Cigarette prices decreased in both tax-cut and non tax-cut provinces, but the decrease was much larger in tax-cut provinces. Smoking prevalence decreased in both groups of provinces from baseline to follow-up, but the decrease was slightly larger in non tax-cut provinces. Overall, tax-cut provinces had a higher proportion of smoking initiation (10.5%) than non tax-cut provinces (8.5%) (Table 1). Cigarette price change was significantly associated with smoking initiation: The greater the price reduction, the greater the likelihood of initiating smoking (OR 1.18, 95% CI , p 0.034) (Table 2). When controlling for individual characteristics, tobacco-control policies, and unobserved provincial variations, cigarette price reduction was significantly associated with smoking initiation (adjusted OR 1.15, 95% Discussion Participants were a representative sample of Canadian young adults aged 20 to 24, representing over 1 million young adults who were nonsmokers in The onset of smoking over 2 years was 9.9% with a significant independent association (adjusted OR 1.15, p 0.042) between decreased cigarette price and smoking initiation. The greater the price reduction, the higher the smoking initiation, after controlling for individual characteristics, tobacco-control policies, and unobserved provincial variations. These findings are consistent with other studies that showed that increased cigarette prices reduce smoking initiation, especially at later stages of smoking uptake. 12,32 34 Whether the magnitude of the effect of a tax cut would be the same if it occurred today, a decade later than , is an empirical question. On the one hand, smoking has become less socially acceptable in large part due to more comprehensive smoke-free bylaws. On the other hand, smoking prevalence among young adults remains very high. The body of literature on tobacco taxes clearly indicates that a tax cut would still encourage more young adults to start smoking. 476 American Journal of Preventive Medicine, Volume 30, Number 6

4 Table 2. Smoking initiation associated with individual characteristics and tobacco control policies among young adults: bivariate analysis, NPHS longitudinal study, Canada, 1994/ /1997 New smokers Yes a (n 101,537) No a (n 929,717) Smoking initiation (%) OR 95% CI b p value Variable (range) INDIVIDUAL CHARACTERISTICS Age, mean (95% CI) b ( ( ) 22.0 ( ) 9.9 c years) Gender Males 57, , Females 44, , Income adequacy Lowest quartile 16, , Lower/mid quartile 32, , Upper/mid quartile 26, , Highest quartile 14, , Educational attainment High school 15,041 58, Graduated high school 16, , Some post secondary 59, , College/university degree 10, , ** Marital status Not married 83, , Married 18, , POLICY VARIABLES Cigarette price change d Mean (95% CI) b ( $3.04 $6.11) 2.6 ( ) 1.5 ( ) * Tobacco control expenditure Mean (95% CI) b ($0.01 $1.87) 1.2 ( ) 0.9 ( ) Smoke-free bylaws for restaurants No 33, , Yes 67, , Bylaw enforcement No 37, , Yes 63, , Bylaw signage No 65, , Yes 36, , a Based on weighted numbers. b 95% CIs were estimated using bootstrap weights. c Proportion of smoking initiation for the full sample. d Cigarette price change was defined as price at baseline minus price at follow-up; a positive value in the range indicates price decreased at follow-up and a negative value indicates price increased at follow-up. *p 0.05; **p 0.01 (all bolded). CI, confidence interval; NPHS, National Population Health Survey; OR, Odds ratio. Price elasticity has been estimated from 0.3 to 0.5 for overall cigarette demand, and 0.26 to 0.74 for smoking participation. 42 Higher price elasticity has been reported for youth at 1.44 for cigarette demand and 1.20 for smoking participation. 43 Although the price elasticity in this study is larger than those reported in other studies, the discrepancy may be due to the differences in the study design, outcome, and participants. To our knowledge, no previous studies have estimated price elasticity based on smoking initiation among young adults aged 20 to 24 using a longitudinal design. Nonsmokers at baseline in this study might include some former smokers, as it was not able to distinguish former smokers from lifelong never smokers. When analysis was restricted to those answering no to the question Have you ever smoked cigarettes at all? asked at baseline, the result showed an even stronger relationship with borderline significance between cigarette price and smoking initiation (adjusted OR 1.22, p 0.062). Cigarette smuggling, misclassification due to missing primary policy data, and residence change between the two surveys could potentially bias the estimate of the effect of cigarette price on smoking initiation. The sensitivity analyses showed very similar unadjusted and adjusted odds ratios, for the full sample and for the three subgroups (those with primary policy data, those with a reduced chance of being June 2006 Am J Prev Med 2006;30(6) 477

5 Table 3. Multivariate logistic regression and sensitivity analysis of effect of cigarette price on smoking initiation among young adults, NPHS longitudinal study, Canada, 1994/ /1997 Adjusted Weighted % of full Sample sample OR 95% CI a p value a Full sample 100 Model 1 b * Model 2 c * Model 3 d * Respondents with primary policy data d * Respondents not prone to smuggling cigarettes d Respondents remained in the original province d * *p 0.05 (bolded). 95% CIs and p value were calculated using bootstrap weights. b Controlled for demographic variables only (age, gender, education, income, and marital status). c Controlled for tobacco control policy variables only (smoke-free bylaws for restaurants, bylaw enforcement, bylaw signage, and tobacco control expenditures). d Controlled for demographic variables (age, gender, education, income, and marital status), tobacco control policy variables (smoke-free bylaws for restaurants, bylaw enforcement, bylaw signage, and tobacco control expenditures), and unobserved province variation (i.e., a dummy variable for each province and Ontario as referent). CI, confidence interval; NPHS, National Population Health Survey; OR, odds ratio. exposed to smuggled cigarettes, and those remaining in their initial province of residence). There are some limitations that cannot be resolved by sensitivity analysis. Changes in smoking behaviors due to the tax cuts may already have occurred by the time of the first interview. However, this would weaken the relationship between cigarette prices and smoking initiation. The longitudinal NPHS file provides information on smoking behavior at only two discrete points in time; thus, we cannot be sure when the smoking behavior changed, and whether there were multiple changes in smoking behavior between the two interviews. Given that this is not a time series study, a trend of delayed smoking initiation, if any, cannot be ruled out. If the observed association is purely due to delayed adolescent initiation, it would be expected that smoking prevalence would be higher in than in However, this was found only in one tax-cut and two non tax-cut provinces. In recent years, multilevel analysis has been used to account appropriately for the connections between individuals within similar environments in influencing individual susceptibility to disease or behavior change. 44 Data in this study have ahierarchical structure, as people living in the same municipal area share the same tobacco-control policies. It would have been preferable to conduct multilevel modeling in this study, but most municipal areas had only two to three respondents and some municipalities had only one case. A multilevel analysis based on such small numbers in each cell would not be meaningful. Conclusion Using a strong design, this study found that decreased cigarette price was associated with higher smoking initiation among young adults. Since smoking habits tend to become firmly established in early adulthood, reducing smoking initiation in this age group is likely to lead to permanent reductions in cigarette smoking through adult life. 18,38 Tobacco taxation should be considered an important strategy to reduce smoking, since taxation is not only effective, but also the most cost-effective way to reduce tobacco-related harm. 45 Some may argue that a tax increase in cigarettes is regressive, as this is the traditional view. 46 However, this statement was made based on the assumption of equal response by all income classes. 46 Theliteraturehasshownthatthepooraremore price sensitive than the rich; that they are more likely to quit or cut back when cigarette prices increase Incorporating such differential responses would make tax increases in cigarettes less regressive. 46 People will respond differently to tax increases in cigarettes. Increase in cigarette prices may prevent some youth from smoking and force some smokers to quit or cut back. Still others will not change their smoking behavior at all. Society as a whole will benefit substantially not only from reduced active smoking, but also from decreased secondhand smoking. Nevertheless, assessment of the equity of taxes should not be neglected. More research on aids for smokers to quit (e.g., nicotine patch and gum) and compensation on these drugs will be needed. This research was supported by the National Cancer Institute of Canada with funds from the Canadian Cancer Society (grant 8096). All analyses were conducted at the Statistics Canada Research Data Centre at the University of Toronto. Ed Adlaf, PhD, Rita Luk, and Charles Victor provided help with the analysis. Thomas Stephens, PhD, helped in collecting and coding some of policy data. No financial conflict of interest was reported by the authors of this paper. 478 American Journal of Preventive Medicine, Volume 30, Number 6

6 References 1. National Cancer Institute. Changes in cigarette-related disease risks and their implication for prevention and control. Smoking and tobacco control monograph 8. Bethesda MD: U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, U.S. Department of Health and Human Services. The health consequences of smoking: a report of the Surgeon General. Atlanta GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Burns DM. Epidemiology of smoking-induced cardiovascular disease. Prog Cardiovasc Dis 2003;46: Twombly R. Tobacco use a leading global cancer risk, report says. J Natl Cancer Inst 2003;95: Centers for Disease Control and Prevention. Cigarette smoking-attributable morbidity United States, MMWR Morb Mortal Wkly Rep 2003;52: U.S. Department of Health and Human Services. Reducing tobacco use: a report of the Surgeon General. Atlanta GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Cancer Institute. Changing adolescent smoking prevalence: where it is and why. Smoking and tobacco control monograph 14. Bethesda MD: U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute, Biener L, Harris JE, Hamilton W. Impact of the Massachusetts tobacco control programme: population-based trend analysis. BMJ 2000;321: Fichtenberg CM, Glantz SA. Association of the California tobacco control program with declines in cigarette consumption and mortality from heart disease. N Engl J Med 2000;343: Kaplan RM, Ake CF, Emery SL, Navarro AM. Simulated effect of tobacco tax variation on population health in California. Am J Public Health 2001;91: Stephens T, Pederson LL, Koval JJ, Macnab J. Comprehensive tobacco control policies and the smoking behaviour of Canadian adults. Tob Control 2001;10: Liang L, Chaloupka FJ. Differential effects of cigarette price on youth smoking intensity. Nicotine Tob Res 2002;4: Tauras JA. Public policy and smoking cessation among young adults in the United States. Health Policy 2004;68: U.S. Department of Health and Human Services. Preventing tobacco use among young people: a report of the Surgeon General. Atlanta GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, Sugarman SD. A balanced tobacco control policy. Am J Public Health 2003;93: Lantz PM. Smoking on the rise among young adults: implications for research and policy. Tob Control 2003;12(suppl I):i Backinger CL, Fagan P, Matthews E, Grana R. Adolescent and young adult tobacco prevention and cessation: current status and future directions. Tob Control 2003;12(suppl IV):iv Chaloupka FJ, Wechsler H. Price, tobacco control policies and smoking among young adults. J Health Econ 1997;16: Sepe E, Glantz SA. Bar and club tobacco promotions in the alternative press: targeting young adults. Am J Public Health 2002;92: Sepe E, Ling PM, Glantz SA. Smooth moves: bar and nightclub tobacco promotions that target young adults. Am J Public Health 2002;92: Ling PM, Glantz SA. Why and how the tobacco industry sells cigarettes to young adults: evidence from industry documents. Am J Public Health 2002;92: Biener L, Albers AB. Young adults: vulnerable new targets of tobacco marketing. Am J Public Health 2004;94: Emmons KM, Wechsler H, Dowdall G, Abraham M. Predictors of smoking among U.S. college students. Am J Public Health 1998;88: Wechsler H, Rigotti NA, Gledhill-Hoyt J, Lee H. Increased levels of cigarette use among college students: a cause for national concern. JAMA 1998;280: California Department of Health Services. Smoking prevalence among year olds: Tobacco Control Section: fact sheet, March Available at: Accessed August 9, Watson JM, Scarinci IC, Klesges RC, et al. Relationships among smoking status, ethnicity, socioeconomic indicators, and lifestyle variables in a biracial sample of women. Prev Med 2003;37: Health Canada. Canadian Tobacco Use Monitoring Survey (CTUMS): summary of results for Wave 1 (February to June) of Ottawa: Health Canada, Douglas S, Hariharan G. The hazard of starting smoking: estimates from a split population duration model. J Health Econ 1994;13: Douglas S. The duration of the smoking habit. Econ Inq 1998;36: Forster M, Jones AM. The role of tobacco taxes in starting and quitting smoking: duration analysis of British data. Working paper. Heslington, York, UK: University of York, DeCicca P, Kenkel D, Mathios A. Putting out the fires: will higher taxes reduce youth smoking? J Polit Econ 2002;110: Dee TS, Evans WN. A comment on DeCicca, Kenkel, and Mathios. Working Paper. Atlanta: Georgia Institute of Technology, School of Economics, Tauras JA, O Malley P, Johnston L. Effects of price and access laws on teenage smoking initiation: a national longitudinal analysis. NBER working paper Cambridge MA: National Bureau of Economic Research, Emery S, White MM, Pierce JP. Does cigarette price influence adolescent experimentation? J Health Econ 2001;20: Tambay JL, Catlin G. Sample design of the National Population Health Survey. Health Rep 1995;7:29 38, Swain L, Catlin G, Beaudet MP. The National Population Health Survey its longitudinal nature. Health Rep 1999;10: Chaloupka FJ, Pacula RL. Sex and race differences in young people s responsiveness to price and tobacco control policies. Tob Control 1999;8: Liang L, Chaloupka F, Nichter M, Clayton R. Prices, policies and youth smoking, May Addiction 2003;98(suppl 1): Health Canada. Smoking by-laws in Canada Ottawa: Office of Tobacco Control, Health Protection Branch, Health Canada, Cunningham R. Evaluation of federal and provincial legislation to control tobacco, Ottawa ON: Canadian Cancer Society, Yeo D, Mantel H, Liu TP. Bootstrap variance estimation for the National Population Health Survey. Proceedings of annual meeting of American Statistical Association, Survey Research Methods Section, Baltimore MD, Chaloupka F, Warner KE. The economics of smoking. In: Culyer AJ, Newhouse JP, eds. Handbook of health economics. Vol. 1. Amsterdam: Elsevier Science, 2000: Lewit EM, Coate D, Grossman M. The effects of government regulation on teenage smoking. J Law Econ 1981;24: Duncan C, Jones K, Moon G. Context, composition and heterogeneity: using multilevel models in health research. Soc Sci Med 1998;46: Shibuya K, Ciecierski C, Guindon E, et al. WHO Framework Convention on Tobacco Control: development of an evidence-based global public health treaty. BMJ 2003;327: Remler DK. Poor smokers, poor quitters, and cigarette tax regressivity. Am J Public Health 2004;94: Chaloupka F. Rational addictive behavior and cigarette smoking. J Pol Econ 1991;99: Townsend J, Roderick P, Cooper J. Cigarette smoking by socioeconomic group, sex, and age: effects of price, income, and health publicity. BMJ 1994;309: Centers for Disease Control and Prevention. Response to increases in cigarette prices by race/ethnicity, income, and age groups United States, JAMA 1998;280: Evans WN, Farrelly MC. The compensating behavior of smokers: taxes, tar and nicotine. RAND J Econ 1998;29: June 2006 Am J Prev Med 2006;30(6) 479