Health Promotion International, Vol. 20 No. 1 © The Author 2005. Published by Oxford University Press. All rights reserved. doi:10.1093/heapro/dah506 For Permissions, please email: [email protected] Advance access publication 28 January 2005

Targeting smokers via tobacco product labelling: opportunities and challenges for Pan European health promotion Downloaded from

ELINOR DEVLIN, SUSAN ANDERSON, GERARD HASTINGS and LYNN MACFADYEN Cancer Research UK Centre for Tobacco Control Research, University of Stirling, http://heapro.oxfordjournals.org Department of Marketing, Stirling, FK9 4LA, UK

SUMMARY Cigarette on-pack messages are one of the principal increased effectiveness of targeting messages to specific vehicles for informing smokers about the risks of smoking segments of the population. Two main research areas were and research has highlighted their role as a valid health addressed. First, the extent to which the new messages were communication tool. Furthermore, they have the potential appropriate for smokers in Europe and second, the to disrupt the powerful cigarette brand imagery associated potential to provide targeted and personally relevant with tobacco packaging. Responding to concerns within messages to smokers via tobacco packs. Fifty-six focus Europe that the old style on-pack messages were ineffective groups were conducted across seven European countries and the introduction of new tobacco product legislation (Finland, France, Germany, Greece, Spain, Sweden and the at World Health Organization on May 21, 2010 across Europe (EU Directive 2001/37/EC), this study was UK) with 17–64-year-old smokers, half of whom were not conducted to explore European smokers’ response to the thinking about quitting (pre-contemplators) and half of changes. The research draws upon two main areas of health whom were thinking of quitting in the next 6 months communication: the need to pre-test messages to ensure (contemplators and preparers). Implications for future they are appropriate for their intended audience; and the labelling practices within Europe are discussed.

Key words: Europe; targeting; tobacco labelling

INTRODUCTION

Tobacco consumption is responsible for 1.2 products is an integral element of this strategy million deaths in Europe [Word Health (Kaiserman, 1993) with many countries such as Organisation (WHO), 1999]. This figure is Australia, Poland and Canada introducing estimated to rise to 2 million and account for one tougher legislation in this area (Borland, 1997; in five of all deaths in the European region by Przewozniak and Zatonski, 2002; Health the year 2020 (Haglund, 2000). A comprehensive Canada, 2003). tobacco control strategy including policy on taxa- On-pack messages are a valid health commu- tion, tobacco promotion, nicotine replacement nication tool. Research in Australia (Borland, therapies, smoke-free public places and health 1997) found that new, more prominent informa- promotion is vital to persuade current smokers to tion on tobacco packs (covering at least the top quit and encourage young people not to start 25% of the front and the top third of the back of (World Bank, 2003). The labelling of tobacco the pack) resulted in: increased noticing of the

41 42 E. Devlin et al. warnings; more potent stimulation of negative Table 1: EU Directive on tobacco product thoughts about smoking; and the premature regulation (2001/37/EC): content of warnings stubbing out of cigarettes already lit. Research Location on pack Statement also supports design enhancements. The effec- tiveness of on-pack information can be increased Front Smoking kills by: being made more prominent (Kaiserman, Front Smoking seriously harms you and 1993); being printed in contrasting colours (Ross, others around you 1981); varying the design and content (Bhalla Back Smokers die younger Back Smoking clogs the arteries and and Lastovieka, 1984); and by being developed causes heart attacks and strokes

for a target audience and informed by creative Back Smoking causes fatal lung cancer Downloaded from input and market research (Jacoby et al., 1982; Back Smoking when pregnant harms Fischer et al., 1993). Cigarette packaging is your baby Back Protect children: do not make them known to play an important role in honing and breathe your smoke supporting the imagery associated with powerful Back Your doctor or your pharmacist can cigarette brands (Carr-Gregg and Gray, 1993). help you stop smoking Research has shown that on-pack messages have Back Smoking is highly addictive, do not the ability to disrupt this brand imagery by start http://heapro.oxfordjournals.org Back Stopping smoking reduces the risk disrupting the available space on packs to of fatal heart and lung diseases communicate brand values and by undermining Back Smoking can cause a slow and the influence of commercial communication painful death (MacFadyen et al., 2001). Furthermore, Mahood Back Get help to stop smoking: (telephone, postal address, internet address, (1999) argues that an effective warning system consult your doctor/pharmacist) will create a situation of informed consent Back Smoking may reduce the blood flow regarding the nature of the risks, the magnitude and causes impotence of the dangers and the probability of occurrence Back Smoking causes ageing of the skin among smokers regarding the risks of tobacco Back Smoking can damage the sperm and decrease fertility smoke. Back Smoke contains benzene, nitrosamines,

Within Europe, until recently, the labelling of formaldehyde and hydrogen cyanide at World Health Organization on May 21, 2010 tobacco products was governed by the 1992 EU Labelling Directive (EU Directive 1992/41/EC), which required a general message to be present on the most visible surface covering at least 4% As a communication device, on-pack messages (6% for countries with two languages) and an should follow the principles of communication alternating specific warning on the back of the theory and practice. It is recognized that good pack covering around 5%. The messages were communication involves the active participation required to be clear and legible, printed in bold of both the sender and the receiver (Lannon letters and printed on a contrasting background. and Cooper, 1983; Meadows, 1983; Buttle, 1991). Research in the UK and across Europe found Effective health communications must therefore that the information provided fell short of be designed in consultation with the target group, informed consent (Health Education Authority, taking account of their needs and wants (Eadie 1998) and that the messages were barely visible and Smith, 1995). Audience research is generally and not noticed by smokers (ASH, 1998). conducted to formulate, develop and evaluate EU legislation has begun to address these health communications (Atkin and Freimuth, shortcomings with the introduction of the EU 1989; Nowak and Siska, 1995). It follows that Directive on Tobacco Product Regulation (EU messages on tobacco packs, designed under this Directive 2001/37/EC). The Directive standard- process, will be more effective in achieving their izes the design and content of messages across desired objective (Ross, 1981; Kaiserman, 1993; Europe and requires messages to be printed in Eadie and Kitchen, 1999). Furthermore, careful black on a white background covering at least targeting of messages will increase their 30% of the front and 40% of the back of the pack, effectiveness (Ross, 1981; Kaiserman, 1993) and with a 5-mm black border. The Directive also it is argued that on-pack health information prescribes a list of 16 messages (two for the front should be specific and targeted to certain segments and 14 for the back), which are to be randomly of the population (Krugman et al., 1994; Mahood, rotated by member states (see Table 1). 1999). Research was therefore conducted to Targeting smokers via tobacco product labelling 43 examine the extent to which the new measures Table 2: Profile of focus group sample per country meet the needs of smokers in Europe and to explore the potential for labels to provide Gender Age Smoking status targeted and personally relevant messages. 1 Male 17–24 Contemplator/preparer The ‘Stages of Change’ model (Prochaska and 2 Male 25–34 Pre-contemplator DiClemente, 1983) was used to segment the 3 Male 35–44 Pre-contemplator target market. Prochaska and DiClemente’s 4 Male 45–64 Contemplator/preparer 5 Female 17–24 Pre-contemplator ‘Stages of Change’ model posits that behaviour 6 Female 25–34 Contemplator/preparer change involves the progression through five 7 Female 35–44 Contemplator/preparer

stages: pre-contemplation, contemplation, pre- 8 Female 45–64 Pre-contemplator Downloaded from paration, action and maintenance. The model forms the basis of effective segmentation and allows messages to be designed that are tailored (those who are not thinking of quitting) and to the specific needs of the individual (Werch and ‘contemplators/preparers’ (those who are thinking DiClemente, 1994). The EU has identified those of quitting in the next 6 months). Throughout who want to quit as a priority group (Commis- this report, descriptions and analysis of sion Report, 1999). However, as the majority of ‘contemplators/preparers’ is simplified to http://heapro.oxfordjournals.org smokers are in the pre-contemplation stage it is ‘contemplators’. The sample profile for each vital to consider their views and ensure that any country is summarized in Table 2. strategy developed does not alienate them. This In order to ensure consistency across the research therefore explores the needs of smokers project, all countries used a specially designed in the pre-contemplation and contemplation/ recruitment questionnaire and followed agreed preparation stages. sampling procedures, although provisions were made for cultural differences. For example, recruitment of smokers in the UK was conducted METHODS door-to-door in residential locations, whereas Sweden advertized for volunteers. Similarly, all

The research was conducted to provide an respondents were offered an incentive, according at World Health Organization on May 21, 2010 understanding of smokers’ response to the range to local practice, to encourage their participation of measures outlined in the Directive including and to thank them for their contribution. labelling messages and design. Given that smokers’ The groups were held in informal venues such response to such stimuli is likely to result from sub- as the recruiters’ homes or a hotel. All groups conscious and entrenched attitudes, a qualitative were audio taped with respondents’ permission research approach was required. This approach and transcribed for thematic analysis. allows the researcher to elicit and understand underlying attitudes that cannot be directly observed or measured (Kumar et al., 1999). Procedure Focus groups were employed; a technique A semi-structured discussion guide was which is widely used in commercial market developed and translated into the relevant lan- research and the social sciences, and is extremely guages. While maintaining consistency, flexibility useful for exploring attitudes and perceptions was afforded to allow for cultural differences in and when exploring taboo or difficult behaviours questioning techniques, to enable the partici- such as smoking. pants to introduce their own salient points and to allow the researcher to explore issues as they arose. Sample Each group began with a warm-up discussion Fifty-six groups were conducted across seven of current smoking behaviour and attitudes countries (Finland, France, Germany, Greece, before focussing on response to both the old and Spain, Sweden and the UK) with 17–64-year-old the new style messages. This was examined by smokers. Groups typically comprised six to eight showing respondents packs of the three most respondents and lasted between 1 and 2 h. The popular cigarette brands in their country plus a groups were purposively sampled according to pack of Marlboro (see Table 3) incorporating age (17–24, 25–34, 35–44, and 45–64-year-olds), the old style messages and also mock packs gender and smoking status: ‘pre-contemplators’ incorporating the new labelling design. 44 E. Devlin et al. Table 3: Most popular brands in each country coding segments of the text according to the discussion guide and themes and issues resulting Country National brands International brand from the research. Important relationships Finland Belmont Marlboro between and among the codes were subsequently L&M North State explored. Each country was instructed to produce France Gauloises (Blondes) Marlboro a report based on their research. As all countries Gauloises (Noir) had used the same discussion guide and visual Winfield prompts, and to make the analysis process man- Germany West Marlboro Lucky Strike ageable, a generic report structure was provided

Camel for each country to follow. A new document was Downloaded from Greece Peter Stuyvesant Marlboro subsequently created which consisted of ‘cut and Assos paste’ text (Wilson, 2003) from the original reports Camel Spain Ducados Marlboro into separate sections relating to each topic. The Chesterfield re-categorized data was then printed out and key Fortuna patterns, relationship and emergent themes were Sweden Prince Marlboro systematically coded and explored. Right http://heapro.oxfordjournals.org Blend UK Benson & Hedges Marlboro RESULTS Lambert & Butler Mayfair Three key themes emerged from the research: the ineffectiveness of old message design and Table 4: Messages tested in research content; the strength of the new EU message format; and the potential of EU content. Statement Appeal category

Smoking kills Fear Old message design and content Smoking seriously harms you Fear/social and others around you When presented with the packs incorporating the

Smokers die younger Fear old message style, very few respondents in any of at World Health Organization on May 21, 2010 Smoking can cause a slow and Fear the countries spontaneously mentioned the painful death messages, even when looking at and describing the Smoking when pregnant harms Social your baby pack. While this may be due to avoidance it Protect children: do not make Social appeared mainly due to a lack of prominence. This them breathe your smoke lack of prominence was blamed on both design Your doctor or your pharmacist Cessation support problems and ineffective message content. As a can help you stop smoking result, the intended message virtually disappears.

Message content was then explored using . . . those little letters you never notice. 1 showcards of eight messages as outlined in the (Spain) Directive (see Table 4). These comprise three That is just part of the package. (France, male, 45–64, contemplator) message types: fear appeals, addressing the health consequences of smoking; social appeals, Many respondents, particularly those who addressing the effect of smoking on others; and exhibited a degree of ambivalence about their cessation support appeals. habit, appeared happy to conspire with these cues, and ignore or rationalize away the message. Analysis They also raised suspicions about the motives behind the messages, seeing them as cynical or The analysis process consisted of two stages. the hypocritical fulfilment of a legal obligation. First, the groups were analysed in each country and a country report was prepared. Second, ana- The tobacco industry washes its hands. They make the lysis of cross-country comparisons was conducted texts so small that it doesn’t bother anyone, but they and reported. can claim having warned people. In line with marketing research techniques, each (Finland, male, 25–34, pre-contemplator) focus group was transcribed for analysis (Wilson, 1demographics for Spanish quotes not provided by 2003). The transcripts were then analysed by research agency. Targeting smokers via tobacco product labelling 45 The content of the messages compounded the Southern Europe, and especially Greece, seemed problem. They were felt to say nothing new and indifferent to the new format, and a minority of had remained unchanged for many years. Conse- male respondents from Greece expressed quently, respondents tended to either reject them irritation and hostility towards them, seeing them as patronizing and worn out, or displace them, as invasive and pointless. arguing that they are only of relevance to It therefore appears that the new format is able children or new smokers. Again, a degree of to gain the attention of smokers, at least on initial rationalization was evident with many smokers presentation. Pre-contemplators are more cynical content to be ‘let off the hook’. than contemplators about the new format and

respondents from Greece and Spain were more Downloaded from They treat us like children, we’re adults by now, we likely to question the role of on-pack labelling. understand it’s not good for us. (Spain) EU content A small number of respondents, particularly Response to new message content was then those from Northern Europe and those consid- explored using a series of showcards as detailed

ering quitting, expressed a desire for larger, more http://heapro.oxfordjournals.org in Table 4. effective messages that addressed issues of relevance to them. Fear appeals EU message format This appeal category appeared to have a number of strengths. First, the messages were generally Respondents were then shown the more promi- clear, short and to the point and respondents nent message format as outlined in the Directive. liked their simple and direct nature. The new message format was generally the first aspect of the pack mentioned and clearly under- I think that was quite effective because it is just simple. mined its ability to communicate brand values. (UK, female, 25–34, contemplator)

It’s blunt, no beating around the bush. at World Health Organization on May 21, 2010 . . . the bigger it is, the better we see the warnings. It is (Spain) unusual so we pay attention to it. (France, male, 17–24, contemplator) Second, the tone of the message was perceived as realistic and appropriate given the seriousness The majority of respondents were supportive of the message. of the new format, perceiving it to add credibility Third, fear appeals evoked a number of to the intended message. Perhaps most impor- emotions among respondents such as anxiety and tantly, the new message format seemed to pro- guilt. This was particularly true for descriptive voke an emotional response among smokers who messages such as ‘Smoking can cause a slow and appeared quite shocked when first presented with painful death’. Finally, in the groups at least, the mock packs. It appeared to evoke feelings of these appeals gained the attention of smokers guilt and prompted both thoughts and discussion and prompted thoughts and discussion around around the negative aspects of smoking. the ill-effects of smoking. This often resulted in negative thoughts and attitudes about smoking. I feel sinful now and that is good. (Finland, female, 25–34, contemplator) However, the fear appeal category ignited a Scary. defensive reaction among some respondents. (UK, female, 45–64, pre-contemplator) Upon discussion some respondents viewed the messages as an over-simplification of the dangers There were exceptions however. Finnish and consequently they tended to rationalize the respondents in the pre-contemplation stage possible dangers. They compared smoking to other expressed some genuine irritation about the new activities such as alcohol consumption or road format and felt that it would not have much accidents in attempts to play down the dangers effect. They did, however, accept that the labels while also expressing the opinion that smoking would help young people and may deter others receives unfair focus. This was particularly true for from starting. In this respect they were willing to respondents from Southern European countries accept the role of the labels. Some respondents in and for those in the pre-contemplation stage. 46 E. Devlin et al. Using that guideline, they should also go after the cars comfortable with messages addressing this issue as well, they also kill. and viewed passive smoking more as an incon- (Spain) venience to others. Even at their most powerful (when focussing Furthermore, some respondents did not on children) these messages only prompted perceive the messages to be providing them with respondents to adapt their behaviour (e.g. any meaningful information. Consequently, they smoking in the garden or on the balcony). They tended to be rejected as patronizing and ‘worn did not suggest the need to quit. out’. This was particularly true for respondents from Northern Europe. Subgroup differences also emerged. The long- Support appeals Downloaded from term health effects of smoking tended to be more Contemplators, and particularly contemplators salient among those respondents in the contem- from Middle and Northern Europe, welcomed plation stage: they were more willing to elaborate the supportive tone of these messages and saw on fear messages and make some attempt to them as a relief from fear-orientated messages. personalize the effects. Respondents in the pre- They found them positive and encouraging, contemplation stage, however, made no attempt appreciating what they felt was a sympathetic http://heapro.oxfordjournals.org to personalize the messages and adopted a very and understanding approach, and indicated a defensive, hostile reaction dismissing the dangers likelihood to seek out such advice and support. as being too distant or unlikely. Furthermore, the long-term health effects of It is nice because if you want to quit, people can help smoking were not a salient issue among young you, even if you previously failed quitting. respondents who also found it difficult to person- (France, male, 17–24, contemplator) alize and relate to this type of appeal category. It’s encouraging, you CAN, we believe in you. They held the view that they would give up (Sweden, female, 35–44, contemplator) smoking before they were at serious risk from disease. Among more committed smokers and particu- larly respondents from the South, this type of

appeal category had little to recommend it. at World Health Organization on May 21, 2010 Social appeals These respondents viewed quitting as an indivi- When messages in this category addressed the dual activity, driven by self-motivation and issue of children, respondents in all countries consequently did not see support as either useful found such messages relevant and important, or feasible. In some countries, especially in believing they have a moral obligation to protect Southern Europe respondents viewed support children. Respondents, particularly females, were from doctors and pharmacists with scepticism. able to easily personalize and relate messages such as ‘Protect children: don’t let them breathe When I make the decision I will go by myself—doctors your smoke’ to their own children and grand- treat you like you’re useless. children and consequently such messages seemed (Spain) to evoke a highly emotional response.

Personally I get furious when adults smoke among kids DISCUSSION AND CONCLUSIONS because they are innocent and can’t do anything about it. (Sweden, male, 25–34, pre-contemplator) This was a cross-sectional, qualitative study and responses were explored in an artificial setting When social appeals did not relate specifically where the new labels were completely novel. to children (e.g. ‘Smoking seriously harms you Long-term impact, therefore, could not be and others around you’) country differences assessed, and generalizations have to be made emerged. The majority of respondents from with caution. Nonetheless a number of tentative Northern and Middle Europe found it reason- conclusions can be drawn, and recommendations able to modify their behaviour around those who made about future labelling practice. do not smoke. They were generally aware of the The research has confirmed that the EU dangers of passive smoking and were uncom- labelling format for tobacco packs, introduced in fortable when smoking around non-smokers. the EU Directive 2001/37/EC, is more noticeable Respondents in the South, however, seemed less than the one it replaces and has the potential to Targeting smokers via tobacco product labelling 47 help smokers who wish to quit. This supports the explore smokers’ comprehension and accepta- available literature (e.g. Kaiserman, 1993; bility. The variations also confirm the wisdom of Borland, 1997), which suggests that larger more segmenting audiences and targeting messages. prominent messages communicate more effec- Targeting on-pack messages is possible if tively. It was also clear that the novelty of the challenging. It would make sense, for example, to new format made an impact. Three mechanisms focus cessation oriented messages on Northern are at work here. First, their bigger more promi- Europe and more straightforward health messages nent format facilitates processing of the health on the South. It would also be possible to target messages and makes it difficult for the smoker according to age by attaching specific types of

who wants to ‘screen out’ or avoid them. The message to youth brands. Even potential quitters Downloaded from message was the first aspect of the pack noticed could be targeted by selecting appropriate brands and mentioned by smokers. Second, the new such as Silk Cut or other former ‘light’ cigarettes. format created ‘noise’ that undermined the This may seem a difficult route to follow, but it brands ability to interact with the smoker; as a is exactly what the tobacco industry does in all its result, brand imagery was diluted and impaired. marketing. Their internal documents show how Third, the fact that the messages are present at potential quitters, new recruits and regions the point of consumption means that they within the UK—let alone Europe—are targeted http://heapro.oxfordjournals.org provide very timely reminders of the risks of not just with customized messages but entire smoking and the benefits of quitting. marketing strategies (e.g. Hastings and However, the new labels do pose some MacFadyen, 2000). One lesson from this research difficulties in targeting messages across Europe. As is that health promoters need to approach this the segmentation and targeting literature suggests issue with the assiduity and resources as tobacco (Krugman et al., 1994), the labels were received executives. The resource would be immediately and interpreted differently between regions. In extended if policy makers decreed that the whole particular, respondents in Southern Europe were pack—and not just the health message area— less receptive to all the messages, and especially should become a platform for health promotion those addressing less familiar concepts such as rather than leaving it as a marketing tool for

passive smoking. They also viewed support from tobacco. This would also maximize the at World Health Organization on May 21, 2010 doctors and pharmacists with skepticism. This does opportunity for innovation and change that helps not necessarily mean different messages are keep messages prominent. needed for every member state, just that some By the same token, on-pack messages should be cultural sensitivity is desirable. seen as just one fragment of the ‘health marketing Two other segmentation criteria emerged from strategy’. The reticent response of the Southern the research: commitment to smoking and age. European smokers seems to reflect a generally Key differences were apparent between pre- more pro-smoking culture there than in the North contemplators’ and contemplators’ response to (Shafey et al., 2003). This suggests a need for a the content of the new labels. In the UK for broadly based tobacco control push to challenge example, respondents who were thinking of this norm and support the on-pack messages. At quitting wanted support and cessation advice and the moment they seem to be something of a voice hence paid attention to on-pack messages that in the wilderness. Similarly, in the North the provide this. In contrast, committed smokers are cessation support messages will only work if good inclined to ignore any messages, and therefore help lines and services are in place. In short it is need to have their attention grabbed. In terms of clear that on-pack warnings have to form part of a age, there were differences between younger and comprehensive tobacco control strategy. older smokers’ information needs. Younger Furthermore, labelling design and content respondents find the short-term health and should be varied and refreshed on a regular basis. cosmetic effects more salient, while older smokers Research conducted in the Netherlands, which are more concerned with illness and premature examined the effect of the new messages on ageing. smokers found that although they resulted in an These variations in response in terms of increase in noticing the messages, made tobacco region, smoking status and age fit with the health products less appealing and increased some communication literature and confirm that, in smokers’ motivations to quit or smoke less, the common with other types of campaign, on-pack impact of the messages decreased over time messages should be pre-tested in order to (Willemsen, 2002). Every opportunity should 48 E. Devlin et al. therefore be taken to maintain the novelty of the Borland, R. (1997) Tobacco health warnings of smoking- new messages. The EU could pursue the Canadian related cognition’s and behaviours. Addiction, 92, 1427–1435. and Brazilian example of introducing pictorial Buttle, F. (1991) What do people do with advertising? messages on packs, while other opportunities International Journal of Advertising, 10, 95–110. could include regularly introducing new state- Carr-Gregg, M. R. C. and Gray, A. J. (1993) ‘Generic’ ments and pictures to the current library. packaging—a possible solution to the marketing of With these measures in mind, tobacco tobacco products to young people. World Smoking and Health, 18, 2. packaging can become an even more effective Commission Report (1999) Progress achieved in relation to tool for both pan-European health promotion public health protection from the harmful effects of tobacco consumption. Commission report to the European

and tobacco control policy. Downloaded from parliament, the Council, the Economic and Social Committee and the Committee for the Regions, Brussels. Eadie, D. R. and Kitchen, P. J. (1999) Measuring the success ACKNOWLEDGEMENTS rate: evaluating the marketing communications process and marcom programmes. In Kitchen, P. J. (ed) Marketing The research was conducted in collaboration with Communications: Principles and Practice, Chapter 26. seven cancer societies throughout Europe and the International Thomson Business Press, pp. 459–476. Eadie, D. and Smith, C. (1995) The role of applied research authors acknowledge their contribution to the http://heapro.oxfordjournals.org in public health advertising: some comparisons with results presented in this paper: Dr Matti commercial marketing. Health Education Journal, 54, Rautalahti (Cancer Society of Finland), Ms 367–380. Sylviane Ratte (Ligue Nationale Française Contre EU Directive 1992/41/EC of the European Parliament and le Cancer), Dr Volker Beck (German Cancer of the Council of 15th May 1992. Official Journal of the European Communities. Society), Dr Nicolas Kordiolis (Hellenic Cancer EU Directive 2001/37/EC of the European Parliament and Society), Dr Dolors Marin-Tuya (Societat of the Council of 5th June 2001. Official Journal of the Catalana per a la Prevenció del Tabaquisme), Ms European Communities. Britt Marie Lindblad and Ms Tanja Tomson Fischer, P. M., Krugman, D. M., Fletcher, J. E., Fox, R. J. and (Swedish Cancer Society) and Ms Jean King Rojas, T. H. (1993) An evaluation of health warnings in cigarette advertisements using standard market research (Cancer Research UK). The principle sponsor of methods: what does it mean to harm? Tobacco Control, 2, the research was the European Commission with 279–285. financial support from Cancer Research UK and Haglund, M. (2000) Regional summary for the European at World Health Organization on May 21, 2010 all participating cancer societies. Thanks go to all region. In Corroa, M. A. et al. (eds) Tobacco Control Country Profiles. American Cancer Society, Atlanta, GA. the participants who took part in the focus group Hastings, G. and MacFadyen, L. (2000) Keep Smiling: No discussions across Europe. one’s going to die. An analysis of internal documents from the tobacco industry’s main UK advertising agencies. The Address for correspondence: Centre for Tobacco Control Research and the Tobacco Elinor Devlin Control Resource Centre. British Medical Association, Cancer Research UK Centre for Tobacco Control London, ISBN 0-7279-1600-9. Research Health Canada (2003) www.hc-sc.gc.ca (last accessed May University of Stirling 2003). Department of Marketing Health Education Authority (1998) Seminar Paper: Young Stirling, FK9 4LA People and Smoking. Health Education Authority, UK London. E-mail: [email protected] Jacoby, J., Nelson, M. C. and Hoyer W. D. (1982) Corrective advertising and affirmative disclosure statements: their potential for confusing and misleading the consumer. Journal of Marketing, 46, 61–72. REFERENCES Kaiserman, M. J. (1993) The effectiveness of health warning messages. Tobacco Control, 2, 267–269. ASH (1998) Summary of the EU Health Warnings Project. Krugman, D., Fox, R. J., Fletcher, J. E., Fischer, P. M. and By Comitè National Contre le Tabagisme; FARES Rojas T. H. (1994) Do adolescents attend to warnings in (Fondation contre les Affections Respiratoires et pour cigarette advertising? An eye-tracking approach. Journal l’Education à la Santé; ASH (Action on Smoking and of Advertising Research, November/December. Health). Kumar, V., Aaker, D. A. and Day, G. S. (1999) Essentials of Atkin, C. K. and Freimuth, V. (1989) Formative evaluation Marketing Research. John Wiley & Sons, Inc. research in campaign design. In Rice, R. E. and Atkin, C. Lannon, J. and Cooper, P. (1983) Humanistic advertising: a K. (eds) Public Communication Campaigns, 2nd edition, holistic cultural perspective. International Journal of Sage, Newbury Park CA, pp. 131–150. Advertising, 2, 195–213. Bhalla, G. and Lastovieka, J. L. (1984) The impact of changing MacFadyen, L., MacKintosh, A. M., Hastings, G. and cigarette warning messages and format. In Kinnear (ed) Devlin, E. (2001) Developing Improved Cigarette Advances in Consumer Research, 11, 305–309. Warning Labels for Young People. Final Report to The Targeting smokers via tobacco product labelling 49

Cancer Research Campaign. Centre for Social Marketing, Ross, K. (1981) Legal and practical considerations for the University of Strathclyde, Glasgow. creation of warning labels and instruction books. Journal Mahood, G. (1999) Warnings that tell the truth: breaking of Products Liability, 4, 29–45. new ground in Canada. Tobacco Control, S, 356–362. Shafey, O., Dolwick, S. and Guindon, G. E. (2003) Tobacco Meadows, R. (1983) They consume advertising too. Control Country Profiles, 2nd edition. American Cancer ADMAP, July/August, 408–413. Society, Inc., Atlanta GA. Nowak, G. J. and Siska, M. J. (1995) Using research to Werch, C. E. and DiClemente, C. C. (1994) A multi- inform campaign development and message design. In component stage model for matching drug prevention Maibach, E. and Parrott, R. L. (eds) Designing Health strategies and messages to youth stage of use. Health Messages: Approaches from Communication Theory and Education Research Theory and Practice, 9, 37–46. Public Health Practice. Sage Publications, Thousand Oaks. World Health Organization (1999) Combating the tobacco Prochaska, J. O. and DiClemente, C. C. (1983). Stages and epidemic. In The World Health Report 1999. World processes of self-change of smoking: Toward an Health Organization, Geneva. Downloaded from integrative model of change. Journal of Consulting and Willemsen, M. C. (2002) The Altered View of Tobacco Clinical Psychology, 51: 390–395. Products: The Effects of the New Health Warnings on Przewozniak, K. and Zatonski, W. (2002) Stygmatizing the Smokers. DEFACTO, the Netherlands. cigarette pack by enlarged health warnings: Polish Wilson, A. (2003) Marketing Research: An Integrated experience and results. 3rd European Conference on Approach. Prentice Hall. Tobacco or Health: ‘Closing the Gaps—Solidarity for World Bank (2003) http://www1.worldbank.org/tobacco Health’, 20–22 June 2002, Warsaw, Poland. (last accessed May 2003). http://heapro.oxfordjournals.org at World Health Organization on May 21, 2010