SPECIAL COMMUNICATION
A strategy for controlling the marketing of tobacco products:a regulated market modelR Borland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
Correspondence to:Ron Borland, VicHealthCentre for TobaccoControl, Cancer ControlResearch Institute, TheCancer Council Victoria, 1Rathdowne Street, Carlton,Vic 3053, Australia;[email protected]
Received 20 July 2003Accepted26 September 2003. . . . . . . . . . . . . . . . . . . . . . .
Tobacco Control 2003;12:374–382
Objective: To outline a novel strategy for controlling the tobacco market.Arguments: More comprehensive controls over the tobacco market are essential and long overdue.Effective controls need to encourage the development of less harmful products; control commercialcommunication to ensure that potential harms are highlighted relative to any benefits; and providemechanisms to move consumers away from tobacco use, or at least towards less harmful alternatives.Achieving this by regulating the existing industry is one strategy. This paper puts the case for analternative: to have marketing controlled by an agency (called here the Tobacco Products Agency, or TPA)which tendered to manufacturers for product and which distributed to retailers in ways that reduceincentives to bend or break the law. The TPA would be backed by legislation that made tobacco acontrolled substance with possession sale and use only allowed as permitted by the regulations, which inreality would be only as provided by the TPA.Conclusions: The overall effect of such a model, which we call a ‘‘regulated market model’’, would be toeliminate most of the incentives and remaining opportunities for commercial promotion of tobacco and tocreate incentives to encourage the development of less harmful tobacco products. Such a model preservesthe competition inherent in a free market, but directs it towards the challenge of reducing the harm fromtobacco use.
Tobacco smoking causes so much disease1 2 that it isestimated to kill approximately half of its regular longtime users.3 There is no known safe exposure level to
smoked tobacco products, although some forms of smokelesstobacco product (for example, the Swedish product calledsnus) are notably less harmful than smoked products.4 Theharm caused by tobacco products is predominantly due tolong term exposure to tobacco related toxins. The worldcommunity is currently in the process of ratifying aFramework Convention on Tobacco Control (FCTC) to helpcontrol the problem.5
The challenge of minimising the social costs from tobaccouse can be encapsulated in four major tasks: discouragingpeople from using tobacco, encouraging existing users toquit, protecting non-smokers from exposure to harmfultobacco smoke, and reducing exposures to tobacco relatedtoxins among any continuing users. The challenges involvedin preventing tobacco use and facilitating smoking cessationwill be minimised if we can eliminate activities of the tobaccoindustry that make the task more difficult by making tobaccouse more attractive than it intrinsically is. This effectivelymeans exercising control over the tobacco market.
Recently, Liberman and Clough6 have canvassed argu-ments about the possible illegality and certain immorality ofmany current tobacco industry practices. They make acompelling case that many current industry practices maynot only be subject to civil sanctions, but may also breakcriminal laws. In Australia, this includes laws to preventconduct endangering life and corporate manslaughter.Liberman,7 in a related paper to this one, argues that thefundamental problem is the nexus between profitability andthe harm caused by tobacco: the more you sell, the moreprofit you make, the more people you harm nexus needs to bebroken by changing the incentive structures for operating inthe tobacco business. He goes on to suggest the desirability ofgovernment control over both manufacturers and retailers.
This paper explores the viability of controlling the tobaccomarket by focusing on controlling two main ways tobaccocompanies influence tobacco use: through designing productsto be more attractive to consumers, and by promoting themin ways that add extrinsic value (fig 1). By promoting theirproducts and by adding ingredients to their products to makethem more attractive to consumers, tobacco companies areincreasing the value of engaging in an inherently dangeroushabit, one that is highly addictive.
The analysis in this paper is grounded in the science baseconcerning the composition and formulation of tobaccoproducts and the behavioural implications of how thatinformation is communicated.8 9 It adapts ideas fromothers8 10 11 and from the work of our group.7 12 There isnow an emerging consensus that tobacco products need to bemore strongly controlled,8 13 but no clear consensus as to howto do this. Currently, no country has anything approaching acomprehensive regulatory framework in place: one that caneffectively control both product composition and productpromotion. One aim of this paper is to broaden thinkingtowards alternative models of control by outlining onepossibility that minimises the role for a formal regulator.
Major gains have been made in tobacco control incountries like Australia that have taken the issue ser-iously.14 15 There is good evidence that strong restrictions onthe promotion of tobacco products and other countermeasures reduce smoking.16 However, even among the mosteffective and best resourced jurisdictions, anything fromaround 15% of the adult population continue to smoke on adaily basis and in most countries many more do so.14 It is alsoproving extremely difficult to prevent young people becoming
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abbreviations: FCTC, Framework Convention on Tobacco Control;RMM, regulated market model; TPA, Tobacco Products Agency
374
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
addicted.17 The tobacco industry has been remarkablysuccessful in undermining public health efforts, probablybecause only short periods of experimentation are needed fordependence to develop, and in part because countries havenot consistently invested the resources needed to overcomethe effects of tobacco industry marketing. It is inefficient tospend public money to overcome company activity, ratherthan more effectively controlling it.
Better controls will not eliminate tobacco use. At leastsome smoking is due to dependence and experienced benefits(independent of any debate about whether these experiencesreflect any enduring benefit). There is likely to be acontinuing demand for tobacco products, and as a resultprohibition will not eliminate use, and could create socialharm by making lawbreakers of otherwise good citizens. Thatsaid, tobacco should not be treated as an ordinary consumerproduct. If tobacco products are to be legally available, thereneed to be controls to ensure that they do not have valueadded to them, and are as harm reduced as possible. It is alsocritical that any reductions in the harmfulness of individualproducts do not have unwanted effects such as encouraginggreater use, or acting as a conduit to use of more dangerousforms.
If controls are to be effective, they need to be framed withan understanding of the realities about tobacco. Inhalingcigarette smoke is currently the most efficient way ofdelivering nicotine in a psychologically desirable way, makingcigarette smoke the most addictive form of tobacco. Most ofthe harm from tobacco use comes from chemicals other thannicotine. Many of the harmful chemicals are created bycombustion. Cigarette smoke is particularly harmful in thisregard. Thus the most attractive form of nicotine delivery alsohappens to be one of the most toxic. Past attempts to breakthis nexus between attractiveness/addictiveness and harm-fulness have failed. This may be partly because of constraintsthe industry has imposed on itself by trying to pretend thatsmoking is not really harmful or addictive.
After decades of deceit and denial some tobacco companieshave admitted that their products are harmful, and nowclaim to be diligently moving towards developing lessharmful products.18 However, there is currently no tobacco
product which has both a lower harm profile and will be usedas a substitute for cigarettes by most tobacco users.
Tobacco companies, like all corporations, exist to maximiseshareholder value. Monetary profits are an important part ofthis, but other less tangible benefits, such as pride in owningshares (or in the case of tobacco companies, shame),contributes to share and thus company value. Tobaccocompanies get no benefit from killing half their long termcustomers. Unfortunately, without less harmful products,they are caught in a dilemma: to keep making money, theyhave to cause harm and can do little to reduce that harm.
Tobacco companies have little alternative than to search, orat least seen to be searching, for less harmful products.Currently some tobacco companies, like Philip Morris,19 areseeking regulative protection. This is presumably mainly toprotect them from endless litigation, but this is notinconsistent with facilitating the search for less harmfulproducts. Any such regulation should give priority to thegoals of public health and broader social wellbeing. A keychallenge for tobacco control is to avoid allowing productsthat appear to have harm reduction properties, only to findthat they provide little or no benefit to health. The false harmreduced innovations such as filters and light cigarettes9 havefuelled scepticism about the value of harm reduction.However, failure to regulate effectively is the surest way ofallowing these mistakes to be repeated. The possibility of lessharmful products is one important reason for urgentlyneeding a comprehensive regulatory framework.
COMPREHENSIVE REGULATION OF THE TOBACCOINDUSTRYThe rationale for controlling the tobacco market, as part of acomprehensive tobacco control strategy, is to minimisepopulation harm from tobacco use by controlling the formand contents of, and information about, tobacco products inways that minimise population exposures to tobacco relatedtoxins.
In principle, the harmfulness of a tobacco product can bereduced in three ways: by making it less toxic per unit used;by making it less addictive per unit used; and/or by making itless palatable. The first of these reduces harm directly, unless
Figure 1 Tobacco marketing andmajor ways it can affect tobacco use.
Controlling the marketing of tobacco 375
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
there is compensation by increased use; while the latter twoshould reduce motivation to use and/or continue use, thusreducing lifetime exposures. The regulatory framework needsto allow progress on all three.
The ways in which the products are used affects theirharmfulness. Use is affected by characteristics of the product,experiences of using, and by beliefs about the consequencesof use. Consumer beliefs and behaviours are affected by whatis communicated to them, including by the ways tobacco ismarketed.
Regulation of tobacco products needs to ensure theprovision of both adequate consumer information andeffective controls on marketing. However, tobacco users aretypically dependent on the product and are severely limited intheir capacity to assess long term consequences of use. Thisextends to low capacity to properly assess harm reduction orexposure reduction claims. Thus, the preferred option forchanges in product toxicity is to continually upgradestandards of exposure (or harm) for products, such that thedirtier forms of products are phased out. This brings changeunder the control of health agencies, rather than relying onconsumers who are not able to assess the claims properly.20
That is, it eliminates the need for claims about relativeharmfulness, which have a high capacity to mislead.
Controls over promotion are needed, among other things,to reduce the creation of extra social value for tobacco use.There need to be controls over packaging, product informa-tion, and sales. Governments are already moving to mandateincreasingly strong and comprehensive warning messages onpackages because tobacco companies have failed to actappropriately. Governments also realise that with productsas dangerous as tobacco, promotion should be eliminated,but perhaps because some form of promotion is essentialwhile companies market to consumers, have failed to ban italtogether. The fundamental question here is: what residualsocietal benefit is being retained in maintaining a directrelationship between tobacco manufacturers and consumers?In our view, there is no practical use, it arguably only acts toretain incentives for tobacco companies to subvert the lawsthat are designed to control their activities.
To be effective, regulators will also need ongoing access tosuch things as information about the composition andengineering of tobacco products, exposures when used,indicators of harm, patterns of use, effects of price, consumerbeliefs, and effects of communication about the product. Theregulatory framework should allow for the stable, yet flexible,control of all aspects of tobacco products and theirmanufacture, promotion, and distribution.
Regulation needs to compatible with the free enterprisesystem because, through the agreements administeredthrough the World Trade Organization, free enterprise iseffectively mandated for ordinary consumer products world-wide. The free enterprise system is one that is extraordinarilydynamic, which uses competition to encourage innovationand efficiency. It is also extremely good at building marketsfor products. Unfortunately, the latter is not a trait that isdesirable for products that society wishes to discourage.
There are a number of other constraints on achievingoptimal regulatory control over tobacco use.
N Consumer perceptions and preferences are critical to thesuccess of any strategy to change patterns of use.
N Competition to develop less harmful products is con-strained because consumers have poor capacity to identifysuch products correctly. This historically has allowed theproliferation of products with the appearance of reducingsome of the harm, but little of the reality.
N Companies have no natural incentive to reduce theaddictiveness of their products because users who become
dependent provide a long term market. Indeed, they haveincentives to increase addictiveness and to hide addictionenhancing modifications from regulators.
N The capacity of manufacturers for innovation in productdesign and their natural advantage in knowledge of theeffects of innovations means that regulators will always betrying to catch up. Regulators will be under pressure to beconservative to reduce the risk of allowing innovationswhich have undesirable effects, but which might other-wise be attractive.
N Companies need to communicate with consumers aboutpositive features of their products to encourage use. Theyonly provide information on harms when required to doso, and they have incentives to try to overshadow suchinformation by spending more resources promoting thepositive features.
N Companies need to identify their products, if consumersare to be able to choose them. Proprietary brand names arethe main means of achieving this. If companies areexpected to compete for the consumer market, thenrestricting branding by mandating completely genericpackaging becomes problematic: there needs to be someresidual link to the manufacturer/marketer.
N Where direct means of communication are restricted,companies have incentives to seek out indirect means. It isvery difficult to prohibit some forms of indirect promotionwhile products are branded by manufacturers. Becauseindirect promotion is often about associating productswith desirable lifestyles, it is doubly dangerous as it notonly promotes use of the products, but does so in a waythat focuses on non-essential elements and not intrinsic(including harmful) aspects.
N Companies compete in part on the basis of the ‘‘quality’’ ofthe products they produce. Quality includes the sensoryexperiences of use, largely taste and smell, but also packimagery and the look of the product. Products withdifferent sensory characteristics that are not immediatelyobvious before purchase (such as so-called light cigarettes)need to have some labelling to identify them if consumersare to be able to choose them.
N Current regulation makes it harder to market potentiallyless harmful nicotine products than it does to marketcigarettes because the latter are often caught up intherapeutic goods laws. Also in some countries, likeAustralia, smokeless tobacco is prohibited, even thoughsome of these products are much less harmful thansmoked tobacco.
Taken together, these issues mean that regulators arecontinually in an antagonistic relationship with tobaccomanufacturers and distributors because for the most partthey do not share common goals. Tobacco markets haveincentives to avoid controls that restrict their capacity toinfluence consumers and have a knowledge advantage overregulators who might wish to stop them. What is the bestway to regulate such products? If a system could be devisedwhere the incentives on tobacco companies were consistentwith the goal of harm reduction, the challenge of regulationwould be greatly simplified. Control over marketing couldprovide such a solution.
Marketing is where the power lies in the modern tobaccoindustry. Marketers can determine which products will besold, and thus can exercise virtually complete control overmanufacture (and thus indirectly, growers), because theychoose which products to sell. They can also build image ondistinguishing characteristics of products, and/or just on thebrand names themselves. Ownership of brands and thecapacity to exploit them is at the core of their business.
376 Borland
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
Marketers also have considerable influence over retailactivity, and can influence retail prices through settingwholesale prices. They need to communicate with consumersabout their products, although in the case of tobacco, mostremain stonily silent about the harmful side of the ledger.Control over marketing provides levers to change incentivesto facilitate more effective tobacco control regulation.
A REGULATED MARKET MODELThe proposed model to control tobacco marketing is a versionof what we call a regulated market model (RMM). Under theRMM, free enterprise companies would retain the right tomanufacture, but a monopsonistic agency would be set up tomarket tobacco products. This agency, which we call theTobacco Products Agency (TPA), would need a charter thatspecifies that it will service the existing market, but shape itto minimise harm. The TPA becomes the sole customer ofmanufacturers and importers. The TPA would requirecapacity to assess performance characteristics of products toallow it to make informed choices. Free enterprise growerswould sell to licensed manufacturers who would tender formarket share from the TPA. The TPA would control wholesaledistribution to retailers. The TPA could allow the currentpractice of for-profit private retailers to continue withretailers having contractual arrangements with the agencyrather than with the manufacturers. Figure 2 is a schematicdiagram of the model. This is a fundamentally different
model to a government monopoly. It is similar in manyrespects to the way alcohol is marketed in Scandinavia, mostCanadian provinces, and some US states, where the govern-ment controls distribution and sales. The key difference fromthe US and Canadian alcohol distribution monopsony is thatthe focus of the TPA is on control of marketing rather than ofdistribution, and it does not necessarily require that the retailoutlets are government run. This distinction is critical.Control over marketing means the TPA can control commu-nication about the products including branding.
The tobacco market would initially consist of most, if notall, of the products that are currently on the market, plus,once the TPA had the capacity to assess new products, arange of potential exposure reduced smoked and smokelessproducts*, plus any nicotine replacement products marketedfor non-cessation purposes. The TPA would have the capacityto withdraw more harmful products from the market asalternatives emerged and/or discourage their use throughhigher prices or other mechanisms. The initial emphasis onproduct modification would be on cigarettes, or their
Figure 2 The regulated market model.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
*In jurisdictions where smokeless products are banned, it may be worthallowing them onto the market to the extent that they are likely to besubstitutes for cigarettes, rather than a largely independent market. Thevalue of doing so would need to be clearly weighed, using theexperience of Sweden as a guide.
Controlling the marketing of tobacco 377
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
alternatives, as they dominate the market. It would move toreduce or eliminate additives and features that mask inherenttoxicity and/or enhance addictiveness. It would be able to setand revise performance and emission standards for toxiccompounds and also be able to introduce new products thatboth promised to reduce exposures and not appeal unac-ceptably to non-users of the more harmful products. Newproducts that in reality performed in unacceptable wayscould be withdrawn from the market or otherwise discour-aged.
The TPA will have a more dynamic and trusting relation-ship with manufacturers because it is in a customer-supplierrelationship, rather than being an independent regulatoryapparatus that stands outside the relationship betweenmarketers and consumers. This simplifies the complexprocess involved in using regulation to progress to lessharmful products, while at the same time maintainingcontrols to prevent unwanted outcomes. In addition to thecapacity to set standards (shared with ordinary regulators),the TPA can directly create incentives to better thosestandards with the offer of increased market share.
Tobacco manufacturers will have incentives to create lessharmful products because the TPA will have reasonablecapacity to make judgements on the harmfulness profile ofproducts, so that it can act like the ideal informed customerwho both wants and recognises less harmful products.Progress to market less harmful products can thus proceedas fast as consumers are prepared to allow. To achieve lessharmful versions of a particular product, the TPA couldspecify maximum levels of undesirable chemicals, and couldgive competitive advantage to those who did even better thanthe specified requirements. Changes in tender requirementscould be introduced in such a manner, and with such notice,as to allow manufacturers and importers the opportunity toadapt their capacities in order to continue to compete. TheTPA would need to take consumer choices into account, andwould have incentives to do so, to prevent it having unsoldstock and to prevent the development of black markets thatwould emerge if its policies were too far out of line withconsumer requirements.
By keeping manufacture in the hands of free enterprisecompanies, the model maintains the capacity for innovationthat is inherent in free enterprise, provided that it can ensurethe operation of competition. Competition could be achievedby having proportional tendering for products done in such away to ensure that several companies were allocated marketshare and thus kept in the market. Market share would beconsidered in terms of both the overall market and of share ofparticular products. Novel products would initially have onlyone supplier. If it looked as if such a product would achieve alarge share of the total market, or where combinations ofinnovations from different manufacturers held out thepromise of even less harmful, but acceptable products,patented innovations could be licensed to other manufac-turers for a negotiated fee or royalty. This would preservecompetition, while preserving incentives for innovation.
Such a system actually creates incentives for manufac-turers to disclose product information to the TPA. The TPAwould be in a position to stipulate that certain, specifiedinformation be supplied about particular products, or thatresearch about such products be provided to it. It is also intheir interests to provide extra information to convince theTPA of the merits of their tender proposals. That informationcould be made publicly available to enhance openness ofoperations. In addition to the information it gets fromtobacco manufacturers, the TPA will also need access toindependent research, plus the capacity to evaluate criticallyall the information. Independent research will be importantnot only to check industry research, but also to monitor the
consequences of decisions the TPA makes: on consumerbehaviour, on exposures to harmful constituents, and, in thelonger term, on observed health consequences. As surveil-lance information comes to hand, policies can be modified tomaximise the harm reduction capacity of the TPA’s activities.
Control of communication and marketing issues isobviously one of the most important parts of any regulatoryscheme. Balanced communication is most likely to comefrom a harm minimising public authority making decisionsand acting under a harm minimising statutory charter. TheTPA would be the main or only organisation with the right tocommercial communication with the public. It wouldexercise controls over what could be said by retailers aboutthe products they sell.
It is extremely difficult to eliminate positive promotion ofproducts when companies are competing to sell to con-sumers. The TPA could and probably would market productsunder its own name. A particular type of cigarettes could besourced from a number of different manufacturers, with theend consumer not knowing which. Where this was done, itwould effectively cut the relationship between manufacturerand consumer, and thus would effectively eliminate anybenefits from manufacturers marketing to consumers. Thusmanufacturers will lack incentives to break or bend lawsrestricting promotion of tobacco products (something thatthey are currently highly motivated to do, as the potentialgains are huge). Such a system also makes it easy to ensuredrab packaging. While the TPA would use its own genericbrands on the more harmful tobacco products (for example,Type 2 cigarettes), less harmful products (if they are everproduced) could be allowed to carry more attractive brands ifthere was any benefit in doing so. The TPA would also specifythe form and content of health warnings and other productinformation, including provision for regular updates asknowledge advances and/or gaps in knowledge are identified.This overcomes the problem of regulators having limitedcapacity to rapidly update warning information because ofthe time taken to introduce or amend regulations.
In summary, the RMM has:
N the innovative capacity of free enterprise to move to lessharmful products with safeguards to prevent marketing ofproducts that appeal to consumers regardless of harm
N capacity to remove incentives to over-promote positives oftobacco use and to inform consumers in a realistic wayabout harms
N capacity to influence price, in particular to prevent pricediscounting being used to grow markets
N capacity to take control over brands, as they are a majormeans of adding extrinsic value to tobacco products (as forother consumer products).
ENSURING THE TPA OPERATES APPROPRIATELYA major area of concern is whether we could trust an agencysuch as the TPA to serve the public interest. The TPA wouldrequire independent and open governance under a charterthat spelled out clearly its objectives. To achieve this, it wouldneed an independent board, with experts in appropriatefields, and sufficient distance from executive government toensure independence. The deliberations, decisions, dealings,and actions of the TPA would need to be as open as possible.This would reduce the risks of regulatory capture and ofcorruption of its purpose by forces within government moreconcerned with revenue than community wellbeing.
A monopsony is likely to be less subject to unbalancedinfluence than a monopoly. By its nature it cannot be assecretive in its operations. Under the proposed system,smaller tobacco companies are likely to proliferate as they
378 Borland
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
can be competitive when they do not require the large sizenecessary for mass marketing. This should result in the bigmultinationals becoming smaller. This should enhancecompetition, but in this instance competition that is moreclosely tied to the community interest. The existence of alarger number of players makes it more difficult for the TPAto do inappropriate deals with some companies withoutothers getting upset and blowing the whistle. Thus, tobaccomanufacturers could become key agents for reducing therisks of regulatory capture; something almost inconceivablewith respect to today’s big tobacco companies. To ensurecompetition, there will need to be rules governing maximummarket share to keep sufficient companies in the market, atleast in jurisdictions big enough to sustain multiple providers.
The other main potential for corruption is from withingovernment. It is not too hard to imagine a government indire financial straits looking at tobacco as a means of raisingrevenue. In the short term, there is a common interest inprice rises for tobacco products as they will reduce demandwhile increasing total revenue to government. However, ifthere is pressure to increase demand, which may occur ifconsumption falls too much, how would it be resisted?Separation of the main revenue related decision making fromthe TPA is an essential part of the solution to maximisescrutiny of decisions taken. The TPA would need to be set upin such a way that the powers that regulate it areadministered by the Health Department (or equivalent) anda different government department (for example, Treasury)collects taxation (or equivalent) revenue. The Treasury wouldset tax rates after consultation with the TPA and ideallywould have the power, on the recommendation of the TPA, toset differential taxes for products of different presumed harmprofiles. The TPA, itself, would probably be created as a non-profit (or non-profit maximising) organization, raising itsoperating costs from the business (although alternatively itcould be given a budget from government revenues). Thisseparates the revenue raising capacity from other functionsdesigned primarily to reduce harm. To bolster further theinterest of the TPA in public health, it could be givenresponsibilities for some cessation and prevention pro-grammes. Staff working in these programmes would bemotivated to resist any efforts to undermine the agency’sintegrity. Taken together, this structure should create abalance that minimises the risk of regulatory capture bymanufacturers or corruption by revenue raising interestswithin governments.
One final guard against corruption could be benchmarkingagainst the performance of TPAs in other jurisdictions. Inlarge countries, the model could be adopted at a state level,with between-state competition driving improved perfor-mance.
External to the TPA is the problem of preventing anysubstantial illicit market (largely due to smuggling). Illicitmarkets develop where there is money to be made. Anysystem that either acts to drive up price relative to production(for example, taxes), and/or mandates changes to cigarettemanufacture that make them less appealing, could makecheaper and/or more appealing illicit products attractive. Ifstates, or countries with permeable borders, acted alone, theywould be restricted in the extent of changes they could makebecause of the difficulty of preventing smuggling if theymoved too far ahead of consumer desires. This is a problemfor any system, not just an RMM. Change will be limitedwhere there is capacity to import product in large quantitiesfrom outside the area controlled by the TPA. However, itwould still be possible to implement the basic apparatus. Thiscan be seen by the lack of problems associated with stateliquor distributors in some US states operating alongsideopen markets in neighbouring states. These work well, but
often do little more than reduce the density of retail outlets.To optimise the potential of the model to move towardsproducing the least attractive and/or least harmful productspossible would require it being adopted within a set ofjurisdictions which could control smuggling.
A final challenge for the form of regulated market arguedfor here is to ensure that retailers do not become the de factomarketers. Small retailers have little capacity to add value orotherwise grow the market. Within their locality they can,but as the distributors will be agents of the TPA, there is somescrutiny to prevent these sorts of practices. Of more concernwould be if one retailer became big enough to make it worthits while to bend or break laws to try to grow the market. Thisrisk can be managed, perhaps by use of anti-monopolieslegislation.
MERITS OF THE REGULATED MARKET MODELIn assessing the relative merits of the RMM, it is useful tocompare it with plausible alternative models. We initiallycanvassed three other possibilities: conventional regulation, amonopoly, and a laissez-faire system relying entirely oninformed consumers. It immediately became apparent thatthe laissez-faire model was not viable; government clearlyneed to be involved in product information (this is acknowl-edged in the FCTC) and there is insufficient consumersovereignty for users to make truly informed decisions. Also,we do not believe that a monopoly is viable; however, it isincluded in the comparisons in table 1, mainly to showreaders that the RMM is fundamentally different to either aconventional regulatory authority or a government mono-poly, even with both set up with the same goals as a RMM.
The analysis summarised in table 1 suggests that while aconventional regulator is as able to mandate less harmfulproducts as an RMM, it is less clear how it could activelyencourage new innovation. Further, an RMM appears toenable effective controls over promotion by removingincentives to promote, something that does not seempossible while manufacturers compete to sell to endconsumers.
Under the RMM, the role of the regulator is made easierbecause most of the work a regulator needs to do under thefree enterprise system is done by the special customer-supplier relationship between the TPA and the manufacturersor their import agents. The role of regulation is thus littlemore than the rules setting up the TPA, prohibiting otherentities engaging in marketing of tobacco products outside ofmechanisms the TPA might permit, and preventing collusionthat could undermine the intent of the system. What wouldnormally be complex sets of rules about what is allowed inparticular products and what marketing activities are and arenot permissible, is simplified to flexible contractual arrange-ments, or is done by an agency without strong incentives todistort communication. For the alternative of a conventionalregulatory approach to work effectively, it will requiresufficient powers and flexibility to do its job properly andresources to support and utilise the necessary expertise. Theintrinsically antagonistic relationship it has with manufac-turers and marketers makes doing this task difficult, and it islikely to be more expensive than doing the same task withinan agency that has a more cooperative relationship withmanufacturers.
For countries with a current monopoly or near monopoly,the RMM has considerable extra attraction as it probablyovercomes World Trade Organization problems with mono-polies, while maintaining government control over what isavailable and how it is promoted. It provides a clearalternative to the privatisations that governments areengaged in or are contemplating in places like Japan andThailand. As Mark Levin notes (personal communication,
Controlling the marketing of tobacco 379
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
Table
1H
owth
ree
poss
ible
syst
ems
deal
with
som
eke
yis
sues
Issu
eC
onve
ntio
nal
regul
atio
nG
over
nmen
tm
onop
oly
Reg
ulate
dm
ark
etm
odel
A.
Prod
uctco
mpo
sitio
nEn
sure
that
prod
ucts
mee
tth
est
anda
rds
that
have
been
esta
blis
hed
Requ
ires
regu
latio
nsth
atsp
ecify
the
stan
dard
s.Th
isis
som
ethi
ngth
atth
eyca
ndo
wel
l,bu
tit
also
requ
ires
mon
itori
ng,
whi
chca
nbe
sque
ezed
intim
esof
budg
etar
ysc
arci
ty.
Typi
cally
can
only
regu
late
wha
tis
,so
limite
din
capa
city
tore
gula
tene
win
nova
tions
.W
illne
edto
best
rong
,fle
xibl
ean
dvi
gila
ntto
wor
k
The
gove
rnm
ent
sim
ply
inst
ruct
sth
epr
oduc
tion
units
asto
wha
tth
eyca
nan
dca
nnot
do.
Ther
eis
none
cess
ary
mec
hani
smof
revi
ew.
The
proc
ess
can
vary
from
the
soun
dly
base
d(g
ood)
toth
eca
pric
ious
(dan
gero
us)
The
TPA
both
spec
ifies
min
imum
stan
dard
sin
tend
ers
for
prod
ucts
and
rew
ards
com
pani
esth
atex
ceed
(bet
ter)
thos
est
anda
rds
with
incr
ease
dm
arke
tsha
re.
The
TPA
will
need
inde
pend
entc
apac
ityto
test
prod
ucts
,bu
tis
likel
yto
getgo
odqu
ality
info
rmat
ion
from
itssu
pplie
rsas
itis
thei
rcu
stom
er.
Ther
eis
ari
ska
mon
opol
istic
dist
ribu
tor
will
mak
epo
orch
oice
s,bu
tin
tern
atio
nalc
ompa
riso
nsh
ould
show
this
upEn
cour
age
the
crea
tion
ofge
nuin
ely
harm
redu
ced
prod
ucts
Has
inno
vativ
epo
tent
ialo
ffr
eeen
terp
rise
,bu
tre
quir
esca
paci
tyto
mak
ecl
aim
sif
itis
toen
cour
age
prod
uctio
nof
prod
ucts
that
bette
rth
em
inim
umst
anda
rds
Mon
opol
ies
typi
cally
are
poor
atin
nova
tion
Can
enco
urag
epr
oduc
tin
nova
tion
tore
duce
pote
ntia
lhar
m,
thro
ugh
capa
city
toof
fer
incr
ease
dm
arke
tsh
are
tosu
chpr
oduc
ts.
Doe
sno
treq
uire
capa
city
tom
ake
harm
redu
ctio
ncl
aim
s.C
ontr
acts
form
am
ore
flexi
ble
base
for
prod
uctin
nova
tion
than
regu
latio
nsA
llow
mov
esto
less
intr
insi
cally
attr
activ
epr
oduc
tsV
ery
diffi
cult
toac
hiev
e.C
ompa
nies
use
tast
ean
dot
her
diffe
renc
esto
com
pete
and
will
stro
ngly
cont
estan
ysu
chch
ange
s
Can
bere
adily
achi
eved
ifth
ew
illis
ther
e,bu
tlim
ited
byco
nsum
erto
lera
nce
Can
beac
hiev
ed,
butlim
ited
byco
nsum
erto
lera
nce
B.C
omm
unic
atio
nan
dm
arke
ting
Ensu
reba
lanc
edco
mm
unic
atio
nab
out
the
bene
fits
and
harm
sof
use
Stro
ngin
cent
ives
for
com
pany
com
mun
icat
ion
toun
ders
tate
harm
s.G
over
nmen
tsha
veta
ken
role
inm
anda
ting
heal
thw
arni
ngs
and
inso
me
case
sfu
ndin
gan
ti-to
bacc
oad
vert
isin
g.G
over
nmen
tsw
illne
edto
play
anex
pand
edro
lein
man
datin
gpr
ovis
ion
ofba
sic
info
rmat
ion
abou
tha
rms
asth
eco
mpa
nies
can
notor
will
notdo
this
.W
hile
itis
prof
itabl
eto
doso
,th
ere
will
bepr
essu
reto
avoi
dor
even
brea
kla
ws
Inpr
inci
ple,
less
prob
lem
with
ince
ntiv
esto
dist
ort
com
mun
icat
ion,
butun
less
the
mon
opol
yha
sa
harm
redu
ctio
nch
arte
r,no
real
ince
ntiv
eto
ensu
reba
lanc
e
Cen
tral
toits
char
ter,
sosh
ould
best
raig
htfo
rwar
d.C
anef
fect
ivel
yre
mov
ein
cent
ives
for
toba
cco
com
pani
esto
prom
ote
toco
nsum
ers,
byse
lling
prod
ucts
unde
rTP
Aliv
ery.
Requ
irem
ents
oftr
ansp
aren
cyan
dpu
blic
invo
lvem
ents
houl
dpr
oduc
eco
nditi
ons
toge
tthe
bala
nce
righ
t
Con
trol
sto
prev
ent
pric
ing
bein
gus
edto
attr
actne
wm
arke
tsPr
ice
cont
rols
wor
k,bu
tth
eyre
quir
eco
nsta
ntvi
gila
nce,
thus
itis
cost
lyto
ensu
reth
eyop
erat
ew
here
ince
ntiv
esfo
rvi
olat
ion
exis
t
Sim
ple
inpr
inci
ple,
butin
cent
ives
tobr
eak
rule
sre
mai
nSh
ould
oper
ate
smoo
thly
.TP
Aha
sno
ince
ntiv
e,bu
tin
cent
ives
for
reta
ilers
rem
ain
asfo
rot
her
syst
ems,
butp
roba
bly
easi
erto
cont
rol
asco
ntra
ctua
lrel
atio
nshi
psw
ithTP
Apr
ovid
esgr
eate
rov
ersi
ght
Trul
yge
neri
cpa
ckag
ing
Lim
ited
asco
mpa
nies
mus
tbe
able
tore
tain
som
eid
entif
iabl
elin
kw
ithth
eir
prod
ucts
,lin
ksth
atca
nbe
used
for
mar
ketin
gPo
ssib
leas
allb
rand
sow
ned
bym
onop
oly
Poss
ible
asal
lbra
nds
can
beow
ned
byTP
A.
Part
ofth
eap
peal
ofth
em
odel
isth
atit
enco
urag
essu
chin
nova
tions
Elim
inat
ion
ofca
paci
tyto
add
valu
eto
toba
cco
prod
ucts
thro
ugh
mod
ern
mar
ketin
gte
chni
ques
Ince
ntiv
esre
mai
nfo
rin
dire
ctpr
omot
ion
prov
idin
git
can
belin
ked
toco
mpa
nypr
oduc
ts.
Proh
ibiti
ngin
dire
ctpr
omot
ion
ofpr
oduc
tsis
extr
emel
ydi
fficu
lt
Shou
ldca
use
nopr
oble
ms,
asco
uld
bepr
ohib
ited
and
even
ifin
cent
ives
rem
aine
d,m
onop
olie
ste
ndto
bele
sscr
eativ
ein
thei
rm
arke
ting
Such
activ
ities
wou
ldbe
inco
mpa
tible
with
the
char
ter
ofth
eTP
A,
and
ther
eis
little
ince
ntiv
efo
rvi
olat
ions
.Th
ere
isal
solit
tlein
cent
ive
for
any
such
prom
otio
nto
cons
umer
sby
man
ufac
ture
rsif
toba
cco
prod
ucts
are
sold
with
TPA
bran
dsC
.Sy
stem
inte
grity
Lega
lint
egri
tyW
illre
quir
em
akin
gin
dust
ryac
tivity
that
mee
tsre
gula
tor
requ
irem
ents
lega
lwhe
reth
ere
curr
ently
may
bequ
estio
nsN
otpe
rmis
sibl
eun
der
Wor
ldTr
ade
Org
aniz
atio
nco
nven
tions
,th
usno
tvi
able
for
WTO
mem
ber
TPA
wou
ldbe
actin
gle
gally
whi
leit
oper
ated
with
inits
char
ter.
Itw
ould
requ
ire
man
ufac
ture
r’s
activ
ityto
bem
ade
lega
lif
they
met
TPA
and
othe
rre
quir
emen
tsM
aint
aini
ngth
ein
tegr
ityof
the
syst
em—
ince
ntiv
esto
com
ply
with
the
spir
itas
wel
las
lette
rof
the
law
Role
ofre
gula
tor
will
beco
mpl
exas
com
pani
esha
vein
cent
ives
toby
pass
regu
lato
ran
dke
epin
form
atio
nse
cret
whe
repo
ssib
le.
Will
requ
ire
stro
ngan
dvi
gila
ntre
gula
tors
with
stro
ngva
lues
ofin
depe
nden
ce.
Stro
ngen
forc
emen
tof
law
sw
illbe
need
edto
ensu
reno
com
mer
cial
lym
otiv
ated
avoi
danc
eor
viol
atio
nof
the
rule
s.H
asso
mew
hatgr
eate
rin
trin
sic
limits
onpo
tent
ialt
rans
pare
ncy
beca
use
ofco
mm
erci
alin
-con
fiden
cere
quir
emen
ts.
As
ther
eis
likel
yto
beon
lya
smal
lnum
ber
ofco
mpa
nies
inth
em
arke
t,re
gula
tory
capt
ure
rem
ains
apr
oble
m
Few
ince
ntiv
esex
istfo
ra
mon
opol
yto
impo
seon
erou
sre
stri
ctio
nson
the
type
sof
prod
ucts
that
they
can
prod
uce.
Ove
rsig
htw
illbe
need
edby
gove
rnm
ent
agen
cyw
ithin
tere
stin
publ
iche
alth
.C
anbe
tota
llytr
ansp
aren
t,bu
tra
rely
dom
onop
olie
sfe
elth
ene
edto
be.
Long
hist
ory
ofm
onop
olie
sbe
ing
corr
upte
dby
desi
refo
rm
oney
and/
orpo
wer
Con
trol
ofpr
oduc
tcom
posi
tion
and
othe
ris
sues
isea
sier
than
inth
eop
enm
arke
tbe
caus
eco
mpa
nies
have
less
ince
ntiv
eto
hide
info
rmat
ion
from
TPA
and
ofte
nha
vein
cent
ives
topr
ovid
eun
solic
ited
info
rmat
ion
asth
eTP
Ais
thei
rcu
stom
er.
Com
petit
ion
inpr
oduc
ing
prod
ucts
for
the
TPA
will
min
imis
eri
skof
regu
lato
ryca
ptur
eby
cart
els.
Sepa
ratio
nof
reve
nue
rais
ing
from
regu
lato
ryov
ersi
ght,
the
latte
rin
heal
thag
ency
hand
s,w
illm
inim
ise
over
emph
asis
ofre
venu
eas
pect
s.Tr
ansp
aren
cysh
ould
beea
sier
than
unde
rre
gula
tion,
and
whi
leit
may
bem
ore
diffi
cult
inpr
inci
ple
than
for
am
onop
oly,
ther
ew
illbe
anu
mbe
rof
affe
cted
part
ies
dem
andi
ngit.
380 Borland
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
July 2003), we should not sit by and let tobacco companies betaken over by interests with no potential interest in publichealth.
The RMM provides improved capacity to minimise theimage creation associated with marketing and capacity tomove to reverse some of the engineering that has maskedinherent toxicity. It should act to reduce demand. The modelcould also allow or encourage retail outlets to promotecessation. While in principle this can be done under thecurrent system, many would be extremely worried about thetobacco industry systematically undermining the effort, whilebeing seen to cooperate. The model can also constrain supply,particularly by restricting more harmful products and, if itwished to do so, by gradually restricting the number of retailoutlets it contracted to sell its products.
BARRIERS TO ADOPTIONThere is little doubt that the initial reaction of most people tothe question of whether an RMM is politically feasible will be‘‘No’’. Part of the reason for this is the distrust of new ideas.It is more useful to think about what might be morepersistent barriers that may remain when the novelty wearsoff.
For an RMM to become a reality, it needs to beimplemented through national laws in ways that areconsistent with international law. The FCTC is creatingunprecedented international focus on tobacco control andforcing all-of-government responses, both of which should behelpful in moving towards the kind of comprehensivesolution the model represents. However, this type of controlover the tobacco market was not considered in the negotia-tion of the convention, because the ideas are too new. It willbe important to ensure that the absence from the conventionof models like the RMM is not used as an excuse to avoidconsidering new ideas. Indeed, the very fact of the FCTCdemonstrates the need for concerted international action todo all that is practical to reduce the death and disease tobaccouse causes.
The other relevant international law is that governing theWorld Trade Organization (particularly the TRIPS agreementwhich deals with intellectual property and trade marks).There appears to be no basic prohibition of monopsonies aslong as they do not discriminate between national and non-national suppliers in their purchasing of product. The goal oftruly generic packaging could run into problem withagreements like TRIPS, particularly if it were to be donewithout compensation, but this is not clear. More work isneeded here, but it needs to be remembered that tobaccorelated trademarks have monetary value because they addconsumer value to these inherently harmful products.Removal of trademarks may be more a problem for somecountries than others. In some, it may be enough to maketobacco a controlled substance,21 moving it outside ordinarycommerce. Such legislation would only allow consumerpossession of tobacco products marketed by the TPA.Alternative pathways include forfeiture of trademarks eithervoluntarily, as part of protection against future liability forproducing harmful products, or forcefully through successfullitigation against such practices.
Another major potential barrier is that governments maysee a risk of becoming associated with the residual tobaccoproblem. This may cause problems as the community comesto forget the benefits that change has provided. In setting upthe system, it will be important that there are symbolic aswell as practical aspects to the legislation defining tobaccouse as socially problematic. The symbolism of making tobaccoa controlled substance and constraining its availabilityshould help to send signals of governmental disapproval,
minimising the risk of it being seen to promote somethingthat is socially undesirable.
A final barrier is to convince people that the problem is bigenough to warrant such a solution. There is no doubt thatvested interests who profit from the current system willoppose the adoption of an RMM and they may gain supportfrom those who are naturally suspicious of governmentinvolvement in either commerce and/or in influencingpersonal choices. For countries like Australia, with strongcommitments to free markets, there are challenges ofrethinking how the model actually fits within a system thatis designed to encourage competition and innovation.However, there are currently major challenges in justifyingthe preservation of a system that grows demand for productssociety has decided are harmful.
To get decision makers to focus on the importance of suchcomprehensive change may require a crisis. This could comefrom the courts. It is possible that successful litigation, eithercivil or criminal,8 could effectively outlaw the current systemof for-profit companies marketing tobacco products toconsumers. In such a context, governments would be forcedto act quickly. An RMM could be an attractive solution.
CONCLUSIONSThe idea for an RMM arose as a response to the proliferationof new potentially harm reduced tobacco products. It wasdesigned to create a context whereby the forces of competi-tion could be marshalled in the interests of reducing theharmfulness of tobacco products. It is also an extremelyeffective system for minimising commercial incentives to usetobacco products, and could be used solely for that purpose.The RMM may be the most practical way of ensuring theelimination of inappropriate promotion. These are potentreasons to consider its adoption.
It is important to realise that the version of the RMMarticulated here is not the entire solution. There will alsoneed to be controls over retailing, ongoing programmes forprevention and cessation, and more extensive controls overwhen and where tobacco products are used. At least some ofthese tasks will be made easier when the powerful forces inthe tobacco industry that are systematically underminingthese efforts are removed. An RMM may mean we will haveto spend less in these other areas to achieve the same result.
Tobacco control needs strategies to minimise forces thatencourage the uptake of inherently harmful products andcapacity to allow smokers who are currently unwilling orunable to quit to switch to less harmful variants, withoutdiscouraging their eventual cessation, or of encouraging newusers. The RMM appears to achieve these goals. We are notaware of any alternative proposal that is likely to be asefficient or effective. Tobacco use causes far too much deathand disease for any but the most effective strategy to becredible. Research and analysis is needed to test the claimsmade in this paper. Governments should seriously considerits adoption, or at the very least add it to the mix of strategiesthey consider when they take the (what seems to us asinevitable) steps towards more comprehensive control of allaspects of the tobacco market.
ACKNOWLEDGEMENTSThe project on which this paper is based is funded by the VictorianHealth Promotion Foundation (VicHealth) and The Cancer CouncilVictoria. The core of the ideas within it come from work done withJonathan Liberman, Michelle Scollo, and Kathy Barnsley. These ideashave been refined by discussions and correspondence with manyindividuals who have contributed comments to presentation of theideas developed in the paper. Also special thanks to JonathanLiberman for thoughtful comments on this manuscript.
Controlling the marketing of tobacco 381
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
REFERENCES1 US Department of Health and Human Services. Reducing the health
consequences of smoking: 25 years of progress. A report of the SurgeonGeneral, 1989. Rockville, Maryland: Public Health Service, Centers forDisease Control, Office on Smoking and Health, 1989. (DHHS Publication No(CDC) 89-8411.)
2 UK Department of Health. Report of the Scientific Committee on Tobacco andHealth. UK Department of Health, 1998.
3 Peto R. Smoking and death: The past 40 years and the next 40. BMJ1994;309:937–9.
4 Ramstrom LM. Snuff—An alternative nicotine delivery system. In: Ferrence R,Slade J, Room R, Pope M, eds. Nicotine and public health. Washington DC:American Public Health Association, 2000.
5 WHO link to FCTC documents http://www.who.int/gb/EB_WHA/PDF/WHA56/ea56r1.pdf
6 Liberman J, Clough J. Corporations that kill: the criminal liability of tobaccomanufacturers. Criminal Law Journal August 2002, 26 (4), Lawbook Co.,Sydney.
7 Liberman J. Where to for tobacco regulation: Time for new approaches? Drugand Alcohol Review (in press).
8 Stratton K, Padma S, Wallace R. Clearing the smoke: assessing the sciencebase for tobacco harm reduction. Washington: Institute of Medicine, 2001.
9 Kozlowski LT, O’Connor RJ. Cigarette filter ventilation is a defective designbecause of misleading taste, bigger puffs, and blocked vents. Tobacco Control2002;11(suppl I):i40–50.
10 Slade J, Henningfield J. Tobacco product regulation: context and issues.Tobacco dependence: innovative regulatory approaches to reduce death anddisease. Food and Drug Law Journal 1998;53(suppl):44–76.
11 World Health Organization. Advancing knowledge of regulating tobaccoproducts. Tobacco Control 2000;9:224–6.
12 Borland R, Scollo M, Farquharson K. Towards a framework for effectivetobacco product regulation. Paper presented to SACTob, Olso, February2002.
13 Royal College of Physicians. Nicotine addiction in Great Britain. London:Royal College of Physicians, 2000.
14 Corrao MA, Guidon GE, Sharma N, Shokoohi DK, eds. Tobaccocontrol country profiles. Atlanta, Georgia: American Cancer Society,2000.
15 Hill DJ, White VM, Scollo MM. Smoking behaviours of Australian adults in1995: trends and concerns. Med J Aust 1998;168:209–13.
16 Jha P, Chaloupka FJ. Curbing the epidemic: governments and the economicsof tobacco control. Washington: World Bank, 1999.
17 US Department of Health and Human Services. Preventing tobacco useamong young people. A report of the Surgeon General, 1994. Atlanta,Georgia: Public Health Service, Centers for Disease Control and Prevention,Office on Smoking and Health, 1994. (US Government Printing OfficePublication No S/N 017-001-00491-0.)
18 Waxman HK. Tobacco industry statements in the US Department of JusticeLawsuit. Tobacco Control 2003;12:94–101.
19 http://www.altria.com/about_altria/01_00_03_philipmorrisusa.asp.20 Gray N, Kozlowski LT. More on the regulation of tobacco smoke: how we got
here and where next. Ann Oncol 2003;14:353–7.21 Sweanor DT. Regulatory mechanisms for nicotine delivering products:
marketing and competition issues. In: Ferrence R, Slade J, Room R, Pope M,eds. Nicotine and public health. Washington: American Public HealthAssociation, 2000.
The lighter side ...................................................................................
EI’m Going Out for a Smoke. By Walt Handelsman. Copyright 2003 Tribune Media Services, Inc. All RightsReserved. Reprinted with permission.
382 Borland
www.tobaccocontrol.com
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from
tobacco products: a regulated market modelA strategy for controlling the marketing of
R Borland
doi: 10.1136/tc.12.4.3742003 12: 374-382 Tob Control
http://tobaccocontrol.bmj.com/content/12/4/374Updated information and services can be found at:
These include:
References #BIBLhttp://tobaccocontrol.bmj.com/content/12/4/374
This article cites 7 articles, 5 of which you can access for free at:
serviceEmail alerting
box at the top right corner of the online article. Receive free email alerts when new articles cite this article. Sign up in the
Notes
http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:
http://journals.bmj.com/cgi/reprintformTo order reprints go to:
http://group.bmj.com/subscribe/To subscribe to BMJ go to:
group.bmj.com on June 19, 2016 - Published by http://tobaccocontrol.bmj.com/Downloaded from