Countering the demand for, and supply of, illicit tobacco: an assessment of the 'North of England Tackling Illicit Tobacco for Better Health' Programme.

BACKGROUND: Illicit tobacco (IT) undermines the effectiveness of tobacco control strategies. We assessed the implementation and impact of a new programme designed to reduce demand for, as well as supply of, IT, in the north of England, where IT was prevalent.
METHODS: 'Mixed methods' research was undertaken. Qualitative methods included stakeholder interviews (at outset and 1 year later) and ethnographic research. Indicators reflecting those supply and demand issues for which data were available were identified and monitored, including relevant items on two cross-sectional surveys carried out in 2009 and 2011 with over 4000 individuals from which a social marketing campaign was also developed. IT reports to two existing hotlines, promoted through the programme, were assessed.
RESULTS: Initially, concerns abounded about the different philosophies and ways of working of local and national enforcement and health agencies, but these were much reduced at follow-up. A protocol was developed which greatly facilitated the flow of intelligence about IT supply. A social marketing campaign was developed highlighting two messages: IT makes it easier for children to start smoking and brings crime into the community, thereby avoiding misleading messages about relative harms of illicit and licit tobacco. Public and stakeholder awareness of IT increased as did calls to both hotlines.
CONCLUSIONS: A partnership of agencies, with competing values, was established to tackle IT, a complex public health issue and, inter alia, implemented a social marketing campaign using novel messages. This improved the flow of intelligence about the supply of IT and increased awareness of IT. Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions.

Introduction

Illicit tobacco (IT, a term we use to include smuggled, bootlegged and counterfeit tobacco,1) undermines the effectiveness of tobacco control strategies by maintaining and encouraging tobacco use. IT is an international problem accounting for 11.6% of cigarette trade at a cost to governments of over US$40 billion in lost revenue.2 In the UK, the availability of tobacco products, particularly among deprived communities, at sometimes less than half the price of legal tobacco weakens the economic disincentive for tobacco use that paid duty provides.3 IT products frequently lack adequate health warnings, increase the potential for young people to start smoking through unregulated sales, and contribute to crime and criminalisation at local, regional, national and international levels. Those involved in the IT market are often involved in other forms of illegal activity as well, with tobacco offering what is often a particularly high profit, low-risk activity; in some cases, tobacco smuggling has funded terrorist activity.4 5 The role of the tobacco multinationals in the illicit market both internationally and in the UK is also well documented.6 7 Like any other consumer market, IT is driven by both supply and demand factors.

The illegality of IT makes it difficult to measure. In the UK, Custom's officers (Her Majesty's Revenue and Customs (HMRC)) assess IT use by comparing tax receipts from legitimate sales with smoking prevalence figures collected by national surveys which assess smoking prevalence and consumption. In 2000, with illicit cigarettes constituting an estimated 20% of the market and hand-rolled tobacco 63%, the UK government introduced a new comprehensive strategy Tackling Tobacco Smuggling,8 which was updated and expanded in 2006,9 20091 and 2011.10 Legally binding agreements to reduce IT are also now in place between the UK Government and tobacco companies.11 Significant decreases in the volume of IT have been measured by HMRC since 200410; latest figures indicate the illicit share to be approximately 9% of cigarettes and 38% of hand-rolling tobacco, costing the exchequer around £2 billion per annum.12 The IT market is complex, dynamic and rapidly evolving. Seizures in the UK have changed over the last decade from predominantly smuggled legally manufactured tobacco to counterfeit and ‘illicit whites’ (cigarette brands produced by smaller, overseas manufacturers purely for smuggling to the UK and elsewhere), although seizures are not necessarily representative of the IT market as a whole.

Tax revenue data collected nationally masks major geographic variations in IT consumption, and the north of England was perceived informally as something of an IT ‘hotspot’. In December 2007, a summit meeting acknowledged the need for better understanding of IT use and insights into how to tackle demand for, and supply of, the evolving IT market in the region. The ‘North of England Tackling IT for Better Health’ programme (hereafter referred to as the Programme) was a response to this need.13 This pilot Programme, launched officially in 2009, received pump-priming funding of £1 million from the English Department of Health (DH), divided equally between the three DH regions involved: the northeast (NE, 2.5 m people), northwest (NW, 6.8 m) and Yorkshire and Humber (Y&H, 5.1 m). If the pilot appeared effective, additional funding was to be sought from elsewhere to enhance or continue the work. The main aim was to increase the health of the population by reducing smoking prevalence through (a) reducing the supply and availability of IT, thus keeping tobacco prices high, and (b) reducing the demand for IT by building on existing tobacco control measures such as media campaigns (figure 1). A multiagency partnership was established, which brought together health, national and local enforcement agencies (including HMRC) and marketing representatives to combat the IT trade. Other key Programme activities over time are listed in table 1. Hitherto, IT had primarily been the policy responsibility of HMRC with a focus on supply, therefore, this Programme represented the first large-scale attempt in the UK (and to our knowledge, internationally) to reduce both IT supply and demand. We were commissioned in 2009 to carry out formative, process and outcome evaluation where feasible, of the Programme, up to September 2011.

Figure 1

Logic model for the North of England Programme.

Table 1

Timeline of Programme and assessment activities and contextual factors

Activity	2007	2008	2009	2010	2011
Programme activity
Summit on Illicit tobacco (IT) in north of England	X Dec)
Developing logic plan and activities		X	X
Setting up management infrastructure: Dedicated programme manager Governance Board (regional governmental health leads, national enforcement (customs and excise), regional enforcement and border agency representatives), local enforcement, marketing representatives) 3 Regional steering groups and local partnerships			XXX
Consumer research			X (July)		X (Mar/Apr) (NW/NE) only due to regional funding stoppage)
Social marketing campaign & associated websites:http://www.get-some-answers.co.uk and http://www.keep-it-out.co.uk				X (June/July)NW/NE only
Establishing main website aimed at professionalshttp://www.illicittobacconorth.org				X
Establishing regional level enforcement posts			X
Agreeing 2 hotlines: Crimestoppers and Customs Hotline				X
Establishing a protocol for sharing intelligence on the supply of IT among local and national enforcement bodies			X	X	X
Development and implementation of tailored training packages for health professionals, magistrates and prosecutors					X
Assessment activity
Ethnographic research			X	X	X
Stakeholder interviews I			X (Nov/Dec)
Stakeholder interviews II				X (Nov/Dec)
Contextual factors
Media ‘freeze’ (ie, government funds could not be used on media campaigns)				X
General election				X (May)
Recession			X	X	X
Funding stopped for Department of Health Regional Offices (Y&H activity stopped)					X (Mar)
Restructuring of local health service, local government, HMRC etc				X	X

HMRC, Her Majesty's Revenue and Customs; NW, northwest; NE, northeast.

Methods

‘Mixed methods’ research was undertaken (table 1). Financial constraints precluded new quantitative research beyond that commissioned by the Programme; qualitative methods included stakeholder interviews and ethnographic research. The Programme could not be expected to have an appreciable impact on prevalence during the assessment period, so indicators reflecting those supply and demand issues for which data were available were identified and monitored, building on a logic model developed by Programme staff (figure 1).

Ethnographic research

Collaborative ethnographic research14–16 was carried out to elucidate gaps in understanding and to contextualise better what was happening. Methods used included participant observation at meetings, events and Programme activities, approximately 50 informal interviews conducted on an opportunistic basis with stakeholders and members of communities in disadvantaged areas targeted by the Programme, analysis of documents prepared for and by Programme stakeholders, and two in-depth locality studies (the researcher spent 1 week in each area). Analysis involved reviewing the information contained in each of the sources and coding by themes which were cross-checked with other data sources to identify important metathemes. Findings from the ethnographic research were fed back to the Programme, provided valuable context, and informed the design of the stakeholder and consumer research described below.

Stakeholder interviews

Semistructured interviews were conducted with key stakeholders at the outset (November/December 2009) and again 1 year later. The focus was on the Programme Governance Board (GB) comprising representatives from local and national enforcement agencies, regional health agencies, and marketing and communication professionals. Sixteen GB members were interviewed at baseline. At follow-up, nine of these members representing the different agencies that comprised the board were randomly chosen and interviewed, together with five individuals external to the Programme selected purposively from regional and national health and local government agencies and two non-governmental organisations working in the UK. Interviews were conducted either face-to-face at participants’ place of work or via telephone, and explored stakeholders’ perceptions, expectations and understanding of the Programme and, additionally at follow-up, thoughts on the Programme's progress and views on its future. Interviews lasted approximately 45 min, were digitally audio-recorded, then transcribed verbatim and analysed using the framework approach.17 Important core themes based on a priori and emergent issues were determined and applied to all transcripts for further development and refinement after which they were grouped into a smaller number of main themes and placed in a framework. Analysis was primarily conducted by one researcher, and second coded by an additional researcher. Illustrative quotes are presented; to maintain anonymity, only the professional background is provided.

Consumer research

The Programme commissioned a market research company (NEMS18) to conduct consumer research to understand the market in relation to the sale of IT and its purchase, and to inform the development of messages which could be used to deter demand for IT. Two independent cross-sectional surveys were carried out, in July 2009 and March/April 2011, of 6084 and 4111 people, respectively; the first was across the three regions, but the second was confined to the NW and NE (see below), and hence, only data from the NW and NE regions are presented here (n=4105, NW/NE 2009). Stratified (by small geographical areas) telephone number sampling from published lists and random generation was used, together with street interviews (using quota sampling), to recruit a minimum number of participants aged 16 years and over from each area. A representative sample of around 2300 people was attained and this was then supplemented with a boosted sample of around 1800 smokers; both data collection methods were used for representative and booster smoker samples. The surveys captured prevalence of IT use and consumption, source/type of tobacco, attitudes to illicit supply and suggested channels and approaches for reporting IT offences; Z tests were used to assess differences between 2009 and 2011.

Telephone hotline data

As discussed below, two existing telephone hotlines were used for consumers to relay IT intelligence confidentially: the HMRC's Customs Hotline (CH), and Crimestoppers. The CH was designed to be used by the public to share information about tax evasion issues (including the distribution and sale of IT). Crimestoppers is a free-to-use, independent charity hotline for the UK public to report criminal activity. HMRC provided CH data from 2008 to 2011, and Crimestoppers provided data from 2009 to 2011.

Results

The main five themes from the stakeholder research are described below alongside the quantitative findings and other contextual information from the ethnographic research. From the second phase of the stakeholder interviews, the external individuals shared similar levels of knowledge about the Programme, in that it involved bringing together and improving engagement of local and national agencies to reduce smoking prevalence by addressing IT, and they were also largely aware of the key components of the Programme. Two external individuals were attempting to mirror a similar approach in their respective regions.

Context

Implementation of the Programme coincided with several events which could have affected its impact (table 1). These include the onset of economic recession in the UK and a rise in unemployment rates across the region, the likely effect of which would have been to draw tobacco users towards cheaper illicit sources. A media spending freeze which precluded government funding being spent on smoking media campaigns meant that only two out of the three regions (NW and NE, who received funding from sources other than the DH) were able to implement the Programme's marketing campaign. Additionally, in May 2010, a change of government resulted in the regional DH tobacco managers who led the Programme having their funding stopped from March 2011 which prevented Y&H carrying out the second phase of the consumer research (see below). There was also major restructuring of all the agencies involved in the Programme causing considerable uncertainly and disruption which were routinely commented on during interviews.

Partnership working

Stakeholders considered the Programme exemplary in how it drew together a variety of national and local agencies around a common goal: “…So many people were brought together from both the health side of tobacco control and the enforcement side, for the first time... that joining up has been a really useful thing to see and we've looked at that and we've started to try and emulate that” [External non-government-organisation, wave 2]

During the early stages of the Programme, there were concerns about the different philosophies of the agencies and a lack of trust between some stakeholders with intelligence sharing a key barrier to progress: “I think the barriers to enforcement are about the barriers that exist between organisations, lack of trust between organisations, different structures of organisations, different cultures of organisations” [Health, wave 1]. The police were a notable omission from the partnership at an early stage. However, follow-up interviews showed perseverance at working together and few negative comments on partnership working although, for a minority, progress still fell short of initial expectations.

Intelligence sharing

National and local partners worked to improve the way intelligence on IT sales could be shared from the outset. All agreed that there would be little point in raising the profile of IT locally and increasing the likelihood of people giving information on local sellers/users without a secure means of doing so, and mechanisms in place to ensure the information supplied was subsequently seen to be acted upon. Additionally, there was the need to streamline operations to prevent more than one team dealing with the same situation: “My biggest worry in all of this is what we term in the trade... ‘blue on blue’. What we are going to end up with if we are not very careful is {local enforcement} take such a route on some intelligence they may have and customs going down exactly the same route and us colliding somewhere in the middle. And it tends to look terribly unprofessional.” [National enforcement agency, wave 1].

A ‘Closer Working’ protocol was developed to enable intelligence sharing between national and local enforcement agencies. Establishing this was time consuming, due to the need for multiorganisational agreement, although stakeholders reported improvements during the process; it was finally signed in June 2011. Programme partners were highly influential in bringing the negotiations to resolution.

Many stakeholders felt that the use of CH to report IT supplies could be problematic, particularly as the social marketing campaign (see below) aimed to encourage the public to view IT as more than a revenue issue. Additionally, there was no mechanism by which local enforcement agencies could routinely access CH information and HMRC expressed concerns about transferring sensitive information to local enforcement officers, many of whom did not have secure email or encrypted systems of communication. For these reasons, Crimestoppers was also promoted as a means of passing on IT intelligence.

The 2009 survey found an overall smoking prevalence of 24%, slightly higher than the 23% (NW) and 22% (NE) in the nationwide General Lifestyle Survey (GLS,19) (table 2). Those who bought IT represented all demographic groups, but included a high proportion of males from lower socioeconomic groups, those struggling financially and younger smokers (aged 16–24 years). Price and convenience of supply were the main motivators for purchase with roughly half the market made up of opportunistic (unplanned) purchases. IT purchasers smoked two cigarettes per day more on average than smokers not purchasing IT. Awareness, attitudes and purchasing behaviour are shown in table 2. Sellers constituted 0.4% of the sample and tended to be young, unskilled or unemployed males. The most common sources for IT purchases were friends (62%) and family (16%), followed by pubs and clubs (14%), shops (6%), private homes (5%) and street hawkers (4%).

Table 2

Illicit tobacco (IT) awareness, purchase and attitudes in the consumer research, phases 1 and 2

	July 2009 (n=4105) (%)	Mar/Apr 2011 (n=4111) (%)	Z test
Awareness of IT (only asked of non-smokers; base non-smokers n=1661, 2029)	54	69	p<0.01
Purchasers of IT (only asked of current smokers; base current smokers n=2444, 2082)	20	18	ns
IT proportion of total consumption (base IT purchasers, n=477, 381)	36	33	ns
Purchase IT from shops (base IT purchasers, n=477, 381)	6	14	p<0.01
IT composition (base IT purchasers, n=477, 381):
(1) Foreign brands not normally found in UK	22	35	p<0.01
(2) Counterfeit/fake	25	26	ns
(3) UK brands	53	39	p<0.01
(IT products) don't do anyone any harm (% disagreeing)	80	82	ns
Buying IT is no ‘big deal’ (% disagreeing)	65	68	p<0.05
IT is a danger to children because they can buy them easily and cheaply (% agreeing)	87	88	ns
Billions of pounds a year of tax which could be spent on schools and hospitals is lost (% agreeing)	75	75	ns
IT brings crime into the local community (% agreeing)	63	64	ns
IT makes it possible for me to smoke when I couldn't afford it otherwise (base IT purchasers, n=477, 381)	63	67	ns
Everyone does it (buy IT) (base IT purchasers, n=477, 381)	69	56	p<0.01
Likely to report someone if you suspected them of selling IT	26	29	p<0.05
Likely to report someone if you suspected them of selling IT to children	75	75	ns

Note: Base total sample unless otherwise indicated.

ns, non-significant.

In the 2011 survey, smoking prevalence was 22% similar to the 21% (NE) and 22% (NW) in the GLS. Awareness of IT increased from 54% to 69% among non-smokers. There was a non-significant reduction in the total market share of IT as the proportion of smokers reporting to purchase IT fell from 20% to 18%, and among these smokers the proportion of IT cigarettes purchased fell from 36% to 33%. Sources of illicit supply were very similar in the two surveys, with the exception of shops, which increased from 6% to 14% of total. There were some small attitudinal and purchasing shifts observed. The proportion of people stating they were likely to report someone selling IT increased by around 3 percentage points to 29%. As in 2009, 75% stated that they would report sales made to children.

Social marketing campaign

Although IT had not previously been the focus of social marketing campaigns, a point commonly made during any media coverage of the issue previously had been that IT was more dangerous and of lower quality than legal tobacco. Programme staff felt this risked legitimising legal tobacco and the tobacco industry, and hence, undermining the de-normalisation agenda. A marketing campaign that used other messages to deter demand for IT, based on the consumer research was, therefore, developed and pretested. The campaign was aimed at people who were already slightly uncomfortable with IT and, under the logos ‘illegal tobacco’ and ‘keep it out’, proposed that people should ‘get some answers’ to key questions about IT. Its two main messages were that IT made it easier for children to start smoking, and that IT brought crime into the community. The campaign ran in June/July 2010 across a range of media, including websites, and all materials were accompanied by information to stakeholders.

Calls to hotlines

Calls to the two hotlines increased during the assessment period (table 3). The greatest increase was observed in calls to Crimestoppers in the NW and NE over the year when the social marketing campaign was run, and calls increased from 100 to 328.

Table 3

Calls to Crimestoppers (CS) and Customs Hotline (CH) concerning Illicit tobacco (IT)

Region	June 2008–March 2009		April 2009–March 2010		April 2010–March 2011*		April 2011–Sept 2011		% increase Apr 09/Mar 10–Apr 10/Mar 11
	CS	CH	CS	CH	CS	CH	CS	CH	CS	CH
Northwest (NW)	75	NA	84	175	229	279	94	141	273	159
Northeast (NE)	13	NA	16	166	99	178	9	99	619	107
NW & NE	88	NA	100	341	328	457	103	240	328	134
Yorkshire & Humber	21	NA	33	212	19	242	44	135	58	114
TOTAL	109	NA	133	553	347	699	147	375	261	126

*Social marketing campaign ran in June and July 2010 in the NW and NE regions only.

NA, data not available.

Numbers in italics are subtotals.

Awareness raising

Towards the end of the evaluation, stakeholders, health and community workers unanimously agreed that the Programme had raised awareness about the scope and significance of IT in their communities. However, there was, occasionally, some uncertainty about what to do with this raised awareness. Some frontline healthcare professionals were worried about jeopardising the fragile trust they worked hard to establish with smokers by discussing IT use with their clients. Others simply felt that the greater harms message would be easy to use, questioning how effective criminality messages would be; some partners continued to use the greater harms message.

Reflections on the success of the Programme

Although stakeholders recognised that any changes in smoking prevalence as a result of the Programme would take time, they indicated that one key objective, bringing together national and local agencies, had been achieved: “I've been impressed with the fact that we've been able to achieve some of the objectives that we've set out…The partnership, the work around partnerships for one thing. Three regions working together very closely in the North of England as well as the regional partnerships that have been developed and also the local partnerships that have occurred within the local alliance, local tobacco control alliances. I think that's probably the key achievement that we've had, is the partnership working.” [Health, wave 2]. Operationally, while a little slow to start, procedures had improved across the region due to better coordination and greater understanding of the other agencies; this resulted in more intelligence sharing about IT in their localities. Additionally, external stakeholders mentioned that data (eg, from consumer research) were being used to inform the development of similar initiatives elsewhere. Finally, the Programme had enabled those partners outside of health to recognise the importance of their roles in impacting public health.

Sustainability

The overriding concern was whether, with the challenging economic climate, the necessary material and human resources would continue to be available to maintain the Programme. It seemed likely that partnership working would continue at least informally, but there was a belief that continued focus was needed on IT to reach a tipping-point. Others were concerned whether this work could, or should, be prioritised over other tobacco control work in tight economic circumstances. There was also a perception that the tobacco industry could over-hype IT to undermine other tobacco control strategies: “…the key thing for me is this should not be seen as becoming a magic wand... This has to be seen as being part of a much broader de-normalisation programme and I think the risk is some people are just catching onto this and saying this is the way we can solve youth smoking. No, it's not…, there are very multiple factors how to influence youth smoking and we must not allow the industry to almost cosy up to this as in they make a fuss about illicit tobacco but I'm convinced that they're part of the problem and we need to be just very mindful to keep saying all tobacco is bad.” [Health, wave 2]

Discussion

This Programme focused simultaneously on both supply and demand for IT and involved a novel partnership between health, marketing and enforcement agencies in the north of England. The increased call volume to IT hotlines reflected the increasing acceptability of reporting by the public, also evident from the consumer research. Additionally, our findings suggest that the Programme was effective in raising awareness of IT. Implementation occurred during a period of economic recession and rapidly shrinking infrastructure in the health and enforcement sectors which would normally be expected to increase IT use, yet this did not occur. Non-significant decreases in self-reported IT purchase and consumption were found instead. The partnership was perceived to be effective and likely to be sustained, at least informally, in the absence of further resources.

Our study has a number of limitations. First, the study was observational in nature and without a control area, we cannot say with certainty whether the effects observed were directly causal in nature. Additionally, HMRC calculations on IT usage could not be disaggregated to regional level, and other indicators had to be found, but these were limited. CH data refer only to intelligence where HMRC received a correct postcode and, therefore, do not include non-geographic and frontier-related intelligence; the data does not relate to results, only allegations received, and may also contain some duplicates where HMRC received multiple pieces of intelligence on one subject. Similarly, with Crimestoppers data, only partial postcode data were supplied, so data have been categorised into geographical areas based on their size. Second, a small minority of calls to Crimestoppers made reference to several addresses, all of which were mapped, so the total exceeds the number of reports made. The confidential nature of the calls precluded us from assessing the quality of the calls received by either helpline. Finally, interviews of stakeholders involved in the Programme or its application elsewhere may have introduced bias although the pilot nature of the work meant that stakeholders were encouraged to give honest feedback.

To our knowledge, this is the first assessment, globally, of any programme developed to reduce the demand as well as the supply of IT. Indeed, the main preoccupation of IT work in the past has been on supply, but the large numbers of buyers (and small sellers) identified across the sociodemographic spectrum highlights the limited likely impact of approaches focusing purely on supply. Misperceptions about IT were also observed which militated against an appreciation of the wider harms of tobacco smuggling arising from its links to organised crime, irrespective of the financial and health costs. The resultant Get Some Answers campaign, was based on evidence collected by the Programme that the messages that would influence demand were those focussing on the harm IT causes children and the criminality that IT brought into communities. Awareness and the volume of calls concerning supply increased, although only small attitudinal shifts were observed, mostly in the preferred direction. The one channel of IT supply that reportedly increased, namely shops, could be a consequence of some of the imagery about street sellers shown in the campaign. The sale of IT through shops is of concern, but retailers in England can currently be fined for doing this. However, there is no licensing system for retailers selling tobacco in England making enforcement difficult. We believe that if retailers needed a licence to sell tobacco, and could have this revoked if caught selling IT, this would be a much greater deterrent.

Measuring and tackling IT is difficult, due to its connections with organised criminal activities. The rapidly changing contraband market emphasises the need for greater intelligence sharing between agencies, an important part of tackling the supply side of IT. Communities needed to be made aware of IT illegality and encouraged to report such activity to law enforcement agencies. However, it was challenging to develop a user-friendly reporting system where the provider of the information could be appropriately anonymised while enabling local geographical information to be obtained. There were further challenges in developing a protocol for securely sharing this intelligence between partner organisations. The successful development of the intelligence-sharing protocol between national and local enforcement agencies would have been unlikely without the high-level partnerships developed through the Programme. This, we suggest, happened because the mismatch between views of the agencies involved were resolved at a level whereby values capable of over-riding such differences could be agreed upon. The partnership approach to tackling IT was, therefore, vital in developing an appropriate pathway for sharing and reporting intelligence. The importance of such resolution is a message relevant to a wide range of problems in public health and policy more generally, when specific issues, such as IT, bring to the table stakeholder organisations with differing and sometimes competing values and priorities. Tensions were also evident in the development of messages beyond those of IT causing greater harms which risked legitimising licit tobacco; the new messages about criminality were not adopted universally.

Sustainability was a concern given the recession, widespread restructuring of posts in the involved sectors, and lack of resources. While overall, the Programme had a large budget, aspects of its work were relatively cheap, such as the development of the partnership, which would therefore be applicable internationally to both high and low-resource countries. Additionally, given tobacco industry hype and focus on IT to counter tobacco control strategies, such as removal of point-of-sale displays and plain packaging, and industry complicity with IT, programmes which aim to understand the nature and size of the IT market and how to reduce it may have increasing importance. Nevertheless, a concern raised was that resources should not be devoted disproportionately to this issue either in response to tobacco industry hype and/or at the expense of other proven effective tobacco control strategies. The importance of tackling IT was highlighted recently by the ratification of the World Health Organisation's Framework Convention on Tobacco Control's IT Protocol20 and a cross-European study.21

Illicit tobacco (IT) is common in some countries, undermines tobacco control strategies, and is the subject of tobacco industry misinformation. There has been little focus on reducing IT to date, and most attention has been on supply. Using mixed-methods research, we report here a Programme designed to reduce both the demand for, as well as supply of, IT. This involved establishing a partnership of disparate agencies, consumer research, a protocol for sharing intelligence on IT across the agencies and a social marketing campaign using novel messages. Our findings indicate that the partnership was successful in improving awareness of IT, increasing the flow of intelligence on the supply of IT and developing a social marketing campaign which avoided misleading messages about the relative risks of illicit versus licit tobacco.