Understanding the emergence of the tobacco industry's use of the term tobacco harm reduction in order to inform public health policy.

OBJECTIVES: To explore the history of transnational tobacco companies' use of the term, approach to and perceived benefits of 'harm reduction'.
METHODS: Analysis of internal tobacco industry documents, contemporary tobacco industry literature and 6 semistructured interviews.
RESULTS: The 2001 Institute of Medicine report on tobacco harm reduction appears to have been pivotal in shaping industry discourse. Documents suggest British American Tobacco and Philip Morris International adopted the term 'harm reduction' from Institute of Medicine, then proceeded to heavily emphasise the term in their corporate messaging. Documents and interviews suggest harm reduction offered the tobacco industry two main benefits: an opportunity to (re-) establish dialogue with and access to policy makers, scientists and public health groups and to secure reputational benefits via an emerging corporate social responsibility agenda.
CONCLUSIONS: Transnational tobacco companies' harm reduction discourse should be seen as opportunistic tactical adaptation to policy change rather than a genuine commitment to harm reduction. Care should be taken that this does not undermine gains hitherto secured in efforts to reduce the ability of the tobacco industry to inappropriately influence policy. 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

In a 2005 speech at the National Press Club in Australia, David Davies, then Philip Morris International's (PMI) Senior Vice-president Corporate Affairs, suggested PMI shared significant common ground with the public health community, and that his company could be ‘a positive contributor, indeed a partner in shaping future policy for tobacco’ (authors’ emphasis).1 Davies was referring to the concept of ‘tobacco harm reduction’, a term that has increasingly entered the tobacco industry's vernacular. For example, in 2008 British American Tobacco (BAT) set up a special website (http://www.bat-science.com) promoting their research and development (R&D) efforts in developing potentially reduced harm tobacco products. David O'Reilly, BAT's Scientific Director, states on this website that he believes that ‘tobacco harm reduction has the potential to be one of the biggest public health opportunities of this generation’ (accessed March 2013).2 Similarly, other BAT employees have begun publicly commenting on academic papers addressing harm reduction issues, offering ‘to work with the public health community and regulators to achieve a reduction in the public health impact of tobacco use’.3 4

The public health community has been divided over the possible benefits of tobacco harm reduction, with the debate in Europe focused hitherto on a low risk smokeless tobacco (SLT) product, snus,5 and more recently on e-cigarettes.6 Public health interest in snus began following observations of the so-called ‘Swedish Experience’, which attributes Sweden's low male smoking prevalence, and resulting low levels of tobacco-related mortality, to high rates of snus use among Swedish men.7–9 With overwhelming evidence that snus is considerably less harmful than smoking, and that nicotine addiction is the key reason why smokers continue to smoke, this raises the potential for lives to be saved if smokers could switch from cigarettes to using nicotine in a less hazardous form such as snus.10

While an Australian modelling study suggests that the introduction of snus should reduce harm at a population level,11 others argue that the extent to which this may occur will depend on how snus is marketed, who takes it up, and whether it successfully enables smokers to quit cigarettes.12 13 Concerns focus mainly on the possibility that marketing of snus may lead to an increase in dual tobacco use rather than a decrease in cigarette smoking or to snus acting as a gateway into tobacco use for non-smokers.

In the European Union (EU), with the exception of Sweden, the sale of snus has been prohibited since 1992.14 Despite this ban, the four transnational tobacco companies (TTCs) operating on the European tobacco market (BAT, PMI, Imperial Tobacco (IMT) and Japan Tobacco International (JTI)) have been acquiring, or entering into joint ventures with, Swedish snus manufacturers, to the extent that small independent manufacturers only account for an insignificant proportion of the Scandinavian snus market (the only snus market in Europe).15 16 Concurrently, TTCs have started publicly opposing this ban on snus sales, supporting their position with a harm reduction narrative.17–19 Yet recent research shows this narrative is inconsistent with the TTCs’ private documentation which suggests that harm reduction is not central to their business strategy.16 In addition, evidence from the USA, an established SLT market, suggests that snus is being marketed to augment cigarette use.20–23

This paper seeks to examine TTCs’ use of the term ‘harm reduction’. It examines when and why harm reduction first entered TTC's discourse, and explores the TTCs’ approach to and perceived benefits of harm reduction. In so doing it aims to assess whether the tobacco industry's interest in harm reduction reflects opportunistic tactical adaptation to policy change, or a substantive commitment to harm reduction.

Methods

Our study was based on internal tobacco company documents, available on the online Legacy Tobacco Documents Library (http://legacy.library.ucsf.edu/) following litigation in the USA.24 The Legacy Library was searched between May 2010 and February 2013, with searches focused on the two largest TTCs, PMI and BAT. PMI has been through various restructures, with its current form established in 2008 when it spun off from Altria (known pre-2003 as Philip Morris Companies).15 For simplicity purposes, we refer to ‘PMI’ throughout this paper. Documents were initially retrieved using broad search terms (eg,‘harm reduction’, ‘smokeless tobacco’, ‘R&D strategy’). The documents retrieved helped identify secondary search terms including relevant internal committees (eg, BAT's Tobacco Executive Committee and BAT's Research Policy Group) and key personnel. Surrounding Bates numbers of key documents were also searched. The iterative process of searching, analysing and refining resulted in a final set of 455 documents, dating from 1971 to 2009. Analysis of these documents was based on a hermeneutical approach to company document analysis,25 complemented by sociohistorical archival techniques.26

To validate and update this documentary evidence, and specifically to explore when and how tobacco industry discourse on harm reduction was introduced and changed, we triangulated the documentary evidence obtained with data from other tobacco industry sources. First, we searched the TTCs’ corporate websites (http://www.bat.com, http://www.pmi.com, http://www.jti.com, http://www.imperial-tobacco.com) to identify their position on harm reduction. Second we searched TTCs’ annual reports for the terms ‘harm reduction’, ‘reduced harm’, ‘smokeless’ and ‘snus’, recording all occurrences. This analysis was undertaken on all reports accessible online, via the British Library and personal contacts in October 2012: 1997–2011 annual reports for BAT (minus the 1998 report); 1997–2011 annual reports for IMT; 2005–2012 annual reports for JTI; and 2002–2011 annual reports for PMI.

As our document and website searches identified BAT as particularly vocal on harm reduction we further explored BAT's use of the term harm reduction. We searched BAT's sustainability reports from 2001/2002 (the first year of publication) to 2011 (prior to 2007 these were referred to as social reports) using the same search terms as used for the annual report searches. In October 2012 and March 2013 we accessed the Internet Archive Wayback Machine (http://www.archive.org), a historic archive of preserved webpages based on web crawler technology, which allows archived versions of websites to be visited. We entered BAT's URL (http://www.bat.com) into the Wayback Machine, selecting the earliest web page that was returned in the results (29 October 1996) and searched this page for BAT communications on harm reduction and SLT (note: the search revealed that BAT has only owned this URL from 2000, with an unrelated company owning the URL from 1996–2000). Relevant content and observations were systematically recorded. We repeated this process for all archived BAT webpages up to the last entry in 2012 (dated 23 December).

Finally, between November 2010 and January 2011 we undertook six semistructured interviews with seven key informants including three Swedish public health experts and four senior tobacco industry representatives from the European Smokeless Tobacco Council (ESTOC), BAT and Swedish Match (Europe's biggest SLT manufacturer). JTI, PMI and IMT were also invited for interview, but either declined or did not respond to our invitation. We deliberately approached Swedish public health experts because of the legal status of snus sales in Sweden. Interviews were professionally transcribed and analysed using a framework approach where emerging data were coded using a thematic structure which drew on themes identified from the existing literature while also allowing for new themes to emerge from the analysis itself.27

Results

Emergence of the TTCs’ discourse on harm reduction

Institute of Medicine (IOM) report, ‘Clearing the Smoke’ and the emergence of the tobacco industry's harm reduction discourse

TTCs only began using the term ‘harm reduction’ from 1999 onwards and consistently from 2000.28–33 This coincided with the US Food and Drug Administration (FDA) commissioning the IOM to study the feasibility of tobacco harm reduction.34 Documentary, interview and other tobacco industry material findings suggest that this study, which took place over 1999–2001 and resulted in the ‘Clearing the Smoke’ report,35 was a turning point in the tobacco industry's approach to, and dialogue on harm reduction. In February 2000, the IOM invited US-based tobacco manufacturers to meet with its Working Group to discuss ‘the lessons learned by the tobacco industry from previous attempts at developing products for harm reduction’.36 This approach was welcomed by the tobacco industry,37–39 and it is around this time that the term ‘harm reduction’ gradually, but consistently, appears in the internal tobacco industry's documents.29 40–48 Prior to 1999 no retrieved BAT or PMI documents mention the term ‘harm reduction’. Instead they refer to a variety of terms including ‘reduced risk’,49 50 ‘smoking and health’,51 ‘risk minimization’,52 or ‘reduction of noxae’,53 which appear to reflect the research focus at the time on developing a ‘safer’ cigarette.54–56

Use of harm reduction in TTC public documentation

Our analysis of TTCs’ annual reports and BAT's social reports and website also suggests that ‘harm reduction’ only began to be used in TTC public documentation from 2002 onwards, its use increasing steadily thereafter albeit more by some TTCs than others (tables 1–3). BAT's first social report (2001/2002) is the first public report in which we identify any TTC mentioning ‘harm reduction’ (table 2).57 The fact that this report was published in 2002 limits our ability to assess whether BAT previously used this term, but the finding that its annual reports did not mention harm reduction until 2004 (table 1) make this unlikely. Similarly, analysis of BAT's website (http://www.bat.com) suggests that BAT started referring to ‘less harmful’ tobacco products on its website in 2005, simultaneously announcing the expansion of its product portfolio with snus (table 3).

Table 1

References to ‘harm reduction’ or ‘reduced harm’ and ‘smokeless tobacco’ or ‘snus’ in TTCs annual reporting

	BAT*		IMT		JTI†		PMI
Year	HR	SLT	HR	SLT	HR	SLT	HR	SLT
1997	0	0	0	0	–	–	–	–
1998	–	–	0	0	–	–	–	–
1999	0	0	0	0	–	–	–	–
2000	0	0	0	0	–	–	–	–
2001	0	0	0	0	–	–	–	–
2002	0	0	0	0	–	–	0	0
2003	0	0	0	0	–	–	0	0
2004	5	0	0	0	–	–	1	0
2005	9	29	0	9	0	0	0	0
2006	12	9	0	8	0	0	0	2
2007	11	14	0	6	0	1	0	7
2008	6	47	0	6	0	0	3	1
2009	7	15	0	14	0	0	3	2
2010	13	47	0	13	0	16	1	1
2011	9	5	0	12	0	9	0	1
2012	–	–	0	9	0	8	–	–

*BAT's 1998 annual report is unavailable.

†Results are based on JTI's annual reports, which also include reporting on JTI's domestic tobacco market and its pharmaceutical and food business. Earlier copies are were not available on JTI's website, or in the British Library.

BAT, British American Tobacco; HR, ‘harm reduction’ or ‘reduced harm’; IMT, Imperial Tobacco; JTI, Japan Tobacco International; PMI, Philip Morris International; SLT, ‘smokeless tobacco’ or 'snus’; TTCs, transnational tobacco companies.

Table 2

Number of references to ‘harm reduction’ or ‘reduced harm’ and ‘smokeless tobacco’ or ‘snus’ in BAT's social reports from 2001/2002 to 2011 (name changed to ‘sustainability reports’ from 2007 thereafter)

	2001/2002	2002/2003	2003/2004	2004/2005	2005	2006	2007	2008	2009	2010	2011
HR	2	2	3	54	72	16	41	41	100	247	232
SLT	0	2	0	29	117	5	38	48	52	35	19

BAT, British American Tobacco; HR, ‘harm reduction’ or ‘reduced harm’; SLT, smokeless tobacco.

Table 3

Timeline of BAT harm reduction discourse on BAT's corporate website (http://www.bat.com), based on archived webpages (dating from 1996–2012) available from http://www.archive.org

Date	Emergence and key changes to harm reduction and snus rhetoric
1996–2000	▸ Website owned by an unrelated non-tobacco company
May 2000	▸ BAT set up website to ‘help you know us (BAT) a little better, and to balance the debate on issues that can be controversial’. ▸ Drop down main menu directs to BAT's views on smoking, marketing, corporate citizenship, environment, health and safety, and people. ▸ Claims that smoking as a cause of certain diseases is the ‘working hypothesis of much of our product modification research’.
May 2005	▸ Press release on BAT's snus investments added, saying it aims to give ‘smokers the chance to enjoy a less harmful form of tobacco’ ▸ BAT also states it supports the lifting of the EU ban on snus sales, pointing at the ‘significantly lower health risks’ and arguing that marketing should be aimed only ‘at adults who have chosen to consume tobacco’
Aug 2007	▸ Significant website restructure, with new main section on ‘health and science’. First time that ‘harm reduction’ and ‘smokeless tobacco and health’ are dedicated subcategories. ▸ BAT claims its investment in snus is ‘in line with our [BAT's] continuing efforts in harm reduction and a response to those public health stakeholders who told us they believe that snus, properly regulated, can contribute to reducing the health impact of tobacco use’.
March 2008	▸ Harm reduction is now referred to as ‘a key element of our [BAT's] business strategy’. ▸ BAT's new website, bat-science.com, is promoted (‘written by scientists for scientists’), with a new external link to the IOM report added.
Oct 2010	▸ BAT adds video ‘The Science of Harm Reduction’ to its website.
April 2011	▸ BAT announces its establishment of Nicoventures, a stand-alone company which will focus on nicotine-only products, calling it ‘a natural extension’ of their ‘approach to tobacco harm reduction that has been developed over years’.
June 2011	▸ BAT indicates it no longer runs snus test markets in South Africa, Canada and Japan, and only sells snus in Sweden and Norway.
March 2012	▸ The ‘smokeless tobacco and health’ section has been moved away from the core part of the Health and Science part, to a drop-down menu on the left margin.
Dec 2012	▸ BAT announces it has acquired an e-cigarette company and claims, as with Nicoventures, it's a natural extension of their approach to harm reduction.

BAT, British American Tobacco; IOM, Institute of Medicine.

TTC approach to and perceived benefits of harm reduction

Dialogue with, and access to, scientists, public health community and policy makers

Shortly after the IOM initiated its study, BAT commissioned consultancy firm, The Lewin Group to establish a public engagement schedule to support the development of a ‘Harm Reduction Strategy’.58 The Lewin Group's proposal suggested that dialogue take place with health professionals and policy makers to explore how they might find harm reduction acceptable, followed by a ‘public forum’ with smokers aiming to ‘stimulate public debate, raise issues relating to harm reduction for committed smokers and provide a platform for further public relations’ (authors’ emphasis).58 It is unclear whether BAT followed through on The Lewin Group's proposal, although in November 2000 BAT organised a risk assessment workshop for academics in the UK59–61 to discuss ‘‘Tobacco Harm Reduction’ Assessment Criteria’.62 Six BAT scientists and nine external scientists attended,63–65 including toxicologist Jim Bridges who chaired the workshop on BAT's behalf.61 65–67 Documents show that BAT attempted to include the UK Department of Health (DH) in this event, but that the DH declined.60 68

Subsequently, ‘The strategy of using the IOM report as a catalyst for further dialogue with UK scientific stakeholders’ was discussed at an April 2001 BAT operational planning meeting.69 In an interview, the BAT representative also stressed the importance of dialogue with the public health community, stating that the tobacco industry and public health community have previously worked in silos, whereas dialogue would enable BAT to determine what, and how much, harm reduction science it should be doing, and would enable public health to better understand smokers and their needs (interview December 2010). Significantly, BAT and ESTOC interviewees attribute BAT's decision to add snus to its portfolio, and brand snus with existing cigarette brands like Lucky Strike (an issue which appears to be controversial within industry, see box 1) to dialogue with, and encouragement from, UK public health experts70 (interviews 25 November and 9 December 2010).

The issue of cigarette branded snus appears to be debated within the tobacco industry. The Swedish Match interviewees noted that their market research indicates that full time snus consumers (including ex-smokers) do not want snus linked to cigarettes or smoking, and expressed discomfort about cigarette branded snus (interview January 2011). Similarly, the European Smokeless Tobacco Council interviewee claimed that cigarette branded snus had not succeeded in the Swedish market, and suggested that non-cigarette branded snus was better able to communicate a reduced risk message (interview November 2010). The BAT interviewee on the other hand reported that its decision to brand snus with a cigarette brand (ie, Lucky Strike and Peter Stuyvesant) had been based on advice from UK public health experts, who had argued that if BAT were to be serious about switching smokers from cigarettes to snus, they should put their biggest cigarette brand on it (interview, December 2010). Interestingly, Euromonitor 2009 sales data indicate that cigarette branded snus represented only 2% market share in Sweden,71 and 3.6% in Norway,72 with BAT's traditional snus brands (Granit and Mocca) showing slightly bigger growth than its Lucky Strike and Camel Snus.

TTCs’ use of harm reduction to signal alignment of tobacco industry interests with public health is also apparent in contemporary corporate materials. Public health experts’ views on the benefits of harm reduction are cited on BAT's and PMI's websites,73 74 and when BAT first integrated a section on harm reduction on its website it stated that BAT would continue ‘to seek common ground on harm reduction with health policymakers, who are looking to achieve a reduction in the net public health impact of tobacco use’ (archived version http://www.bat.com 9 August 2007, Harm Reduction section, accessed 13 March 2013). PMI's 2008 annual report states that it supports regulations that ‘are aligned with the public health objective of reducing harm caused by tobacco products’ (yet noting that it won't ‘support every proposal made by public health groups’).75

Similarly, TTCs’ websites76–78 and BAT company reports57 79–81 underline a role for harm reduction in establishing dialogue with and access to policy makers. BAT suggests in its first Social Report published in 2002 that the tobacco industry and governments ‘work together as partners, rather than adversaries’.57 In 2009 BAT's website reported that BAT had contacted the FDA (which had just been given the authority to regulate tobacco products in the USA) to share information about BAT's proposed harm reduction framework and ‘aid regulators in assessing any such (potentially reduced harm) new products’ (archived version http://www.bat.com 30 September 2009, Harm Reduction section, accessed 13 March 2013).

TTCs’ use of harm reduction to gain access to policy makers was also reflected in the internal documents. For example, in July 2000 BAT arranged a meeting with DH officials to discuss risk communication and ‘safer’ cigarettes.82 Afterwards BAT noted there was an appetite among government officials to ‘reach consensus’ on the harm reduction issue, and that harm reduction could be used to establish ongoing access to, and dialogue with, senior government officials.82 Likewise, documents suggest that in 2000 PMI had had dialogue on reduced harm product development with IOM, Health Canada, the EU Commission, World Health Organization (WHO), among others,83 and had approached the DH to set up a meeting ‘to share further information’ on reduced harm products.84–86 On 12 November 2001 DH representatives visited PMI's R&D facility in Germany, INBIFO.87 Whereas, in the 1970s and 1980s, PMI had attempted to ensure its ownership of INBIFO was kept secret (a time when the tobacco industry was still claiming that evidence of the toxic effects of smoking was inconclusive),87 PMI now welcomed government officials to this facility to discuss its harm reduction programme.43 88–90 Significantly, documents suggest this study trip was initiated by the DH.90–92

That policy makers remain a key target of the TTCs’ harm reduction discourse was confirmed by the BAT interviewee who noted that BAT was trying to engage with governments on harm reduction (interview December 2010). Despite BAT's efforts to engage in dialogue with DH on harm reduction in 2009 and 2010 (the latter after a new government came into power subsequently entering into ‘responsibility deals’ with the food and alcohol industries93), efforts had been unsuccessful (interview December 2010).

Responsibility and ‘Improvement of credibility’

Simultaneous with the public health community's emergent interest in tobacco harm reduction,42 TTC's corporate social responsibility (CSR) agenda (initiated by PMI and BAT in 199793–95) was increasing in prominence and our findings suggest they were mutually reinforcing. Existing research shows the CSR agenda emerged once evidence from US lawsuits began to seriously damage the tobacco industry's reputation, ending its insider status with governments and,96 its golden era, dominated by voluntary agreements between government and an industry that ‘could be relied on to act responsibly’.16 97 To regain the policy influence TTCs had previously enjoyed, BAT recognised the importance of rebuilding its reputation.93

Shortly after the IOM invited the tobacco industry to engage on harm reduction, BAT's Adrian Marshall (who had established BAT's CSR programme in 199998) identified harm reduction as one of six key ‘reputation management initiatives’99 that would help BAT rebuild its reputation as a responsible company.39 Shortly afterwards, BAT's scientist, Eian Massey, highlighted the reputational opportunities the IOM study created, recommending that its US subsidiary should ‘address the agenda of us being a responsible Company’40 in its correspondence with IOM. Previous research shows that BAT considered harm reduction a preferred CSR initiative because it was politically salient, was unlikely to be seen as undermining other tobacco control policies, and dialogue on the issue could be represented as ‘the morally right thing to do’.93

Like BAT, PMI also considered harm reduction an opportunity to rebuild corporate legitimacy. Handwritten comments on the agenda of a September 2000 PMI meeting in Hong Kong to brief staff on PMI's research strategy, noted that ‘improvement of credibility’ was to be the overall result of PMI's harm reduction efforts.42 Similarly, an April 2001 draft statement of CSR principles declared that PMI ‘will focus our efforts on harm reduction as it applies to our product, and to or policies, programmes and positions’100 and PMI staff surveys in 2002 identified harm reduction as central to corporate reputation.101

That harm reduction became important to TTCs’ CSR strategies, was confirmed by the BAT interviewee who reported that BAT's early stakeholder dialogue events, initiated in 2001 as part of its emergent CSR programme,94 made BAT understand the importance of doing, and being seen to do, reduced harm research (interview December 2010). BAT's subsequent social reports57 appear to have addressed this concern; mentions of harm reduction increasing over time to become a key focus of these reports (table 2). BAT's website was also established in 2000 ostensibly as part of BAT‘s emerging CSR strategy given that the website's purpose was ‘to explain more about what we do and what we believe in’,102 including highlighting that BAT was involved in ‘product modification research’ to address health concerns (table 3). By drawing attention to their attempts to reduce the risk of smoking, TTCs positioned themselves as partners rather than adversaries in achieving public health gains.57 In its 2010 Sustainability Report, BAT remarked that ‘stakeholders who do talk with us often seem surprised by our candour and willingness to listen. Some are people and groups who would not have engaged with us a few years ago’.103

Discussion

This study demonstrates that TTCs’ adoption and subsequent use of the term harm reduction occurred in direct response to the public health agenda (which TTCs had been monitoring closely). The ‘Clearing the Smoke’ study by IOM appears to have been pivotal in shaping the tobacco industry discourse on harm reduction. Our findings demonstrate that harm reduction serves TTC interests in two main, mutually reinforcing, ways. First TTCs use harm reduction to facilitate access to, and dialogue with scientists, public health experts and policy makers, presenting themselves as ‘partners, rather than adversaries’57 who share a common goal. Second, TTCs considered harm reduction a ‘reputation management initiative’,99 facilitating the tobacco industry's image rehabilitation as a ‘responsible business’. From the beginning, harm reduction was intimately linked to BAT's and PMI's emerging CSR strategies, providing a means of increasing corporate credibility with stakeholders; with social reports and corporate websites used to communicate this message.

Our findings are consistent with previous research showing that the TTCs saw HR as a means of improving their corporate reputation to regain access to regulators and, with that, influence over tobacco control policy.93 96 Both support our earlier work16 which also challenges TTCs’ purported commitment to harm reduction by demonstrating that TTCs’ interest and investments in SLT in Europe were defensive and originated from a desire to generate new tobacco sales ‘without cannibalising existing profits from cigarettes’,104 rather than a desire to reduce harm from tobacco. This paper details how TTCs intended to harness those opportunities and suggests that TTCs’ harm reduction discourse should be seen as opportunistic tactical adaptation to policy change rather than a genuine commitment to harm reduction. While our analysis strongly suggests that BAT's and PMI's main interests in harm reduction were access and reputation, we cannot rule out the possibility of a genuine commitment to harm reduction. We note, however, that there was very little evidence of this relative to the evidence of reputational and access benefits.

A further finding of interest is the difference between tobacco companies in their approach to snus branding. The cigarette manufacturers have branded some of their snus with cigarette brands (according to BAT in response to public health advice), while Swedish Match, which no longer has any cigarette interests, cites market research that snus users (including ex-smokers) prefer snus brands. Both Swedish Match and ESTOC appear to contest whether the cigarette branded snus produced by cigarette manufacturers can contribute to harm reduction. To our knowledge there are no published data on this. This highlights the complexity and the lack of detailed knowledge about the potential for snus as a harm reduction product.

Limitations

This study has several limitations. The nature of tobacco industry document research means that we made decisions about which search terms and documents were most relevant. Inadvertently, this may have resulted in some relevant documents not having been included in the analysis. Furthermore, the archives comprise only those documents obtained through the discovery process in litigation. Documents may have been innocently omitted, intentionally destroyed or omitted, or inappropriately classified as privileged. To overcome these limitations we were as comprehensive as possible in our searching and reached a point of document saturation, where new searches led to documents already retrieved; an indicator that most important documents have been identified. We also used contemporary materials and interviews to triangulate our findings and overcome the issue that most retrieved documents predate 2002. The document collections do not include documents from IMT, JTI or Swedish Match; these companies may take a different approach to harm reduction than BAT and PMI (eg see box 1).

Using the Wayback Machine (archive.org) has its limitations as it is based on crawler technology which only indexes a fraction of the available content. Thus, the archived pages of BAT's website do not necessarily reflect the entire content of BAT's website at a given time. Rather, the archived website versions give a broad indication of which year BAT introduced communications on harm reduction and SLT.

Policy implications

Our analysis suggests that TTCs’ harm reduction discourse should be seen as an opportunistic tactical adaptation to policy change, and shows that TTCs, in the past and present, have tried to use harm reduction discourse to access public health policy makers. Care must be taken that the harm reduction debate does not allow TTCs to re-enter the policy arena from which they have increasingly been excluded in line with Article 5.3 of WHO's Framework Convention on Tobacco Control, which aims to protect public health from the commercial and other vested interests of the tobacco industry.105

We note that various meetings discussing harm reduction, including public health conferences in the UK and FDA policy forums in the USA,106 107 have included tobacco industry representatives. In part this is because of blurring of the definition ‘tobacco industry’ given TTCs’ recent investments in pure nicotine and e-cigarettes. Our findings however suggest that this will merely serve to ‘re-normalize’ an industry that is determined to be seen as a responsible business with a legitimate product, exactly as, this paper shows, industry intended.

The ‘Clearing the Smoke’ study by the Institute of Medicine has been pivotal in shaping tobacco industry discourse on harm reduction. Transnational tobacco companies (TTCs) adopted the term ‘harm reduction’ in response to this study, and proceeded to heavily emphasise the term in corporate messaging.

Apart from trying to make a safer cigarette, documents show that TTCs’ use of harm reduction was aimed to gain access to scientists, the public health community and policy makers, and rebuild its reputation as a responsible industry.

Although we can't rule out the possibility of a genuine commitment to harm reduction, this study suggests that TTCs’ harm reduction discourse should be approached critically and seen as opportunistic tactical adaptation to policy change.

This study underlines the importance of ensuring that the harm reduction debate does not allow TTCs to re-enter the policy area from which they have increasingly been excluded in line with Article 5.3 of WHO's Framework Convention on Tobacco Control.