key: cord-0793601-bzlnim6s
authors: Matthes, Britta Katharina; Lauber, Kathrin; Zatoński, Mateusz; Robertson, Lindsay; Gilmore, Anna B
title: Developing more detailed taxonomies of tobacco industry political activity in low-income and middle-income countries: qualitative evidence from eight countries
date: 2021-03-23
journal: BMJ Glob Health
DOI: 10.1136/bmjgh-2020-004096
sha: 58fdf76962d8e5bb7f26fa3f7e606fec661ea024
doc_id: 793601
cord_uid: bzlnim6s

INTRODUCTION: Historical evidence, predominantly from high-income countries (HICs), shows that the tobacco industry uses a recurring set of arguments and techniques when opposing tobacco control policies. This data formed the basis of a model of tobacco industry political activity known as the policy dystopia model (PDM). The PDM has been widely used in tobacco control research and advocacy and has subsequently been shown relevant to other unhealthy commodities industries in both HICs and low-income and middle-income countries (LMICs). Before it can be validated as a generic tool for researching corporate influence on policy, one needs to determine whether the PDM successfully captures contemporary corporate political activities in LMICs. METHOD: We conducted semistructured interviews with 22 LMIC-based advocates and used the transcripts as the primary data source. The discursive and instrumental taxonomies constituting the PDM served as the starting point for the coding framework. Using thematic analysis, we combined deductive and inductive coding to ensure we captured all strategies from the PDM and the interviews. RESULTS: This study found that the tobacco industry uses a set of discursive and instrumental strategies that is largely consistent across LMICs and with the PDM. We identified several minor contextual nuances absent from the PDM. Some of these nuances were characteristic to individual countries, while others to LMICs more broadly. They included the argument that tobacco control policies unfairly punish reputable tobacco industry actors, and an emphasis on instrumental strategies centred around maintaining a good image, rather than rehabilitating a tarnished image as emphasised in the PDM. CONCLUSIONS: Allowing for the nuances identified in this study, the PDM has been found to be fit for purpose. The revised model should now be tested through in-depth LMIC case studies and could be used to facilitate comparative studies of unhealthy commodity industries’ political activities.

Over recent decades, significant progress has been made in understanding and exposing the corporate political activities of unhealthy commodity industries; activities which pose a substantial obstacle to advancing public health policies. [1] [2] [3] [4] [5] [6] [7] [8] This research is furthest

What is already known? ► Historical evidence, predominantly from high-income countries (HICs), indicates that the tobacco industry has repeatedly used a similar set of arguments and techniques when opposing tobacco control policies. ► Based on this evidence, a model of tobacco industry political activity known as the policy dystopia model (PDM) was developed and has been used in tobacco control and in research on other unhealthy commodities industries.

What are the new findings?

► This research shows the PDM effectively captures corporate political activities in low-and middleincome countries (LMICs), while also identifying context-specific variations in arguments and techniques the tobacco industry uses. ► Context-specific arguments identified across multiple LMICs included claims that tobacco control policies would negatively impact development, and tobacco farmers, were not in the national interest but driven by a foreign agenda, and that policies used in HIC-settings may not work in LMICs. ► Looking at the techniques, we found that reputation management emphasised building and maintaining a favourable image rather than the image rehabilitation seen in HICs, and that the industry sought to build long-term relations with media professionals and intimidate advocates.

What do the new findings imply?

► The PDM provides an appropriate tool for studying, predicting, and countering tobacco industry political activity across countries of all incomes, but it requires some flexibility to capture context-specific strategies.

BMJ Global Health advanced in tobacco control in part due to the release of millions of pages of internal tobacco industry documents through litigation. 9 To make sense of this growing literature which primarily took the form of country-specific or policy-specific case studies, we previously conducted two systematic literature reviews, one on tobacco industry influence on tobacco taxation 10 and one on tobacco industry influence on marketing regulation. 11 These showed that the industry consistently uses the same strategies to influence policy. The two reviews were used to create an evidence-based model of tobacco industry political activity known as the policy dystopia model (PDM). 12 The purpose of the PDM was to enable to predict, pre-empt and counter tobacco industry efforts to influence tobacco control policies and to facilitate comparative research on other unhealthy commodity industries. 12 13 The PDM conceptualised a dynamic model of political influence with five discursive (argument-based) and five instrumental (action-based) strategies. All of these were found to act synergistically to achieve the industry's preferred policy outcomes (see figure 1 ).

The PDM's ongoing relevance to contemporary tobacco industry influence in high-income countries (HICs) has been established 14 15 and research using the PDM and the previous reviews also showed that alcohol, food and gambling companies use broadly similar strategies. [16] [17] [18] [19] [20] [21] [22] [23] [24] However, a potential weakness remains that the PDM mostly draws on evidence from HICs. Eighty per cent of the papers (52 papers) included examined corporate political activities in HICs, and less than 8% (five papers) focused on low-income and middle-income countries (LMICs) (the remaining eight papers had a regional or transnational focus), reflecting the literature base at the time. Further, all the LMIC-based articles were published before 2010 and often described corporate political activities before 2000.

While two recent studies using the PDM to code corporate political activities of the tobacco industry in Nigeria and Brazil found broadly similar tactics, 25 26 there remains a need to validate whether the model successfully captures contemporary tobacco industry political activities in LMICs, particularly given structural changes to the tobacco industry. 27 Such work is particularly urgent since 80% of all smokers live in LMICs, 28 and because the adoption and implementation of the Framework Convention on Tobacco Control (FCTC) tends to be slower and weaker in those countries. 29 30 This study examines the tobacco industry's political activities in a diverse set of LMICs using insights from tobacco control advocates. It aims to answer the following questions: ► What discursive and instrumental strategies does the tobacco industry use in LMICs? ► Do the strategies used by the tobacco industry vary by individual country and context? ► Do these strategies vary from those used in HICs? ► To what extent does the PDM provide a suitable tool for studying and predicting tobacco industry political activities across countries and income contexts? A better understanding of tobacco industry political activities globally can help strengthen tobacco control and support advocates' efforts to pre-empt and counter industry activities. 31 It can also facilitate comparative studies on the tobacco industry and other unhealthy commodity industries.

We purposively selected eight LMICs which had recently adopted or consulted on a key tobacco control policy. These included comprehensive regulations (Ethiopia, Uganda, Zambia), health warning regulations (Bangladesh, India, Sri Lanka) and tobacco tax increases (Colombia, Ukraine) (table 1). Since we sought to record experiences from a diverse group of countries, the eight LMICs are located in four WHO regions and represent the three LMIC income-economy groups.

Participants had to have at least 3 years of experience in tobacco control advocacy at the national level. They were also required to be fluent in English, which limited the pool of potential interviewees, but helped avoid challenges of working with multiple interpreters. 32 Figure 1 Policy dystopia model. 12 

To document varying perspectives from each country, we sought to include two to four participants from at least two civil society organisations (CSOs) per country. We focused on CSOs rather than on non-governmental organisations (NGOs), as it is a broader category that includes, among others, NGOs, charities and professional bodies 33 and hence more adequately reflects the landscape of tobacco control organisations.

We identified potential participants in and through our networks of tobacco control advocates and researchers, followed by snowball sampling. We approached potential participants with an information sheet via email.

We developed a semistructured interview guide to explore participants' experiences of the discursive and instrumental strategies the tobacco industry used in their country. The questions were openended but we also probed important elements of the PDM if they had remained unmentioned. The interview schedule also explored participants' accounts on countering corporate political activities and what would facilitate these efforts; those data form part of a separate study. 31 We piloted the guide with a tobacco control advocate and researcher who was, as most participants, not an English native speaker. All interviews were conducted in English, recorded with participants' permission, and subsequently transcribed.

Thematic analysis, facilitated by NVivo 12, was used to identify patterns across the interview transcripts. 34 The taxonomies of the PDM 12 served as the starting point for the coding framework (see online supplemental tables 1 and 2). In addition, we used an inductive approach to identify strategies not included in the PDM.

BKM led the coding and KL and MZ second-coded 50% of the transcripts each. BKM met regularly with ABG, KL, MZ and LR to discuss key findings. To ensure the validity of our results, we compared the newly identified strategies with academic case studies of tobacco industry political activity in LMICs.

All participants gave consent to participate.

No patient or public involvement.

Between June and October 2019, we conducted 20 interviews with 22 participants from eight countries and 18 different organisations. Five interviews were in-person; 15 took place online, using Microsoft Teams; 2 interviews involved 2 participants. The average duration per interview was 90 min. Table 2 summarises the distribution of interviewees by type of change and country.

Consistent with the PDM, advocates' accounts suggest that the tobacco industry's overall approach was to create a comprehensive and convincing narrative which proposed that tobacco control policies are undesirable by exaggerating their potential costs and dismissing potential benefits.

The most important and commonly mentioned arguments: illicit trade and farming Participants consistently identified the claim that the proposed policy would increase illicit tobacco trade as the tobacco industry's preferred argument. The industry would use this argument to fight a range of proposed measures, including taxes, health warnings and plain packaging. As one participant put it, 'they were very comfortable with the argument of smuggling. They use it for everything' (P15). In countries where tobacco farming takes place, participants consistently reported that the industry argued the policy would hit tobacco farmers and farming communities, often claiming that farmers would struggle to switch to other crops or activities. This argument went beyond job losses and highlighted the impact of the policy on farmers' livelihoods more broadly. When farmers' groups appeared, for example in a public hearing, they would emphasise the impact of the proposed policies on their communities 'they said you can't stop tobacco, our school fees come from tobacco, our food comes from tobacco' (P3). The industry portrayed tobacco farmers as victims of tobacco control policies and in some countries, it even claimed that tobacco control measures led to suicides among farmers.

Other commonly mentioned arguments: economic impact and legal concerns Two more arguments were common across the studied countries but were mentioned less frequently and seen as somewhat less critical by the interviewees. First, the industry often referred to potential job and income losses in the retail sector as a result of the proposed policy. Such claims were often repeated by organisations which participants identified as industry front groups, including retailer associations. The industry also highlighted the potential damage the proposed policies would cause to the country's economy and development. This argument was particularly emphasised when the industry tried to target ministries of finance. In one example, a participant recalled that a business group, which was perceived as a tobacco industry front group, held a meeting that 'was really aimed at looking at the FCTC and how it was going to damage the economy' (P22). The Minister of Finance attended this meeting.

The second argument related to legal concerns that the proposed policies would be unconstitutional or that the relevant public body does not have the power to adopt them. Participants reported that such arguments often formed the basis for legal actions (see instrumental strategies-litigation). The unconstitutionality claim was supported by a range of arguments, including that policies would violate individual rights and the industry's right to trade or practice free speech. In one country, interviewees recalled that the industry claimed that the Minister of Health would not have the power to issue the regulation; the legal actions initiated by the industry delayed the implementation of the regulation by years.

A prevalent argument focusing on the industry's role: penalisation of a reputable industry Participants consistently reported that the industry often highlighted its beneficial contribution to the country's development. It argued that proposed tobacco control measures would unfairly penalise the tobacco industry, a legitimate and reputable actor. [ This argument is multifaceted and was reported in reference to the tobacco industry being a major taxpayer, engaging in Corporate Social Responsibility (CSR) activities or purporting to lift farmers out of poverty. Participants reported that such claims appeared to carry weight and were repeated by policy makers, especially those from ministries of finance and agriculture.

A relevant context-specific argument: national interest Participants from one country reported that the industry claimed the proposed policy was not in the national interest and that a foreign agenda was driving tobacco control.

[They said] these NGOs [with external funding] are trying to do aggressive tobacco control, which is aimed at destroying the domestic tobacco industry, so that the foreign tobacco industry, which is basically multinational companies, can come and establish their market. (P12) This argument was perceived as impactful as it was accompanied by efforts to stop state funding for tobacco control groups (see instrumental strategies-coalition management).

Containing and denying benefits of the proposed policy Consistent with the PDM, the interviewees indicated that the industry tried to deny the public health benefits of the proposed policies. At the same time, it sought to avoid any discussions of the costs the policy would impose on the industry. Instead, as outlined above, the industry focused its arguments on the potential costs for other groups, including farmers and retailers.

A tobacco industry claim mentioned by participants from all countries was that the proposed policies would BMJ Global Health not work. The industry used several lines of arguments to support this claim. In the case of health warnings, it argued that it would not have sufficient time to prepare for the changes or that it would be impossible to implement them ('they argued that they didn't have any machinery to print graphical health warning on the pack' (P6)). To prevent a tax increase, participants reported that:

They said that if you continue to tax cigarettes, people will switch to cheaper forms, [country name] has cheaper forms of tobacco available. (P9) Across policies, the industry emphasised the difference between HICs and LMICs ('the industry said this works for the developed, but not in a developing country' (P20)). In one country, the industry linked this argument explicitly to the evidence base for tobacco control: Finally, another LMIC-specific justification for why the policy would not work linked to a lack of state capacity in implementing and enforcing regulation ('they said (…), the government is weak on implementation. So, it doesn't affect our market.' (P11)).

Consistent with the PDM, participants reported that the tobacco industry employed a range of instrumental strategies to construct and disseminate its dystopian narratives and convince policy makers to decide in its interest.

Interviewees viewed the tobacco industry's direct involvement in the policy arena as the most important strategy through which it sought to influence policy. Public hearings were one easy way for the industry to access the policymaking process. Sometimes, the industry spoke for itself ('I've seen four or five of their top officials attending the public hearing, and they were actively defending their interests.' (P10)). However, in several cases, the industry used front groups to defend its interest, for example, farmers. 'They bring farmers [to the public hearing] who clearly indicate if this law is ratified their livelihood would be significantly affected.' (P11)

Participants also frequently mentioned that the industry exerted influence on governmental decision making in less formal ways ('It mostly happens behind the doors, not through any public hearing.' (P9)). The lack of transparency of this approach was seen as particularly problematic: 'We don't know for instance, who they meet in the Ministry of Finance… we really have suspicions that there is a lot of underground influence.' (P3).

Furthermore, participants shared suspicions of how the industry used incentives and threats to secure influence but admitted that they were frequently unable to find concrete evidence. A common suspicion was centred around financial incentives ('They gave bribes to politicians who are moving our bill.'(P2)). In some instances, financial incentives were perceived as part of longer-term strategies aiming at securing wide support and building networks of influence: '[The tobacco industry] has a lot of influence in terms of sponsoring candidates for elections. [Company name] also funded political parties.' (P18)

Participants also recalled that the industry put policymakers under pressure. In several instances, companies threatened that they would leave a country if taxes increased or if 'you [the government] continue giving us a very difficult legal environment' (P19). Although no examples of industry following through on such threats were identified, the industry was thought to have successfully used them to avert undesired policy outcomes. One participant reported that the industry had temporarily stopped production in their country after health warnings became mandatory. 'But the health minister didn't really budge so then they resumed the production and went back to normal.' (P14)

Finally, interviewees suggested that individual policy makers, for example, parliamentarians leading on tobacco control legislation, had experienced intimidation from the industry. '[The industry] would go and intimidate his [MP leading the bill] voters, and intimidate him through his voters. So the voters would say "you're killing our livelihood, we will not give you a vote".' (P2)

Participants detailed the industry's widespread use of information management strategies, which in the PDM include producing and amplifying industry-favourable data or evidence while suppressing public health evidence; concealing industry links to evidence; and engaging in reputation management.

Of these, reputation management was most widely mentioned by our interviewees. Participants from half of the countries recalled the tobacco industry working with PR firms, and all participants detailed numerous CSR activities. Table 3 lists the CSR activities reported by participants. Each participant mentioned at least one example.

While most CSR activities were reported to occur countrywide or focus on specific vulnerable groups, some participants suspected that 'they claim that they are doing CSR for all, but actually they are doing CSR for their own contracted farmer' (P6).

Participants also reported that the industry sought to establish itself as a trustworthy government partner by offering support in difficult times.

One of the things they've done successfully… trying to show that they are a legitimate company… that is good to society… that they are authentic. (P9) This was considered highly impactful in facilitating access to policy makers, especially those from ministries of finance and agriculture who would often repeat such narratives. Participants consistently raised concerns about the lack of tobacco industry denormalisation and BMJ Global Health how this rendered the policy environment favourable to the industry.

Simultaneously, the industry sought to publicly discredit opponents by questioning advocates' motives ('[they claim] we are you know recruited and financed by other vested interests' (P10)) and practices ('[they say] we are not professionals' (P16)).

The industry has been able to effectively portray the tobacco control community as people who are against farmers, against economic development and that they are in fact against the national interest. (P14)

The industry also built relationships with media professionals to facilitate the amplification of industry-friendly information. Participants noted that the industry would offer financial and in-kind incentives to media professionals, run training and study tours for local journalists.

One participant shared that I used to work for these economic magazines, and we got the order from the editor was 'you don't change this', this came straight from the tobacco company's PR people and was directly printed as a regular article. (P15)

Interviewees pointed out that often newspaper articles that fitted this pattern did not mention industry links. The same lack of transparency applied to reports which skewed evidence to promote arguments against proposed policies. The industry was reported to pay think tanks, market research companies and professional service providers to produce such reports including 'jacked up numbers' (P9).

Participants reported that industry actors and front groups would amplify that information through the media and use it when accessing policy-makers. In at least one instance, the government was reported as drawing uncritically on industry data. The industry also occasionally contested public health evidence, in one case even questioning the claim that smoking causes cancerhowever, this was not a widely used strategy according to participants.

It was clear from the interviews that the industry made significant efforts to build a broad coalition to support its political activity. As in the PDM, this included forming coalitions between the major tobacco companies, collaborating with smaller companies (eg, bidi producers), working with umbrella industry organisations, as well as looking for allies in broader constituencies (eg, chambers of commerce, farmers' associations, retailers' associations, manufacturers' associations, printing business associations, law societies and think tanks). However, participants reported a lack of evidence on these relationships.

In tobacco farming countries, farmers' and tobacco growers' organisations were crucial supporters of the industry's cause. Interviewees recalled that while in some cases such organisations were long-standing, others appeared suddenly to coincide with a policy debate. Where formation of associations was prohibited, the industry would recruit individual farmers. Participants suspected the industry provided in-kind and financial benefits or put farmers under pressure.

We can put two and two together… I mean, there was a poor farmer and you're talking [about] the [luxurious hotel), a two day meeting, and he was booked. (P1) 

A common concern was that farmers' associations did not include or represent tobacco farmers working small plots, but middlemen or 'big commercial farmers, who are real friends to industry' (P22).

Participants also reported significant efforts to undermine the public health constituency, including through intimidation against advocates. For example, 'they [tobacco industry] started working through other ministries, trying to cancel their [tobacco control organisations'] foreign funding and trying to make things difficult.' (P9). Or, in relation to prominent advocates, 'it was like people following you, an anonymous letter reaching your office, anonymous phone calls.' (P4).

Legal action was another ubiquitous strategy used by the tobacco industry to prevent policies from being adopted or entering into force-especially when other strategies had failed. Participants from all countries reported that the industry took legal action, in some countries extensively ('Every single measure has been sued.' (P13)). Participants also recalled industry coalition members filing lawsuits.

In almost all countries, interviewees suspected tobacco industry involvement in illicit tobacco trade, but reported challenges in obtaining evidence for it. One participant suggested a direct link between a period of surge in illicit trade and a concrete upcoming policy event:

It's very hard to claim. But the level of illicit products circulating 90 or 120 days before the adoption of the law was significant in [capital city]. Every street vendor had at least two illegal products. (P12)

Context-specific factors shaping corporate political activities Despite considerable consistency across the countries included, this study uncovered some context-specific variation in industry strategies. We identified several factors which may shape these differences and can help explain differences between countries: 1. Corruption and conflict of interest. In countries with high levels of reported corruption, the industry's emphasis was on informal access strategies. In over half of the countries, participants reported that there were public bodies whose mandates conflict with tobacco control, and which could therefore provide easy routes for industry influence. Examples included a government owning tobacco industry shares, and ministries of agriculture and commerce establishing public bodies to promote tobacco growing and manufacturing. Several participants also shared examples of policy-makers with conflicts of interest (owning shares in a tobacco company, having a family business in the sector) and cases of industry staff taking up government positions. 2. Economic structure. In tobacco farming countries, the industry focused much political activity around farm-ing-it claimed negative impacts on farming and farmers as a key argument, built coalitions with farmers and farmers' organisations and targeted farming communities with CSR efforts. Furthermore, in a country with a sizeable domestic tobacco industry, the industry repeatedly used the argument about tobacco control constituting a foreign agenda. 3. Political institutions. One advocate reported that the industry targeted key members of government but no members of parliament since the latter would all be from the governing party and following the government agenda. In another country, where parliamentarians were perceived as very active in initiating legislation, they were reported to be a key industry target.

where most people's religious beliefs prohibited smoking tobacco, the tobacco industry's strategies did not focus on building public pressure but on influencing policy stakeholders directly.

Where the industry recently had a legal challenge overturned in a neighbouring country, it was less likely to use legal threats.

Our findings from eight LMICs suggest that the tobacco industry uses a set of discursive and instrumental strategies that is largely consistent across LMICs and with the PDM and its taxonomies 12 (which mainly drew on evidence from HICs). However, we identified specific nuances suggesting that the detailed arguments and techniques within the overarching strategies vary somewhat not only depending on the country's income group but also other contextual factors-economic, political and social. This is consistent with the LMIC-focused tobacco control literature emphasising that the industry tailors its approach to specific contexts. [35] [36] [37] [38] [39] With this caveat in mind, our findings suggest that the PDM provides an appropriate tool for studying and predicting tobacco industry political activities across countries of all incomes.

In the case of discursive strategies, there was remarkable consistency across the included LMICs and with the PDM's taxonomy (see online supplemental table 3) . Notably, the overall approach was to claim the policy was unnecessary and undesirable by exaggerating its costs and dismissing its benefits. All bar one strategy (benefits for undeserving groups) of the discursive strategies described in the PDM were detected in this study, with each mentioned by at least four participants. However, we identified a few new detailed LMIC-specific arguments within these broader strategies. For example, when suggesting that the policy would negatively impact on the country's economy, the industry would frequently claim a negative impact on development and, where relevant, on farmers and farming. Under the politics and governance domain, a new argument was that the policy was BMJ Global Health not in the national interest. Perhaps most importantly, we identified one additional discursive strategy-the industry's claim to be a good and reputable actor that is unfairly punished by tobacco control policies. This was closely linked to a widely used instrumental strategyreputation management. Finally, social justice arguments which are prominently used in HICs 40 were of minimal prevalence in our data, potentially because the industry did not expect them to gain traction.

In the case of instrumental strategies, there was again considerable consistency across the LMICs and with the PDM's taxonomy (see online supplemental table 4). All key instrumental strategies identified in the PDM were used, although two specific techniques (intelligence gathering and fronting) were mentioned by fewer than four participants. However, there was variation in the use of individual techniques across LMICs and with the PDM. This appeared to reflect the context in which the industry operated, notably levels of corruption and the progress towards tobacco industry denormalisation.

Four findings relating to specific techniques are particularly important. First, reputation management was key in the countries included in this study, while little emphasis was placed on producing misleading evidence-a dominant strategy in contemporary industry activity elsewhere. [41] [42] [43] Our participants described reputation management as far wider-reaching than rehabilitating the industry's image, as emphasised in the original PDM and research on HICs. 44 We found that the tobacco industry's image was still largely positive in the included LMICs and that the industry spent much effort maintaining this reputation through CSR and public relations. While CSR covered a range of areas, a health focus was most frequent, which is consistent with evidence that the tobacco industry uses corporate philanthropy to distract from its harms. 8 The participants also emphasised how successful such strategies were for the industry in times of crisis when resources and capacity are scarce. Given the tobacco industry's intensified CSR efforts during the COVID-19 pandemic this is an important warning to the public health community. 45 Second, direct access to policy-makers was identified as the dominant strategy while less direct approaches, including the use of front groups, were less prominent. This finding contrasts with the increasingly prevalent use of front groups in HICs, where opportunities for direct access appear more limited. 46 47 Third, the industry's close links to, use of, or even control over media content was another crucial activity accented more in this study than in the original PDM. The emphasis was on building long-term relationships with editors and journalists. Recent evidence from HICs suggests that there still remains a need to further tobacco industry denormalisation among journalists and editors. 48 Finally, one technique identified in this study that was not entirely captured by the PDM was the intimidation of advocates. Although forms of intimidation that are included in a PDM technique (discrediting public health organisations and advocates) have been documented in the academic literature, mostly focusing on HICs, [49] [50] [51] [52] the examples found in this study were more informal and covert. This resonates with LMIC-focused newspaper articles on this topic. 53 Given the serious implications of such activities, more research is needed to further explore instances and types of such intimidation.

As the first paper on tobacco industry political activities that draws on interview data from a range of LMICs, its key strength is that it includes voices from a diverse set of countries. These represent four WHO regions, three income groups (low, lower middle, upper middle) and three areas of tobacco control (comprehensive policies, health warnings, taxation). Our interview data offer unique insights into how advocates report and understand industry interference, including the language advocates use. Such knowledge and input from advocates will be essential to maximise user-friendliness and impact of any advocacy tool informed by the study's findings.

A methodological limitation is that we required our participants to speak English in order to avoid working with multiple interpreters. 32 Another limitation is that most participants were not native speakers of English. We mitigated this by refining the interview guide following a pilot interview with a non-native speaker. Despite these limitations, there were no clear gaps compared with the PDM that are not anticipated given the context. Interviewees suspected, for example, informal links or involvement in illicit trade, yet they reported a lack concrete evidence. Similarly, evidence on intelligence gathering would have been hard for participants to identify. The absence of such evidence cannot be taken to indicate that the industry does not employ such techniques but simply that advocates do not have insights into some areas of corporate political activity. To this end, and given that the papers included in the PDM mainly draw on internal tobacco industry documents which do provide such insights, the consistency of our findings with the PDM and the more recent literature on tobacco industry political activities in LMICs 25 26 30 54-58 is remarkable. For instance, existing literature 37 59 presents evidence for industry involvement in illicit trade that was suspected by participants. This consistency also suggests that other potential limitations of qualitative research 60 61 -the potentially limited generalisability of participants' experiences and participants' recall bias-do not apply and affirms the 'fittingness' 62 of our findings.

Given our ultimate aim of using the PDM to develop tools for advocacy and research, the next step is to test the slightly revised taxonomies (online supplemental tables 3 and 4) through in-depth case studies using diverse sources of data. 63 This will ensure that the evidential BMJ Global Health standards on which the original PDM 12 was based on are maintained.

Our findings suggest that advocacy tools need to be flexible, 'living' resources, open to include new and context-specific industry strategies. This is not only in line with the literature on tobacco industry interference [35] [36] [37] [38] [39] but also with research on the corporate political activities of other unhealthy commodities industries. [64] [65] [66] Since the relevance of the PDM to alcohol, food, and gambling has already been identified, 16-22 24 we see great potential for using the model to facilitate cross-industry comparisons.

Looking at practical implications, this research illustrates the informal and covert nature of tobacco industry political activities in LMICs and the frequent lack of transparency in how policy-makers engage with the industry. This underscores the importance of governments introducing mechanisms ensuring transparency in their contacts with the industry, of the continued work exposing tobacco industry misconduct, 55 and pushing for the denormalisation of the tobacco industry in LMICs to ensure that any form of engagement with the industry is deemed unacceptable. 67 Twitter Kathrin Lauber @kathrin_lauber and Mateusz Zatoński @ZatonskiMateusz Acknowledgements The authors would like to thank the participants for their time and trust. We would also like to thank the two anonymous reviewers for their helpful feedback.

Contributors ABG gained project funding and LR gained ethics approval. ABG, LR and BKM conceptualised the project and designed the interview questionnaire. BKM and LR collected the data. BKM coded the data with KL and MZ as cocoders and prepared the draft manuscript. ABG contributed substantially to the writing of the paper. KL, MZ and LR provided feedback during manuscript preparation.

Funding This project was supported by Bloomberg Philanthropies Stopping Tobacco Organisations and Products project funding ( www. bloomberg. org). The opinions expressed are those of the authors alone. The funders had no role in study design, data collection, analysis, decision to publish or preparation of the manuscript.

Patient consent for publication Not required.

Ethics approval Ethical approval for the study was obtained from the University of Bath's Research Ethics Approval Committee for Health (REACH) (Reference: EP 18/19 012).

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http:// creativecommons. org/ licenses/ by-nc/ 4. 0/.

The Lancet Global Health. Stop industry interference, save lives

Alcohol: global health's blind spot

WHO FCTC Impact Assessment Expert Group. Impact assessment of the who framework convention on tobacco control: introduction, general findings and discussion

Impact of implementation of the WHO FCTC on the tobacco industry's behaviour

Addressing NCDs: challenges from industry market promotion and interferences

Interference in public health policy: examples of how the baby food industry uses tobacco industry tactics

Corporate philanthropy, lobbying, and public health policy

The limits of corporate social responsibility: techniques of neutralization, Stakeholder management and political Csr

Open doorway to truth: legacy of the Minnesota tobacco trial

What is known about tobacco industry efforts to influence tobacco Tax? A systematic review of empirical studies

How does the tobacco industry attempt to influence marketing regulations? A systematic review

The policy Dystopia model: an interpretive analysis of tobacco industry political activity

The Policy Dystopia Model": Implications for Health Advocates and Democratic Governance

It will harm business and increase illicit trade': an evaluation of the relevance, quality and transparency of evidence submitted by transnational tobacco companies to the UK consultation on standardised packaging 2012

Factsheet: literature review of tobacco industry responses to tobacco control measures

The policy dystopia model adapted to the food industry: the example of the Nutri-Score SAGA in France

Help or hindrance? the alcohol industry and alcohol control in Portugal

Systematic examination of publicly-available information reveals the diverse and extensive corporate political activity of the food industry in Australia

Following in the footsteps of tobacco and alcohol? Stakeholder discourse in UK newspaper coverage of the soft drinks industry levy

Alcohol industry strategies to influence the reform of the Finnish alcohol law

Applying corporate political activity (CPA) analysis to Australian gambling industry submissions against regulation of television sports betting advertising

Analysis of alcohol industry submissions against marketing regulation

Analysis of the corporate political activity of major food industry actors in Fiji

Mapping of food industry strategies to influence public health policy, research and practice in South Africa

The taste of smoke: tobacco industry strategies to prevent the prohibition of additives in tobacco products in Brazil

Role of stakeholders in Nigeria's tobacco control journey after the FCTC: lessons for tobacco control advocacy in low-income and middle-income countries

A brief overview of the tobacco industry in the last 20 years

Fact sheet: tobacco

Global tobacco industry interference index

Needs of LMIC-based tobacco control advocates to counter tobacco industry policy interference: insights from semi-structured interviews

International public health research involving interpreters: a case study from Bangladesh

What can "thematic analysis" offer health and wellbeing researchers?

Poisonous pandas. Redwood City

A Massive Spin-Off Effect in West Africa and Beyond": The Tobacco Industry Stymies Tobacco Control in Nigeria

Moving East: how the transnational tobacco industry gained entry to the emerging markets of the former Soviet Union-part I: establishing cigarette imports

Balkan smoke: tobacco and the making of modern Bulgaria

Tobacco industry efforts to keep cigarettes affordable: a case study from Hungary

Smoking and social justice

Representation and misrepresentation of scientific evidence in contemporary tobacco regulation: a review of tobacco industry submissions to the UK government consultation on standardised packaging

Inventing conflicts of interest: a history of tobacco industry tactics

Tobacco industry manipulation of research

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

The two faces of the tobacco industry during the COVID-19 pandemic

The revision of the 2014 European tobacco products directive: an analysis of the tobacco industry's attempts to 'break the health silo

A critical evaluation of the volume, relevance and quality of evidence submitted by the tobacco industry to oppose standardised packaging of tobacco products

Lethal news: the dexterous infiltration of news media by the tobacco industry agenda

Asia is now the priority target for the world anti-tobacco movement": attempts by the tobacco industry to undermine the Asian anti-smoking movement

Tobacco control policy making in North Dakota: a tradition of activism

Consent and counter-mobilization: the case of the National smokers alliance

Structural elements in achieving legislative tobacco control in NSW, 1955-95: political reflections and implications

In poor countries, antismoking activists face threats and violence

A systematic review of tobacco industry tactics in Southeast Asia: lessons for other low-and MiddleIncome regions

Exposing and addressing tobacco industry conduct in low-income and middle-income countries

Exceeding who framework convention on tobacco control (FCTC) obligations: Nepal overcoming tobacco industry interference to Enact a comprehensive tobacco control policy

Tobacco control in Namibia: the importance of government capacity, media coverage and industry interference

The vector of the tobacco epidemic: tobacco industry practices in low and middle-income countries

Complicity in contraband: British American tobacco and cigarette smuggling in Asia

Generalizability in qualitative research: a tale of two traditions

Recall bias: a proposal for assessment and control

Effective evaluation: improving the usefulness of evaluation results through responsive and naturalistic approaches

The use of triangulation for completeness purposes

The perils of ignoring history: big tobacco played dirty and millions died. How similar is big food?

Reassessing policy paradigms: a comparison of the global tobacco and alcohol industries

Governing drug reimbursement policy in Poland: the role of the state, civil society, and the private sector

Tobacco industry denormalisation as a tobacco control intervention: a review

Alphabetical list of who member states

World bank country and lending groups

Tobacco control laws: legislation

Taxing tobacco: a win-win for public health outcomes and mobilizing domestic resources

Author note Tobacco Control Research Group @BathTR ORCID iDs Britta Katharina Matthes http:// orcid. org/ 0000-0003-2346-066X Kathrin Lauber http:// orcid. org/ 0000-0003-0073-3004 Mateusz Zatoński http:// orcid. org/ 0000-0001-7254-569X Anna B Gilmore http:// orcid. org/ 0000-0003-0281-1248