Tobacco Industry Data on Illicit Tobacco Trade
Total Page:16
File Type:pdf, Size:1020Kb
Review Tob Control: first published as 10.1136/tobaccocontrol-2018-054295 on 22 August 2018. Downloaded from Tobacco industry data on illicit tobacco trade: a systematic review of existing assessments Allen W A Gallagher, Karen A Evans-Reeves, Jenny L Hatchard, Anna B Gilmore ► Additional material is ABStract global accountancy firms such as KPMG, Deloitte published online only. To view, Objective To examine the quality of tobacco industry- and PricewaterhouseCoopers.25 28 32 At least one of please visit the journal online (http:// dx. doi. org/ 10. 1136/ funded data on the illicit tobacco trade (ITT) through a these firms has expressed concern that the TTCs tobaccocontrol- 2018- 054295). systematic review of existing assessments of industry- have used their research findings in a misleading funded data on ITT. manner.36 TTCs use such self-funded data and Tobacco Control Research Data sources Papers and reports assessing tobacco the threat of the ITT in efforts to oppose tobacco Group, University of Bath and industry-funded data on ITT were obtained via searches control policies,1 37 38 arguing that tobacco control UK Centre of Tobacco and Alcohol Studes, Bath, UK of 8 academic databases, Google searches and measures will increase ITT and its associated crim- correspondence with ITT experts. inality.1 37–44 Correspondence to Study selection Inclusion criteria identified 35 English- As a consequence of TTCs' use of self-funded Mr. Allen W A Gallagher, language papers containing an original assessment of data, a growing number of independent studies have Department for Health, tobacco industry-funded data. scrutinised the quality of these data in Australia,45 University of Bath, Bath, BA2 Data extraction Using a coding framework, Europe,46 Asia47 and South Africa,48 levelling a 7AY, UK; 31 46 a. a. gallagher@ bath. ac. uk information was extracted from the assessments number of criticisms against them. To date, regarding the quality of tobacco industry data. there has been no attempt to systematically Received 6 February 2018 Documents were second-coded, achieving 94% summarise this literature. Undertaking these assess- Revised 16 April 2018 intercoder reliability with all disagreements resolved. ments is expensive, time consuming and difficult Accepted 25 April 2018 Published Online First Data synthesis Of the 35 assessments reviewed, 31 to achieve quickly enough to be useful within the 22 August 2018 argued that tobacco industry estimates were higher than rapidly moving policy cycle. independent estimates. Criticisms identified problems This paper therefore aims to systematically review with data collection (29), analytical methods (22) and existing studies, which assess tobacco industry presentation of results (21), which resulted in inflated data on ITT (hereafter ‘assessments’) to provide copyright. ITT estimates or data on ITT that were presented in a substantive overview of the characteristics of a misleading manner. Lack of transparency from data such data and to identify the nature of critiques of collection right through to presentation of findings tobacco industry data/reports on ITT. By compiling was a key issue with insufficient information to allow this information, this review will aid public health replication of the findings frequently cited. responses to any future data on this topic. Conclusions Tobacco industry data on ITT are not Growing tobacco industry funding of research reliable. At present, the tobacco industry continues to on ITT underlines the importance of such work. fund and disseminate ITT research through initiatives Philip Morris International’s (PMI) latest initiative, http://tobaccocontrol.bmj.com/ such as PMI IMPACT. If industry data on ITT cannot meet PMI IMPACT, has pledged US$100 million to fund the standards of accuracy and transparency set by high- research on illegal trade and related crimes and, quality research publications, a solution may be to tax as of early 2017, had committed US$28 million to tobacco companies and administer the resulting funds 32 projects across the European Union (EU).49 50 to experts, independent of the tobacco industry, who use Outlining the findings of existing assessments of previously developed reliable models for measuring ITT. past data funded by the tobacco industry is a useful and necessary step towards better understanding future data and how to respond to it. INTRODUCTION The illicit tobacco trade (ITT) is difficult to measure METHODS on September 28, 2021 by guest. Protected due to its illegality,1–3 its global4 and changing Search nature2 5 and data collection and analysis complexi- To identify existing independent assessments ties.2 3 6 7 While methods such as Empty Pack Surveys of TTC-funded data or reports on the ITT, the (EPS), consumer surveys, econometric modelling following search string was applied to eight data- and tax gap approaches have been used effectively bases (Business Source Complete, Embase, the by multiple non-industry sources;3 8–15globally, International Bibliography of Social Sciences, Ovid, there is no agreed ‘gold-standard’ methodology PubMed and PubMed Central, ScienceDirect and for estimating ITT16 and estimates vary greatly in Web of Science): rigour and approach.17 (("Philip Morris" OR "PMI" OR "British Amer- In recent years, transnational tobacco companies ican Tobacco" OR "BAT" OR "Imperial Tobacco" (TTCs) (box 1) have been a major funding source of OR "Imperial Brands" OR "Imperial" OR "ITG" OR To cite: Gallagher AWA, 18–28 Evans-Reeves KA, data on ITT. They heavily publicise these data, "Japan Tobacco" OR "JTI" OR "Tobacco company" Hatchard JL, et al. especially when a tobacco control policy is being OR ‘"transnational tobacco company" OR "TTC" Tob Control debated.1 The tobacco industry has commissioned OR "TTCs") AND ("research" OR "evidence" OR 2019;28:334–345. reports on ITT,19 25–27 29–35 often produced by data* OR "study" OR "studies" OR report*) AND 334 Gallagher AWA, et al. Tob Control 2019;28:334–345. doi:10.1136/tobaccocontrol-2018-054295 Review Tob Control: first published as 10.1136/tobaccocontrol-2018-054295 on 22 August 2018. Downloaded from Box 1 Important terminology Box 2 Inclusion and exclusion criteria and key definitions. Transnational tobacco company/companies Inclusion criteria Transnational tobacco company/companies (TTCs), the major This review aimed to identify documents that assess tobacco four currently being British American Tobacco,105 Imperial industry-funded data on illicit tobacco trade (ITT) (assessments) Tobacco,106 Japan Tobacco International107 and Philip Morris and was conducted in two stages: International.108 1. Title/abstract screening: – Document must be written in English. Counterfeit – Document must include data on ITT (a key term search Products bearing a trademark of a cigarette manufacturer of the document was conducted when this could not be that are manufactured by a third party without the cigarette 46 determined from the title or abstract). manufacturer’s consent. 2. Full-text screening: Tobacco industry illicit – Document must not have received funding from the Tobacco company product that was en route to, imported tobacco industry. into, distributed in or sold in a jurisdiction in violation of the – Document must assess data on ITT that has received applicable fiscal laws of that jurisdiction,46 for example, tobacco funding from the tobacco industry. industry product present in the illicit market. The fact that this – Document must clearly identify the data that are being product was manufactured by the TTCs does not imply they assessed, eg, the source of the data has to be identifiable are always responsible when that product ends up on the illicit from the contents of the document. market. Key definitions Contraband ‘Assess’=to provide an evaluation of tobacco industry data Any tobacco product (including counterfeit and tobacco industry) on illicit trade. This could be a positive or negative statement imported in a jurisdiction in violation of the applicable fiscal laws regarding any element of the data such as how it was collected, of that jurisdiction.55 analysed, presented, etc referring to or citing data without providing any critical comment on it was not considered an Non-domestic assessment of that data. Solely referring to pre-existing critiques Tobacco products brought in from an overseas market. This can of data was also not considered an assessment. include overseas purchases that were then transported legally ‘Industry-funded data on illicit trade’=any data on illicit (legal non-domestic), as well as contraband products (illicit/ tobacco that has been funded fully, or in part, by tobacco copyright. 109 illegal non-domestic). companies including industry-commissioned research and research conducted by those that receive industry funding. This includes data that transnational tobacco companies claim as (illicit* OR illegal* OR smuggl* OR "contraband" OR counter- their own or have commissioned, as well as data featured in a feit*) AND ("tobacco" OR cigar*)). newspaper, website, public event or advertising campaign that Minor variations were made in order to identify the most comes from an industry source. effective search for each database. Additional searches were ‘Source’=where the assessed data were taken from. Sources conducted in the specialist peer-reviewed journals Addiction, may include industry-commissioned reports, internal industry http://tobaccocontrol.bmj.com/ Health Economics and Tobacco Control to ensure that poten- documents, industry