Article Text
PDF
Abstract
The ‘common knowledge’ defence is a legal strategy which has been successfully used by defendant tobacco companies to avoid legal responsibility for the harms caused by smoking. Tobacco companies have hired professional historians to try to persuade courts about a longstanding high level of public awareness regarding the risks of tobacco use. To support this argument, they have used archival news clippings and media reports. Two historians were hired by tobacco companies to offer this defence during a recent class action trial in Canada, following which they were required to submit to the court the collection of media materials which had been gathered by history students to assist their testimony. Included in this collection were tobacco advertisements and other news items about tobacco products which the students had inadvertently also collected. Quantifying this collection reveals that even by the tobacco industry's own construct, the information environment surrounding Quebec smokers in the middle 20th century included more prosmoking messages than information about the risks of smoking.
- Advertising and Promotion
- Litigation
- Media
- Tobacco industry documents
Statistics from Altmetric.com
Introduction
The engagement of 50 historians as expert witnesses in more than 314 US tobacco trials was recently described by Delafontaine, building on an earlier analysis by Kyriakoudes and Proctor.1–4 These researchers report that historians hired by defendant tobacco companies frequently testify that the risks of smoking were ‘common knowledge’ even before governments required that warning labels appear on cigarette packages. Archival news reports citing scientific discoveries about the link between smoking and disease are often used as the evidentiary basis for this assertion.
The use of the common knowledge defence strategy in tobacco lawsuits outside the USA has not been as thoroughly examined, perhaps because these suits are less common elsewhere.5 Nonetheless, additional examples of historical evidence being cited as grounds to dismiss claims against cigarette manufacturers can be found in Finland,6 the UK7 and in New Zealand.8
A further example of this defence being attempted is found in the recent class action suits in Quebec, Canada. In this case, the historians testifying on behalf of the defendant companies were required to make public the archival news reports on which they had based their opinion. In addition to news stories about smoking and health, this collection also included tobacco advertising and news stories about tobacco that were unrelated to health and which had been inadvertently gathered by history students assigned to review newspaper archives. This accidental collection provides an opportunity to assess the broader public information environment facing some North American smokers in the mid-20th century.
Historians at the Montreal tobacco trials
Two tobacco class actions were filed in Quebec in 1998: the first was a claim for $5000 on behalf of each of the estimated one million smokers who had become addicted,9 the second was a claim for $100 000 on behalf of the estimated 100 000 smokers who had become ill with lung cancer, throat cancer or emphysema.10 Between March 2012 and December 2014, these suits were tried simultaneously before Justice Brian Riordan of the Quebec Superior Court.
The defendant companies were the Canadian affiliates of the world’s three largest multinational tobacco companies: Imperial Tobacco (wholly owned by British American Tobacco), Rothmans, Benson & Hedges (wholly owned by Philip Morris International) and JTI-Macdonald (wholly owned by Japan Tobacco International). Imperial Tobacco Group PLC, based in the UK, is not related to the similarly-named Canadian company and does not sell manufactured cigarettes in Canada.
The lawsuits accused tobacco companies of having failed to warn consumers of the inherent dangers of cigarette smoking and of having trivialised those dangers through advertising and public relations efforts.11 Evidence to support these accusations was limited to the class period, which began in 1950 and which ended in late 1998, when the suits were first filed.
In their defence, tobacco companies argued that no harm experienced by any individual smoker could be demonstrably linked to any wrongful act on their part and that they had no duty to warn clients who were already aware that smoking could cause illness because this information was ‘common knowledge’ at the time they started smoking.
Their key witnesses, and the first ones they presented to court, were those hired to support their claim that : “At all material times, members of the class were aware of or could have been aware that there were risks associated with smoking, including the risk that it may be difficult to quit smoking."12
They began their defence by presenting to the court two historians who gave expert testimony in support of this view. One was a retired (emeritus) professor from the University of Western Ontario, Flaherty,13 and the other was a well-known Quebec popular historian, Lacoursière.14 A third historian to testify on their behalf was Robert Perrins, whose testimony did not address ‘common knowledge’ but was linked to the role of government.15 Duch16 was another witness who spoke on behalf of the companies about ‘common knowledge’, but his testimony was based on public polling data.
In several respects, the testimony of Flaherty and Lacoursièrere followed the pattern of historian testimony in US trials as described by Kyriakoudes.1 The Canadian historians, for example, were recruited without previous professional experience in tobacco history. One (Flaherty) had worked with the US tobacco industry's Special Trials Issue Committee. Both men used material provided to them by the companies, and both failed to look at internal industry documents. They both based their evidence primarily on newspaper clippings, and both declined to consider the impact on public knowledge about tobacco use created by other information in the print media, such as cigarette advertising or by industry denials of a link between smoking and disease.17 ,18
The collection of newspaper and media clippings which became the reliance material for their testimony had been gathered from newspapers, magazines and other material circulated in Quebec during the class period. Flaherty had first been engaged by the companies in 1988 to conduct ‘historical awareness research’ and had supervised the collection of material for many of the intervening years.19 Several historians assisted in the collection efforts, including 19 students at the Université du Québec à Montréal, who had been hired in 1999 to review the microfiche records of the four major francophone daily newspapers published in Quebec. The papers reviewed in this systematic way were La Presse (for the years 1950–1991), Le Devoir (1950–1991), Le Soleil (1950–1991) and Le Journal de Montréal (from its launch in June 1964 to 1999).
Flaherty17 testified that the “students were instructed not to collect tobacco advertising”, but they do not appear to have followed any such direction. The research tracking sheets used by each student, which were also filed with the court, suggest that these researcher assistants attempted to gather all material that contained any reference to tobacco and its use. In addition to stories about health effects, they also copied advertisements for tobacco products, business stories, letters to the editor, human interest and entertainment stories, comics, editorial cartoons and even classified ads. That is to say, they gathered news stories reporting on the risks of smoking and also collected other published material which could have shaped public understanding of these risks.
The rules of court as applied to this trial required that all of the material reviewed by these experts be filed with the court as trial exhibits, making this collection available for independent review and further analysis. A large subset of the material given to the court were the clippings collected by the student researchers along with hundreds of tracking sheets which allow confirmation of the completeness of court records. The immediate usefulness of this set of materials was hindered by the absence of an index at the time it was filed with the court. The analysis reported here is the result of the indexing and classification of more than 20 000 records which was performed following the trial's conclusion.
Methods
We began this analysis by creating a consolidated index of the 35 500 electronic records that were included in the reliance material of the two ‘common knowledge’ defence historians. Discounting duplicates, tracking sheets, file folders and cover sheets reduced this collection to about 15 700 unique newspaper items. Among these were 14 240 clippings from the four Francophone daily newspapers, most of which (13 169) were accompanied by a tracking sheet.
We subsequently conducted a subject-area analysis on the subset of 9765 newspaper items which met the criteria of having tracking sheets available and which dated from 1950 to 1989. We chose this timeframe as these were the decades before tobacco advertising was banned in Canada and before health warnings were mandated by federal law.20 These records represented 62% of the total collection, as shown in table 1. Nineteen categories of newspaper content were established for this purpose, and they are described in table 2.
Consolidated reliance material submitted by historians testifying for defendant tobacco companies in Montreal
Categories of newspaper items
In classifying these records, we assigned each item to only one category. To ensure the widest capture of all stories related to health, any reference to health resulted in categorisation as a health story. For example, news stories about tax rates which made reference to the health rationale for taxes were grouped among health stories. The only exceptions to this hierarchical protocol was tobacco advertising which was classified with other ads even if it contained references to health.
We further screened health stories for statements which denied or cast doubt on a causal relationship between smoking and any disease, including addiction, or which presented other factors, such as air pollution, as a greater contributor to lung cancer, heart or lung disease.
Results
One third (3521) of the newspaper items were branded tobacco advertisements. Both the absolute and proportional number of such ads declined over the period, and there were none after 31 December 1988 when the federal law banning direct tobacco advertising came into effect.20
The second largest category of material was the 2830 newspaper items (29%) which discussed or alluded to the health consequences of smoking. Not all of these stories clearly communicated that smoking was risky. Included in this category were newspaper items which would likely have had a more nuanced influence on public understanding of risks, such as 46 reports on cigarettes becoming safer (as a result of filters, tobacco blends, etc) and 144 stories which included denials of the health consequences of smoking or otherwise cast doubt on them. The number of health-related items increased over the 40-year period, but it was only after 1985 that these articles constituted a majority of tobacco-related items in newspapers. It was only after 1980 that the average number of such items exceeded two stories per paper each month.
Reinforcement of other negative aspects of smoking were found in an additional 748 newspaper items, including 565 reports about restrictions on smoking in public places, 125 advertisements for cessation products and 58 stories about cigarette-caused fires.
There remain 2666 newspaper items which do not fit into the above categories. The most numerous of these were 426 news stories about business issues, including plant openings, share purchase opportunities or dividend payments to shareholders. Other large categories were stories about cigarette related crime (such as theft and contraband), government tax policy, or advertisements for other products or services which displayed tobacco products, including tobacco retailing. The content of editorial columns, opinion pieces and letters to the editor (totalling 651) was not further analysed to establish whether it would or would not have assisted the public in understanding the risks of tobacco use.
Flaherty describes newspaper reports about the health risks of tobacco use and smoke-free places as ‘anti-smoking stories’.13 By this description, there were about the same number of ‘anti-smoking’ stories in the collection as there were advertisements for branded tobacco products, which can by the same logic be viewed as ‘pro-smoking’ (37% vs 36%). Three-quarters of the items collected can be viewed as likely to influence ‘common knowledge’ in one direction or the other using this tobacco industry construct of the information environment. (In the absence of a further content analysis, the potential effect of one-quarter of the collection material stories in the collection is unknown.)
These results are shown in table 3 and in figure 1.
Number of newspaper items per category in four Quebec French-language newspapers, 1950–1989
Number of newspaper items per category in four Quebec daily French-language newspapers, 1950–1989.
Limitations
There are several limitations to this effort that might temper any conclusions that might be drawn from it. Some of these relate to the content analysis we conducted on the tobacco industry's collection of media clippings. Others relate to the restricted lens applied to the industry's construct of ‘information environment’.
This study is believed to be the first quantitative analysis of reliance material used by historians testifying in a tobacco lawsuit. As a result, the classification of trial documents does not follow any previously validated method. In addition, the classification exercise as it was conducted was inherently and unavoidably subjective.
This study attempted only to identify categories of media stories and did not attempt to execute a more in depth content analysis. Factors which could influence the impact of a newspaper item on the information environment were not reviewed, such as the prominence of the article, the circulation of the paper, authorship, placement or length. Nor was there an attempt to establish more nuanced categories, as might be warranted for stories about technological changes to cigarettes (filters, new tobacco types) that might have communicated that the harms of smoking could be reduced.
Comprehensive clippings were available only for French language newspapers and not for the two English daily newspapers published in Quebec during that period. The information environment created by print media may have been different for Anglophone smokers in Quebec during that period as it could have been for consumers of other news media, like radio or television.
Human error is a further factor to consider. The students who reviewed the newspaper files may have overlooked some records. The indexing and classification reported here was conducted by one individual and undoubtedly contains both clerical and analytic errors.
A more substantive limitation on the use of the findings is the risk that they are grounded in the ‘information environment’ construct developed by tobacco industry lawyers. This analysis shows that tobacco industry witnesses who testified to the ‘deluge’ of anti-smoking messages sidestepped an even greater flood of prosmoking messages and challenges their conclusions on the basis of their own logic. The results do not overcome the faulty logic and narrow focus of the construct, as detailed by Proctor and Kyriakoudes,2–4 and do not shed light on other missing factors that might have retarded or accelerated public knowledge of the harmfulness of smoking.
Discussion
By opening a window on part of the information environment surrounding Francophone Quebec smokers in the middle of the 20th century, the testimony of defence historians had the unintended consequence of facilitating further research into tobacco use and tobacco control and the factors which influenced them. The records from this trial are now on the Industry Documents Digital Library of the University of California at San Francisco,21 and the index is also available for public use.22
The initial efforts of this analysis to measure the types of information which were available and may have influenced public understanding of the harms of tobacco use can help contextualise the statements made to courts by historians in support of the defence of common knowledge. It complements the testimony of Canadian historians who excluded the majority of tobacco-related newspaper content, including advertisements, in their presentations to a Quebec court and challenges their claims that the media carried a ‘deluge of anti-smoking material’ or that the ‘industry's message was drowned out by the overwhelming flood of anti-smoking stories’.13
The 2013 testimony by Flaherty and Lacoursière in the Montreal tobacco trials was the first time that the common knowledge defence was used in a tobacco trial in Canada. There are other lawsuits facing tobacco companies in Canada, including attempts by each of 10 Canadian provinces to recover the healthcare costs associated with treating tobacco-related disease.23 Should these suits ever come to trial, the defence legal teams may hesitate to repeat this testimony given the hostile response it was given by the judge in the Quebec class actions.
In ruling in favour of the plaintiffs, Justice Riordan of the Quebec Superior Court essentially disqualified the common knowledge defence and the historians who presented it. Their failure to include advertising in their reviews and their use of media attention as a measure of common knowledge were among the reasons the judge gave for his decision to not ‘give any credence’ to the views of these historians about when Quebec smokers might be expected to have been informed about the risks of smoking.[I]t seems inconsistent, to say the least, that these experts should be so chary to opine on the effect of newspaper and magazine ads on people's perception when they have absolutely no hesitation with respect to the effect of articles and editorial cartoons in the very same newspapers and magazines in which those ads appeared. They seem to have been tracing their opinions with a scalpel in order to justify sidestepping such an obviously important factor. In doing so, they not only deprive the Court of potentially valuable assistance in its quest to ascertain one of the key facts in the case, but they also seriously damage their credibility.
The judge not only rejected the testimony of the defence historians, he used the testimony offered by the plaintiffs’ expert historian, Robert Proctor, to rule that the public should have known about the risk of smoking by January 1980. He also rejected the philosophical underpinning of the common knowledge defence:Professor Flaherty talks of “common knowledge,” but all either he or Professor Lacoursière is showing is the level of media attention given to the issue. That is not knowledge. That is exposure. On that basis, how can they opine on anything more than surveying what was published and publicly available?
The industry defence suffered a further injury when the judge used the common knowledge argument against them, saying that it was an admission by the companies that they too would have been aware of the risks of smoking while they were selling these products without warnings.24
Tobacco companies in Canada have to date failed in their attempt to use the common knowledge argument to protect them from being held liable for harming smokers. A consequence of their first attempt to do so was the public release of a large archival collection. On the basis of this initial study, we might expect that additional analysis of this material will further weaken the ability of the companies to convince courts that the information environment surrounding smokers in the mid to late 20th century was one which clearly and unambiguously communicated that smoking was harmful.
What this paper adds
Tobacco companies engaged the common knowledge defence strategy during a recent Canadian tobacco trial. The material they assembled to support this argument inadvertently established a comprehensive archival record of newspaper coverage of tobacco issues for at least four decades, including records of tobacco advertisements and denialist statements issued by tobacco industry representatives. A preliminary analysis of the archive shows that the public was exposed to more prosmoking than anti-smoking messages in the newspapers surveyed until at least the 1980s.
Footnotes
Contributors CC was responsible for the study design and report. The collection of news clippings on which this study was based was conducted by history students at the Université du Québec à Montréal in 1999.
Funding This research was conducted as part of a project supported by the Quebec Ministry of Health and Social Services.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The records analysed for this report are all available, as is the classification conducted in this review.