LL2-07 hindsight check: Tobacco industry manipulation of research (Bero; Panel 7.1 by Smith, Gilmore and Fooks), Late lessons from early warnings: science, precaution, innovation (EEA Report No 1/2013), Ch. 7, pp. 151–178#

Checked 25 September 2026. The check covers what happened between publication (January 2013) and September 2026 that bears on the chapter’s claims, evidence, predictions and recommendations. Page numbers are report pages (PDF page minus 2).

The chapter is an expansion of Bero (2005) and rests mainly on the author’s own meta-research and on internal documents released through US litigation (p. 152). Its core evidence dates from 1993–2005. Bero has since continued this research programme (industry sponsorship, conflicts of interest, industry documents), and Panel 7.1’s authors (Smith, Gilmore, Fooks) have extended their EU work. Much of the later literature therefore comes from the same research community. I have tried to separate independent corroboration (for example Cochrane methodology reviews, courts, regulators) from work by the original authors and their close collaborators.

Annex 3 note. Not applicable. The chapter is new in the 2013 report; the 2001 volume had no tobacco case, so there is no Annex 3 update.

Access note. The session’s web-search quota was exhausted before this check began. Everything below was verified by direct retrieval: - scholarly records and abstracts: Europe PMC’s REST service, Crossref and OpenAlex; - US documents: the Federal Register API (rules, executive orders) and CourtListener (court opinions); - EU texts: the EU Publications Office (cellar); - organisations’ own pages: fetched directly (CDC, WHO, IARC, CMS, ICMJE, FDA, European Commission, ERIF, LobbyFacts, STOP, ANRF).

Some sources were blocked and were not bypassed: - NCBI Bookshelf (reCAPTCHA), so the 2006 and 2014 Surgeon General reports could not be opened directly; - HHS and CDC Stacks (403); - BMJ, ScienceDirect and Nature (403 or login).

Items that rest on a title only, or on facts I could not re-fetch, are flagged.

Overview#

This chapter is not a conventional early-warning case. It is a documented account of how one industry manufactured controversy about second-hand smoke (SHS) to delay regulation, plus recommendations on disclosure and funding bans. Thirteen years on, the record splits as follows.

Vindicated or strengthened - The harm itself. Second-hand smoke causes lung cancer (IARC Group 1, reaffirmed 2012). The cardiovascular case strengthened: - stroke was added as a causal outcome (US Surgeon General 2014 [not re-fetched]; CDC: 20–30% higher stroke risk); - a Cochrane review (2016, 77 studies) found smoke-free laws consistently reduce acute coronary admissions; - WHO now estimates SHS “kills more than 1.6 million people prematurely every year”, up from about 603,000 in 2004 (Öberg et al., 2011). Much of the rise reflects changed methods rather than a trend. - The fraud finding. In US litigation, a federal court (2006, affirmed 2009) found that the major manufacturers had for decades conspired to deny the health effects of smoking. It ordered corrective statements on five topics, one of them “the adverse health effects of exposure to secondhand smoke”. These ran in US newspapers and on television from 2017–18, and at point of sale by 2024. - Sponsorship bias (claim 2). Corroborated in direction by independent methodology reviews: - Cochrane 2017 (drugs and devices): risk ratio 1.34 for favourable conclusions, with a bias “that cannot be explained by standard ‘Risk of bias’ assessments”; - Nejstgaard et al., BMJ 2020; - a blinded study of e-cigarette papers (2019) found odds ratios of about 67–92 for “no harm” findings among industry-conflicted studies. That is the same order of magnitude as the chapter’s OR 88.4 for passive-smoking reviews. - Constancy of tactics. The claim extends past the 1990s. Document studies of the PMI-funded Foundation for a Smoke-Free World (FSFW, 2017–2023) and of leaked Philip Morris Japan documents (2012–2020) found covert research funding, laundered credibility and “vehicle reinvention” of the kind the chapter describes. - Cross-sector reach (claim 1). Much better documented than in 2013: - sugar industry: Kearns et al., 2016 and 2017; - PFAS manufacturers: Gaber, Bero and Woodruff, 2023; - fossil fuels: Supran and Oreskes, 2017 and 2023; - a scoping synthesis across eight sectors (Legg et al., 2021) that found “remarkable consistency” in macro-strategies.

The author’s caveat that more document evidence was needed has been partly met. - “Disclosure is necessary but not sufficient” (claim 8). Strongly supported. Disclosure spread, through the ICMJE form (updated 2021 and 2026) and the US Open Payments database (US$14.67 billion in program year 2025). But: - a meta-analysis of 40 studies found 81% of articles had discrepancies between authors’ self-disclosures and payment records; - a 2025 scoping review (Bero co-author) concluded disclosure policies “are not particularly effective in mitigating risk of bias”.

Contested, weakened or mixed - Breast cancer from SHS (claim 3) is still contested. The chapter reports it on the strength of California EPA (2005). - IARC (2012) called the evidence “inconsistent”, and the US Surgeon General (2006) “suggestive but not sufficient”. - Later meta-analyses split by design: pooled RR about 1.24 when case-control studies dominate, about 1.01–1.07 in prospective cohorts, and null in a nine-cohort Japanese pooled analysis (2024). - Sponsorship bias is independent of quality (claim 2). Only partly true in later syntheses: - industry-sponsored drug trials had similar or better risk of bias (Cochrane 2017); - but systematic reviews with financial conflicts tend to have lower methodological quality (Cochrane 2019); - food-industry associations are weaker or non-significant in some syntheses (Chartres et al., 2016, 2020).

The OR of 88.4 is an imprecise outlier, not a general effect size. - The effect of the “sound science” riders (claim 4). The mechanism the chapter predicted recurred. Its realised effect is less clear than its recurrence: - EPA’s 2018–21 “transparency” rule was vacated within weeks of finalisation. The court held it “determined outcomes rather than process”. - Executive Order 14303 (May 2025) now routes scientific-integrity “corrections” through a political appointee under the Information Quality Act. It limits “overly precautionary assumptions” and protects “confidential business information”. Its effects are not yet measurable. - Panel 7.1’s claim that BAT “contributed to” EU Better Regulation (claim 5) gained support: - document-and-interview studies (Smith et al., 2015; Peeters et al., 2016; Costa et al., 2014); - the Forum’s own claim to have “successfully championed” the Commission’s impact-assessment system, the Regulatory Scrutiny Board and the “innovation principle”, which entered EU law in Horizon Europe (2021, recital 6).

But Better Regulation had many drivers. The EU still adopted and defended strong tobacco law (Tobacco Products Directive 2014, upheld by the CJEU in 2016 partly on the strength of the Commission’s own impact assessment). Precaution remains in the Treaty. The panel’s own verdict, “not yet possible to assess” the campaigns’ success, remains about right.

What the chapter could not see or did not cover - The WHO Framework Convention on Tobacco Control, Article 5.3, adopted 2003 with guidelines in 2008, is omitted. It is the main institutional embodiment of the chapter’s lesson, but only for tobacco. Advocacy monitoring (STOP, 2025) reports that interference persists. - The EU Transparency Regulation (EU) 2019/1381 moved in the opposite direction from the US riders. It applies from 27 March 2021 and: - requires business operators to notify EFSA of the studies they commission (an anti-suppression register); - requires EFSA to publish industry studies “in a proactive manner”. - The harm-reduction era changed the terrain. Industry now funds and publishes peer-reviewed science on its own “reduced-risk” products, and regulators must rely partly on manufacturer data. That makes blanket funding bans harder to apply and is contested (Polosa and Crawley, 2017).

Weight for the lens. High confidence for: - the documented intent and strategies; - the direction of sponsorship bias; - the claim that procedural and evidentiary rules are battlegrounds; - the claim that disclosure alone is insufficient.

Moderate confidence for: - cross-sector generalisation, which is well evidenced for the existence of the playbook but subject to observability bias from litigation; - the causal size of delay attributable to industry tactics.

Low confidence for: - the breast-cancer claim; - the specific magnitude (OR 88.4); - treating the five diagnostic questions as a validated tool. They are plausible but untested.

Claim 1. Tobacco-industry strategies (eight listed in Box 7.1) were “remarkably constant” from the 1950s to the 1990s, and are shared by other corporate sectors: pharmaceuticals, food, chemicals, lead, asbestos and fossil fuels (pp. 152, 155, 167–169)#

Original claim. - The strategies “have remained remarkably constant since the early 1950s” (pp. 152, 154). They were controlled by lawyers and executives “rather than scientists” (p. 152). - The eight strategies (Box 7.1, restated p. 169) are: fund, hide, publish, suppress, criticise, change standards, lay press, policymakers. - They are “similar to those of other corporate interests” (p. 167), with examples from: - pharmaceutical and food funding studies; - ExxonMobil, via the Union of Concerned Scientists (2007) (p. 168); - Markowitz and Rosner on lead, asbestos and vinyl chloride (p. 169). - But the chapter itself cautions that “the public health community must learn more about the internal behaviour of corporations other than the tobacco industry in order to make conclusions about similarities”, since document insight “is not available for most corporate sectors” (pp. 168–169).

What happened since.

Constancy within tobacco, extended beyond the 1990s. - Foundation for a Smoke-Free World. PMI launched FSFW in September 2017 with a pledged US$1 billion over 12 years (van der Eijk, Bero and Malone, Tobacco Control, 2019). - Legg, Clift and Gilmore (Tobacco Control, 2024) tracked it prospectively for 2017–2021. They found “marked similarities” with earlier industry practice: producing industry-friendly research, “obscuring industry involvement in science”, funding third parties that denigrate unfavourable science, and promoting industry credibility. - They call FSFW “a new vehicle for agnogenesis”, the deliberate production of ignorance. - Philip Morris Japan. Braznell et al. (Nicotine & Tobacco Research, 2025) analysed 24 leaked documents from 2012–2020. They found that Philip Morris Japan: - covertly funded a Kyoto University smoking-cessation study through a contract research organisation, with no public record of its role; - paid a consultancy about ¥3 million a month (2014–2019) for work that was “hidden internally and externally”.

This mirrors the chapter’s “hide” and “fund” strategies (pp. 155–158) and its account of reinventing vehicles once exposed (p. 157). - What changed. The object changed. Industry science now largely promotes new nicotine products rather than denying harm from cigarettes, which Legg 2024 and Braznell 2025 both treat as continuity of method with a new aim.

Cross-sector document evidence (the author’s own caveat, p. 168). - Legg, Hatchard and Gilmore (PLoS One, June 2021), the “Science for Profit Model”, is a scoping synthesis of 68 sources. - It found eight sectors repeatedly influencing science: alcohol; chemicals and manufacturing; extractive; food and drink; fossil fuels; gambling; pharmaceuticals and medical technologies; tobacco. - It reported “remarkable consistency” in macro-strategies. All eight sectors used four of five macro-strategies, and five of the eight used the fifth (“manufacture trust”). - Its own limitations: strategies are best documented where litigation released documents (tobacco, chemicals, pharmaceuticals); covert strategies are under-detected; some meso-strategies appear in only four of eight industries. Tobacco was the only sector found to use all 19 meso-strategies. - Sugar. Kearns, Schmidt and Glantz (JAMA Internal Medicine, Sept 2016) showed that the Sugar Research Foundation secretly funded a 1967 NEJM review that downplayed sucrose and blamed fat. It “set the review’s objective, contributed articles for inclusion, and received drafts”, and its role “was not disclosed”. - Kearns, Apollonio and Glantz (PLoS Biology, Nov 2017) showed the Foundation terminated an animal study (“Project 259”) without publishing results suggestive of harm. - These map onto “fund”, “hide” and “suppress”. - PFAS. Gaber, Bero and Woodruff (Annals of Global Health, June 2023) found in the UCSF Chemical Industry Documents archive that manufacturers knew PFAS were “highly toxic when inhaled and moderately toxic when ingested” by 1970. They “suppress[ed] unfavorable research and distort[ed] public discourse”. - Importantly, they “did not find evidence in this archive of funding favorable research or targeted dissemination of those results”. Not all eight strategies appear in every sector. - Fossil fuels. Supran and Oreskes (Environmental Research Letters, Aug 2017) analysed 187 ExxonMobil communications from 1977 to 2014. About 80–83% of internal documents and peer-reviewed papers acknowledged that climate change is real and human-caused, versus 12% of paid advertorials, 81% of which expressed doubt. - Supran, Rahmstorf and Oreskes (Science, 13 Jan 2023) assessed the company’s own internal warming projections. - ExxonMobil disputes these conclusions [company position not re-fetched]. - Comparative tactics list. Goldberg and Vandenberg (Environmental Health, March 2021) compared tobacco, coal, sugar, an atrazine manufacturer and the Marshall Institute. They identified 28 doubt-manufacturing tactics, of which five were used by all five. - The policy arena. Ulucanlar, Fooks and Gilmore (PLoS Medicine, 2016) modelled tobacco industry political activity (the “policy dystopia model”). The Lancet commercial determinants of health series (Gilmore et al., March 2023) generalised the argument to tobacco, ultra-processed food, fossil fuels and alcohol.

Caveats on the corroboration. - Most of this literature comes from a small research community, often the same authors or centres (UCSF, Bath). It uses deductive coding against typologies derived from tobacco, which predisposes it to find tobacco-like patterns. - Litigation-driven document release produces observability bias. Industries that were sued are over-represented, and there is no comparison group of industries that did not behave this way. - The evidence establishes that the playbook exists across sectors. It does not establish how prevalent it is within them.

Verdict: strengthened. Constancy within tobacco is now documented into the 2020s. The chapter’s call for cross-sector document evidence has been partly answered, and the answers largely confirm it, with sector-specific gaps (PFAS) and acknowledged selection bias.

Weight. Strong support for the lesson that where revenue depends on non-regulation, organised doubt-production is a recurring pattern, and that exposure leads to reinvention of vehicles rather than abandonment (digest insights 1, 11). Moderate support for treating the full eight-strategy toolkit as a template for any given industry. Individual sectors use subsets.

Claim 2. Industry sponsorship or affiliation predicts sponsor-favourable conclusions independently of methodological quality and peer review; among 106 passive-smoking reviews, industry affiliation had OR 88.4 (95% CI 16.4–476.5), and quality and peer review were not significant (pp. 154, 160–161, Table 7.1; p. 167)#

Original claim. - In Barnes and Bero (1998), “the only factor” associated with a review concluding that passive smoking is not harmful was tobacco-industry affiliation (p. 161). - Generalising to pharmaceuticals, “the funding source does not appear to influence the methodological quality of the published research”, so bias arises from how questions are asked, how research is conducted and whether it is published (pp. 167–168).

What happened since.

Independent corroboration of direction. - Lundh et al., Cochrane Methodology Review MR000033 (16 Feb 2017; 75 studies). - Industry-sponsored drug and device studies more often had favourable efficacy results (RR 1.27, 95% CI 1.17–1.37) and favourable conclusions (RR 1.34, 1.19–1.51). - They did not differ in most risk-of-bias domains, and had lower risk of bias from blinding. - Agreement between results and conclusions was lower in industry studies (RR 0.83). - Conclusion: “an industry bias that cannot be explained by standard ‘Risk of bias’ assessments”. - Bero is a co-author, but this is a Cochrane review with pre-specified methods. - Nejstgaard et al. (BMJ, 9 Dec 2020; 21 studies). Financial conflicts were associated with favourable recommendations across guidelines, advisory committee reports, opinion pieces and narrative reviews (combined RR 1.26, 1.09–1.44). Individual document types were imprecise. - Same domain, new product. Pisinger, Godtfredsen and Bender (Preventive Medicine, 2019) blinded assessment of 94 e-cigarette content and in-vitro studies. - 95.1% of papers without a conflict found potential harm, versus 39.4% with a conflict, and “only 7.7% of tobacco industry-related studies”. - OR for “no harm” with an industry-related conflict: 66.92 (95% CI 8.1–552.9); with a strong or moderate conflict: 91.50 (10.9–771.4). - The magnitude and imprecision closely resemble Table 7.1. - Food. - Bes-Rastrollo et al. (PLoS Medicine, 31 Dec 2013): sugar-sweetened beverage reviews with food-industry conflicts were five times more likely to conclude no association with weight gain (RR 5.0, 1.3–19.3; 17 reviews). - Mandrioli, Kearns and Bero (PLoS One, 2016): sweetener-industry-sponsored reviews had more favourable results (RR 17.25, 2.34–127.29) and conclusions, bias “that could not be explained by other sources of bias”.

Evidence that qualifies the claim. - Quality is not always independent. Hansen et al., Cochrane MR000047 (5 Aug 2019), looked at systematic reviews, the same document type as Barnes and Bero. - Reviews with financial conflicts more often had favourable conclusions (RR 1.98, 1.26–3.11; 7 studies, 411 reviews). - They also “tended to have lower methodological quality”. - It was “uncertain” whether conflicts were associated with the results.

So for reviews, sponsorship and lower quality travel together. That differs from the chapter’s “quality was not significant”. - Food-sector associations are weaker and less consistent. - Chartres, Fabbri and Bero (JAMA Internal Medicine, Dec 2016): RR 1.31 (0.99–1.72, non-significant) for favourable conclusions in food-industry-sponsored nutrition studies; no association with methodological quality. - Chartres et al. (BMJ Open, 2020) on dairy and cardiovascular disease: “no clear evidence” of an association between industry ties and favourable results or conclusions, though industry-sponsored studies showed smaller risk estimates. - Magnitude. No later synthesis approaches OR 88.4 except other small, sharply polarised tobacco and nicotine datasets. The Table 7.1 interval (16.4–476.5) signals very sparse cells. - The chapter’s own text. It paraphrases the result as tobacco-affiliated reviews being about “90 times more likely” (digest caveat). It is an odds ratio for affiliation, not a probability ratio.

Verdict: held up in direction and in the central point that sponsorship bias is not captured by standard quality checklists (for primary studies). Qualified on three counts: - for reviews, conflicts co-occur with lower quality; - effect sizes vary widely by sector; - OR 88.4 should not be quoted as a general effect size.

Weight. Strong for digest insight 2: sponsorship shapes conclusions in ways that method-quality screening does not catch. Moderate for insight 3 (peer review did not predict conclusions). No later synthesis tested peer-review status as such, and later work shows peer-reviewed literature is itself affected.

Claim 3. Second-hand smoke causes lung cancer and cardiovascular, respiratory and developmental harm; California EPA (2005) also found it causes breast cancer in younger, premenopausal women (p. 153)#

Original claim. - The chapter cites the 1986 US Surgeon General and National Research Council reports, the 1992 US EPA Group A classification and California EPA (1997, 2005). - California’s 2005 list of causal effects includes developmental (foetal growth, SIDS, pre-term delivery), respiratory, carcinogenic (“lung cancer, nasal sinus cancer, breast cancer in younger, premenopausal women”) and cardiovascular outcomes (p. 153). - It reports IARC’s 1998 multicentre study (16% and 17% increases) without noting that both intervals include 1 (digest caveat).

What happened since.

Lung cancer and overall carcinogenicity. IARC Monograph 100E (2012) reaffirmed: “There is sufficient evidence in humans for the carcinogenicity of second-hand tobacco smoke. Second-hand tobacco smoke causes cancer of the lung.” Larynx and pharynx show a “positive association”. It concluded “Group 1”.

Cardiovascular and other outcomes. - CDC (page updated 31 Jan 2025). - SHS causes “more than 7,300 lung cancer deaths” and “nearly 34,000 premature deaths from heart disease each year” among US non-smokers. - Non-smokers exposed to SHS “increase their risk of stroke by 20−30%”. - About 2.5 million non-smokers have died from SHS exposure since 1964. - 2014 Surgeon General’s report. Added stroke as a causal outcome of SHS [not re-fetched: NCBI Bookshelf blocked by reCAPTCHA; the CDC page reflects it]. - Smoke-free laws. Frazer et al., Cochrane CD005992.pub3 (2016; 77 studies, 21 countries): - “consistent evidence of a positive impact of national smoking bans on improving cardiovascular health outcomes”; - clearest for acute coronary syndrome admissions; - less consistent for respiratory and perinatal outcomes.

Ireland’s 2004 national ban was followed by estimated 26% and 32% reductions in ischaemic heart disease and stroke mortality in people aged 65 and over (Stallings-Smith et al., 2013). A contrary US national analysis (Shetty et al., Journal of Policy Analysis and Management, 2011) found no statistically significant short-term declines. The size of the effect remains debated, though not its direction. - Global burden. WHO’s fact sheet (current version dated 26 June 2026) says SHS “kills more than 1.6 million people prematurely every year”. The 2004 estimate was 603,000 (Öberg et al., Lancet, 2011). The difference reflects changed methods and endpoints rather than a measured rise.

Breast cancer: not adopted by the main bodies; still contested. - IARC (2012) summarised the positions: IARC 2004 characterised the evidence as “inconsistent”; the 2006 Surgeon General found it “suggestive but not sufficient”; CalEPA 2005 and a Canadian panel (2009) found it “consistent with a causal association in younger primarily premenopausal women”. IARC 2012 concluded that “the results for an association with premenopausal breast cancer are also inconsistent”. The Working Group noted unresolved concerns about “recall or publication bias”. - The 2014 Surgeon General’s report kept “suggestive but not sufficient” for SHS and breast cancer [not re-fetched]. - Glantz and Johnson (Cancer Epidemiology, Biomarkers & Prevention, Jan 2014) argued that the Surgeon General had become “increasingly cautious and defensive”, and that the breast-cancer evidence was stronger in 2005 than the lung-cancer evidence was in 1986. Glantz is a long-standing collaborator of the chapter’s author. - Later meta-analyses split by study design: - Possenti et al. (British Journal of Cancer, 2024; 63 articles): pooled RR 1.24 (1.15–1.34), with dose-response. Home exposure 1.17; workplace 1.03 (0.98–1.08). - Macacu et al. (Breast Cancer Research and Treatment, 2015): prospective studies 1.07 (1.02–1.13); retrospective studies 1.30 (1.10–1.54) with high heterogeneity. - Yang et al. (PLoS One, 2013; 10 cohorts): RR 1.01 (0.96–1.06). - Wada et al. (International Journal of Epidemiology, 2024; nine Japanese cohorts, 166,611 women): “no association between adulthood or childhood exposure to secondhand smoke and breast cancer”.

The pattern of stronger case-control than cohort associations is the signature of possible recall bias that IARC flagged.

Verdict: partly held up. - Strengthened for lung cancer, cardiovascular disease (stroke added) and child respiratory and developmental effects. - Contested for breast cancer: no major review body has adopted CalEPA’s 2005 causal finding, and the cohort evidence is weak.

Weight. The chapter’s argument does not depend on breast cancer. The case that industry manufactured doubt about a real hazard is secure. The breast-cancer example also shows the other side of the chapter’s own concession (p. 166) that both sides “construct” evidence. Public-health-aligned bodies (CalEPA) and more cautious ones (IARC, the Surgeon General) disagreed on the same data. That is a reminder that sceptical scrutiny should apply to harm claims as well as safety claims.

Claim 4. The US data-access (1998) and Data Quality (2000) riders were driven by covert tobacco-industry coordination; they could block non-industry science, slow regulation and shift evidentiary standards towards industry’s, while applying only to publicly funded data (pp. 162–163; Panel 7.1, p. 164)#

Original claim. - Internal documents “reveal that these policies were driven by tobacco industry efforts to coordinate corporate interests” (Baba et al., 2005) (p. 163). - The riders “have the potential to block agencies from using emerging science from non-industry sources and to slow the regulatory process”. They “could shift the scientific standards of data used for policy purposes to favour standards promulgated by industry”. They “only apply to data generated with government funding, not industry funding” (p. 163). - Panel 7.1 adds Philip Morris’s “good epidemiological practice” standards, including a proposal to disregard relative risks below 2.0 (p. 164).

What happened since.

Recurrence of the mechanism. - EPA “Strengthening Transparency in Regulatory Science.” Proposed 30 April 2018 (83 FR 18768), supplemented March 2020, and finalised 6 January 2021 (86 FR 469) as “Strengthening Transparency in Pivotal Science Underlying Significant Regulatory Actions and Influential Scientific Information”. - The final rule told EPA to “give greater consideration to studies where the underlying dose-response data are available in a manner sufficient for independent validation”. - The editors of Science, Nature, Cell, PNAS, The Lancet and PLOS issued a joint statement against the proposal (Nov 2019) [characterised from title and standard summaries; text not re-read]. - The US District Court for Montana (Environmental Defense Fund v. EPA) held on 27 January 2021 that the rule was substantive, not procedural. In EPA’s summary, the court “reasoned that by determining how the Agency weighs particular scientific evidence, the 2021 final rule determined outcomes rather than process”. The court vacated the rule on 1 February 2021, and EPA removed it (86 FR 29515, 2 June 2021).

The court’s reasoning matches the chapter’s reading of such rules as changes to evidentiary standards dressed as transparency. - Tobacco origins. Goobie, Rice and Carlsten (Annals of the American Thoracic Society, Oct 2021) traced the rule’s lineage explicitly to the tobacco campaign, citing Truth Tobacco Industry Documents: - a 1996 memo advising “procedural hurdles” focused on “process as opposed to scientific substance”; - Philip Morris’s 1997 “Sound Science Project Plan” to mandate epidemiological data sharing and link ETS findings “to junk science”. - Deregulatory science. Mansfield (Social Studies of Science, 2021) analysed EPA frameworks of 2017–2020 as “deregulatory science” that justified rollbacks “in terms of evidence-based decision-making, transparency and the separation of science from politics”. - Executive Order 14303, “Restoring Gold Standard Science” (signed 23 May 2025; 90 FR 22601). Among other things it: - requires agencies to publish data and models behind “influential scientific information”, “consistent with relevant policies that protect … confidential business information” (s. 4(b)); - directs that “highly unlikely and overly precautionary assumptions and scenarios should only be relied upon … where required by law or otherwise pertinent” (s. 4(e)); - mandates a “weight of scientific evidence” approach (s. 4(f)); - assigns enforcement to “a senior appointee designated by the agency head”, with corrections “consistent with … section 515 … commonly known as the Information Quality Act” (s. 7(a)); - lists “without conflicts of interest” as one of nine tenets of Gold Standard Science (s. 3(a)(ix)).

EPA now describes its literature reviews as “consistent with Executive Order 14303”, for example the glyphosate open-literature search (91 FR 54870, 25 Aug 2026). Critical commentary includes Carter, Michaels and Wagner, BMJ, June 2026 [title only], and Thorp, Science editorial, Sept 2025 [not critical of the order as such; abstract only].

Evidence that complicates the prediction. - The 2021 rule never took effect. Challenges under the Information Quality Act have had limited judicial enforceability [not re-fetched]. - Whether EO 14303 will, in practice, down-weight non-industry science cannot yet be measured. It also contains a conflict-of-interest tenet that points the other way. - Data transparency and reproducibility are legitimate aims, pursued independently of industry by journals (the ICMJE requires data-sharing statements) and funders. Not every transparency rule is a tobacco-style manoeuvre. - The asymmetry claim is jurisdiction-specific. The EU moved in the opposite direction. Regulation (EU) 2019/1381 (applying from 27 March 2021): - requires “business operators” to notify EFSA of “any study commissioned or carried out” to support an application (Article 32b); - requires supporting data to be made public “in a proactive manner” (recital 28), noting concerns “with respect to studies commissioned by the industry”.

That is transparency applied to industry data, the reverse of the US riders. EO 14303’s carve-out for confidential business information preserves the US asymmetry the chapter identified. - I found no later revival of the “relative risk below 2.0” rule in EU law. Panel 7.1 already reported that none had been adopted.

Verdict: strengthened as a diagnosis of a recurrent mechanism. Transparency and data-quality rules can be designed to change which science counts, and the pattern recurred in 2018–21 and 2025–26 with documented tobacco lineage. Unclear on realised effects. The predicted blocking of non-industry science has not been demonstrated at scale; the 2021 rule was vacated, and EO 14303 is too recent to assess.

Weight. Strong for digest insight 6 (changing evidentiary rules outlasts attacking individual studies) and insight 7 (transparency can be applied asymmetrically). The contrasting EU Transparency Regulation shows the asymmetry is a design choice, not inherent to transparency.

Claim 5. Panel 7.1: BAT, working through a Brussels think-tank front group (the Risk Forum, later the European Risk Forum), “contributed to” embedding business impact assessment and “structured risk assessment” in EU policymaking; the effect on precaution was “somewhat successful” but it is “not yet possible to assess” (pp. 164–165)#

Original claim. - BAT sought a “legislated demand for structured risk assessment” to “remove the possibility of introducing public smoking restrictions” (BAT, 1995). - Advised to work through a “front group”, it recruited the European Policy Centre. The EPC’s invitation-only Risk Forum included oil, chemical and pharmaceutical firms. - These efforts “contributed to” Treaty amendments (1997) requiring the Commission to “consult widely” and minimise “burden” on “economic operators” (p. 165). - The Risk Forum became the separate European Risk Forum (2006–07), self-described as “expert led” despite “solely representing corporate interests” (p. 165). - The panel judged these efforts “somewhat successful” in redefining responses to risk, but said “it is not yet possible to assess the success of the BAT or Philip Morris campaigns”. It noted that corporate input is legitimate if transparent and not excessive (p. 165).

What happened since.

The panel’s authors extended the analysis. - Smith, Fooks, Gilmore, Collin and Weishaar (Journal of Health Politics, Policy and Law, Feb 2015) added interview data. They argued that BAT deliberately constructed “a vaguely defined idea that could be strategically adapted to appeal to diverse constituencies”, and that regulated industries favour “evidence tools” such as impact assessment. - Peeters et al. (Tobacco Control, published online Feb 2015) analysed the 2014 Tobacco Products Directive lobby, drawing on 581 documents obtained by freedom-of-information request and 28 leaked PMI documents. - PMI alone employed over 160 lobbyists. - Efforts to “Push” or “Delay” the proposal were “partially successful”: plain packaging and a point-of-sale display ban were removed during a three-year delay in the Commission. - The “Smart Regulation mechanism contributed to changes and delays”. It facilitated undisclosed meetings “contrary to Article 5.3”. - Costa et al. (Tobacco Control, 2014) used automated text analysis. The Commission’s position shifted towards the industry’s submissions between 2010 and 2012 (score 0.52 to 0.40 on a 0 = industry, 1 = public-health scale).

Institutional developments consistent with the panel’s concern. - Regulatory Scrutiny Board. COM(2015) 215 (19 May 2015) announced “a new Regulatory Scrutiny Board” to replace the 2006 Impact Assessment Board “with a strengthened role”. - “One in, one out”. COM(2021) 219 (29 April 2021) introduced it and said compliance costs “will be scrutinised by the Regulatory Scrutiny Board”. - Treaty language. The 1997 “burden” wording survives in Protocol No 2 (Lisbon): draft acts “shall take account of the need for any burden … falling upon … economic operators and citizens, to be minimised”. - The “innovation principle” entered EU law in Regulation (EU) 2021/695 (Horizon Europe), recital 6: activities should “take advantage of and inspire innovation-friendly regulation, in line with the innovation principle”.

The Forum’s own account (fetched 25 Sept 2026). - Renaming. The European Risk Forum is now the European Regulation and Innovation Forum (ERIF), a registered association since 2013. - Claimed achievements. Its “About us” page says it “has successfully championed” key developments in EU Better Regulation, including: - “the design and scope of the Commission’s impact assessment system”; - “the establishment of its oversight and scientific advice bodies (currently the Regulatory Scrutiny Board and the Group of Chief Scientific Advisers)”; - “the introduction of the ‘Innovation Principle’”. - Current agenda. Its 2024–25 key messages call for Regulatory Scrutiny Board recommendations to be “mandatory” unless overruled by the College, and for estimating the “opportunity cost” of proposals. On expert bias, it says bias “should be addressed through public disclosure and transparency and not through exclusion because of its origin”, a direct counter to the chapter’s p. 170 case for bans. - Board. Its chair is a former senior adviser at Dow, and one board member previously worked in regulatory affairs at British American Tobacco. - Registration. In the EU Transparency Register (as shown by LobbyFacts, registration of 12 January 2026) it declares that it “does not represent commercial interests”. It lists “member organisations: none declared”, while stating that its funds “come almost exclusively from annual fees paid by our members, comprised of both international companies and sectoral trade associations”. Total budget: €100,000.

This echoes the panel’s 2010 observation about an “expert led” body representing corporate interests. Current tobacco membership could not be verified.

Evidence that complicates attribution and effect. - Better Regulation had many drivers. These include OECD regulatory reform, member-state programmes and Commission initiatives [general literature, not re-fetched]. The Forum’s self-description is advocacy, not independent evidence of causation. Lauber and Brooks (Globalization and Health, 2023) and Radaelli (Journal of European Public Policy, 2023) treat Better Regulation as a contested meta-regulatory agenda shaped by many actors. - Strong tobacco law still passed and was defended. - The 2014 Tobacco Products Directive (graphic warnings, characterising-flavour ban) was upheld by the CJEU in Philip Morris Brands (C-547/14, 4 May 2016). - The Court relied in part on the Commission’s own impact assessment of 19 December 2012 to find the measures justified. Impact assessment cut both ways. - The EU adopted a strengthened Council Recommendation on smoke- and aerosol-free environments on 3 December 2024. - Precaution remains a Treaty principle (Article 191 TFEU). The innovation principle appears only in a recital, not in the Treaties.

Verdict: partly held up. - Supported by later document-and-interview research and by the Forum’s own claims: the tobacco-initiated coalition helped push EU policymaking towards impact assessment, “burden” minimisation, scrutiny bodies and an innovation principle. - Better Regulation cannot be attributed to BAT. The tools have also been used to defend public-health measures. - The net effect on precaution remains, as the panel said, hard to assess.

Weight. Moderate to strong for the lesson that meta-regulation (how rules get made) is a strategic target and that “evidence tools” are not value-neutral (digest insight 6). Weak for any claim that precaution in the EU was substantially neutralised.

Claim 6. Industry procedural tactics delayed protective action: 64% of public submissions on the 1992 EPA risk assessment called it invalid, 71% of those industry-affiliated; the industry won a procedural court case in 1998; a “cloud” over the assessment lasted until 2002, delaying protective legislation; the industry subverted the IARC study (pp. 153, 164, 166, 171)#

Original claim. - The volume of industry comments “probably delayed” EPA’s report, though EPA “revised the report in response to valid criticisms” (p. 153). - 69/107 submissions (64%) claimed the draft invalid; 49/69 (71%) came from tobacco-industry-affiliated individuals (p. 166). - Flue-Cured Tobacco Co-op v. EPA (1998) succeeded “on procedural grounds” (p. 153). The campaign placed “a cloud over its validity” until 2002, “leading to delays in subsequent introduction of protective legislation” (p. 164). - Table 7.2: “1993–1998 Tobacco industry subverts the WHO International Agency for Research on Cancer (IARC) study” (p. 171). - A federal OSHA workplace smoking rule “failed” (p. 167).

What happened since.

Court record. - The EPA case. The district court’s ruling (Flue-Cured Tobacco Coop. Stabilization Corp. v. EPA, 4 F. Supp. 2d 435, M.D.N.C., 17 July 1998) was “vacated and remanded” by the Fourth Circuit on 11 December 2002 (313 F.3d 852). The appeal court treated the report as not reviewable final agency action [holding characterised; opinion located but not re-read in full]. This confirms the 2002 end-point of the “cloud”. - The fraud case. United States v. Philip Morris USA: the district court (2006) found that the defendants “had for decades conspired to deny the health effects of smoking in violation of RICO” (as summarised in the 2017 memorandum opinion, 257 F. Supp. 3d 1). The D.C. Circuit largely affirmed in 2009 (566 F.3d 1095). - The remedy required corrective statements on five topics, including “(e) the adverse health effects of exposure to secondhand smoke” (topics listed in 907 F. Supp. 2d 1, 27 Nov 2012; Smith et al., 2011). - In 2015 the D.C. Circuit struck the proposed preamble announcing that the companies had “deliberately deceived the American public” [holding characterised from the opinion’s opening; not re-read in full]. - Statements ran in newspapers and on TV from late 2017. An estimated 40.6% of US adults recalled seeing one in 2018 (Blake et al., 2020). - Point-of-sale statements followed: 36.5% of adults reported seeing them in 2024 (Gibson et al., 2026).

This is a judicial finding, independent of the chapter’s authors, that corroborates the core of the chapter.

The lag from causal conclusions to protective law. - Timeline of evidence. First epidemiology in 1981; causal conclusions from the US Surgeon General and NRC in 1986; EPA Group A in 1992. - Ireland became “the first country in the world to implement a national workplace smoking ban” on 29 March 2004 (Stallings-Smith et al., 2013). That is 18 years after the 1986 conclusions. - European Union. - A Council Recommendation (30 Nov 2009) called on member states to legislate for smoke-free “enclosed public places, workplaces and public transport” within three years. - A new Recommendation (3 Dec 2024) extends protection to “key outdoor areas” and to e-cigarettes and heated tobacco products. - United States. - OSHA withdrew its Indoor Air Quality proposal on 17 December 2001 (66 FR 64946). It cited the fact that “a great many state and local governments and private employers have taken action to curtail smoking”. There is still no federal workplace rule. - As of 1 July 2026, 28 states and 62.8% of the US population were covered by 100% smoke-free laws for workplaces, restaurants and bars (state plus local; American Nonsmokers’ Rights Foundation, an advocacy organisation). - World. WHO (2026 fact sheet): “over one-third of the global population across 79 nations” is covered by comprehensive national smoke-free legislation.

Complications. - The chapter concedes that some industry criticisms were “valid” and improved the EPA report (p. 153). Procedural participation did produce better science as well as delay. - The OSHA withdrawal notice attributes federal inaction partly to state, local and private action having overtaken it. Delay at one level coincided with progress at others. - The size of the delay caused by industry tactics, as opposed to public opinion, hospitality-sector opposition and federalism, has not been quantified. Peeters et al. (2016) quantify delay mechanisms for the EU Tobacco Products Directive, not for smoke-free laws. - IARC. The IARC study’s published estimates were imprecise (the 16% and 17% increases had intervals including 1). IARC nonetheless classified SHS as Group 1 in 2004 and 2012. The subversion campaign did not change the outcome, though it contributed to early misreporting (Ong and Glantz, 2000, cited by the chapter).

Verdict: held up. The 2002 appellate outcome and the 2006 RICO findings on SHS corroborate the account. Lags of roughly two decades from causal conclusions to comprehensive laws, and continuing gaps (no US federal rule; most of the world’s population uncovered), are consistent with the chapter. The causal share attributable to industry tactics remains qualitative.

Weight. Strong for digest insight 8 (procedural safeguards favour the well-resourced) and for “flooding” consultations as a delay tactic. The same record shows that comment processes also caught valid errors, so the lesson is about imbalance, not participation as such.

Claim 7. Print coverage of SHS research (1981–1995) left readers with an impression of “continuing controversy” in 66% of newspaper and 55% of magazine articles, at a roughly constant rate; tobacco-industry officials were cited in 52% of newspaper articles (p. 166)#

Original claim. - The data come from Kennedy and Bero (1999): 180 newspaper and 95 magazine articles. The share conveying controversy “remained relatively constant” even as evidence accumulated. - 52% cited tobacco-industry officials, 56% government officials and 46% independent scientists. - Citing industry officials as experts “could have contributed to the emphasis on controversy” (p. 166).

What happened since. - No replication. I found no later replication or re-analysis of the SHS press sample. - Experimental support for the mechanism. Koehler (Journal of Experimental Psychology: Applied, 2016) showed across several experiments that presenting two disagreeing experts (“balance”) alongside a full count of expert opinion made participants’ perceptions of agreement “not discriminate as clearly between issues with and without strong expert consensus”. They were also “less likely to think that there was enough agreement … to guide government policy”. False balance “can distort perceptions of expert opinion even when participants would seem to have all the information needed to correct for its influence”. - Related work on vaccine–autism coverage (Dixon and Clarke, Science Communication, 2012/13) and climate (Imundo and Rapp, Journal of Applied Research in Memory and Cognition, 2022) reports similar effects [titles and bibliographic records only]. - Institutional acknowledgement. The BBC Trust (July 2014) endorsed a follow-up to the 2011 Jones review. It noted an “over-rigid” application of impartiality rules to science. It said that judging the weight of scientific agreement correctly “will mean that the BBC avoids the ‘false balance’ between fact and opinion”. - Shifting practice. Later media studies speak of coverage moving “beyond false balance” in some outlets (Brüggemann and Engesser, Global Environmental Change, 2017) [title only; abstract not retrievable]. Media amplification of manufactured doubt is therefore not fixed; it depends on newsroom norms.

Verdict: held up. The descriptive figures stand unchallenged. The inferred mechanism (balance norms plus industry spokespeople sustain perceived controversy) now has experimental support. The causal contribution of industry citation in the 1981–95 SHS coverage specifically remains untested.

Weight. Moderate to strong for digest insight 10 (news practice can sustain apparent controversy after consensus forms). The later literature adds a remedy: weight-of-evidence framing and editorial guidance.

Claim 8. Recommendation: full disclosure of funders’ roles at every research stage and of personal financial ties, in publications, letters, press citations and policy or legal testimony; disclosure is “necessary, but not sufficient” (p. 169)#

Original claim. - Funding sources “for all published research should be fully disclosed”. So should the sponsor’s role in “designing, conducting and publishing”, and personal ties (consulting, stock, honoraria). - Disclosures should appear in “scientific publications (including research articles, letters to the editor and editorials), citations in the lay press, and testimony in policy or legal settings”. - Disclosure “often provides incomplete information” (p. 169), and it is “necessary, but not sufficient”. The chapter cites Krimsky’s view that disclosure can be a “public relations” response (pp. 169–170).

What happened since.

Uptake (journals and payment registries). - ICMJE Recommendations (updated January 2026). Articles must declare “sources of support for the work, including sponsor names along with explanations of the role of those sources if any in study design; collection, analysis, and interpretation of data; writing of the report; any restrictions regarding the submission”. Authors “must avoid entering into agreements with study sponsors … that interfere with authors’ access to all of the study’s data”. The ICMJE disclosure form was last updated in February 2021.

This is close to the chapter’s recommendation for journals. - Payment registries (medicine only). - US: the Sunshine Act (Affordable Care Act s. 6002) created CMS Open Payments. Program year 2025 covered 17.07 million records and US$14.67 billion in payments from drug and device makers to clinicians. - Similar schemes exist in France and under the European pharmaceutical industry’s (EFPIA) code [not re-fetched]. - These cover drug and device manufacturers. There is no equivalent registry for payments from tobacco, food or chemical companies to researchers. - Press citations and policy testimony. I found no systematic uptake of disclosure requirements [not exhaustively searched, given the search quota]. Lobby registers can obscure funders; see ERIF’s declaration under Claim 5.

Does disclosure work? - Accuracy is poor. Taheri et al. (BMJ Open, 2021; 40 studies) found pooled discrepancies between physicians’ self-reported conflicts and objective payment data of 81% at article level (95% CI 72–89%) and 66% at author level. “A median of 45% of non-disclosed payments were directly or indirectly related to the work.” - Undisclosed conflicts still bias. Cherla et al. (Journal of General Internal Medicine, 2019; 590 articles) found a 69% discordance rate. Authors who failed to disclose were more likely to reach industry-favourable conclusions than authors without conflicts (73% vs 62%). - Effectiveness is limited. Graham et al. (Research Integrity and Peer Review, May 2025; 81 studies; Bero co-author) found disclosure policies “are not consistently designed, implemented, or enforced”. It concluded they “are not particularly effective in mitigating risk of bias”. The review called for evaluating mechanisms “other than disclosure”. - Behavioural evidence is mixed. - Disclosure can backfire: Loewenstein, Sah and Cain, JAMA, 2012 [title]. Sah, Loewenstein and Cain (Journal of Personality and Social Psychology, 2013) found it can increase “pressure to comply with … advice” that is distrusted. - It can help when conflicts are avoidable: Sah and Loewenstein (Psychological Science, 2014) found mandatory and voluntary disclosure “can deter advisors from accepting conflicts of interest”. - Counter-view. Some argue the conflict-of-interest agenda had become excessive or moralised, for example Rosenbaum’s three-part NEJM series (2015) [title only] and Babor and Miller, Addiction, 2014, on “McCarthyism” accusations [title only]. These do not dispute that disclosure is insufficient; they dispute how far to go beyond it.

Verdict: strengthened. Disclosure became near-universal in medical journals and in payment registries for drug and device makers. The chapter’s caution that it is “not sufficient” is now backed by systematic evidence of widespread non-disclosure and limited effect on bias. Uptake beyond medicine, and in press and testimony settings, remains thin.

Weight. Strong for digest insight 12 (disclosure is insufficient; structural measures may be needed). This is one of the chapter’s best-vindicated recommendations. Note also that the preferred remedy of industry-aligned bodies is disclosure instead of exclusion (ERIF, Claim 5), which fits Krimsky’s warning.

Claim 9. Recommendation: bans on tobacco-industry research funding are “warranted”; sequestration (independent institutes funded by industry) and total bans on researchers’ financial ties are canvassed as debated options (p. 170)#

Original claim. - “Bans on tobacco industry support for research are warranted in view of the industry’s history of deception”. The industry’s motives include “distracting attention”, “gaining credibility” and public relations (p. 170). - The chapter cites institutional bans (Harvard, University of Sydney; Legacy Foundation grant conditions). - It describes Schafer’s “sequestration thesis” (an independent research institute funded by companies) and notes Shamoo and Resnik’s view that eliminating ties “may not be realistic” (p. 170). - It does not mention the WHO Framework Convention on Tobacco Control (2003).

What happened since.

Bans spread. - Journals. In October 2013 the BMJ, Heart, Thorax and BMJ Open announced they would no longer consider tobacco-industry-funded research (Godlee et al., “Journal policy on research funded by the tobacco industry”, BMJ 347:f5193, 15 Oct 2013) [title and bibliographic record; text blocked]. Tobacco Control changed its own policy in 2013 (Malone, “Changing Tobacco Control’s policy on tobacco industry-funded research”) [title only]. - Professional societies. The International and American Associations for Dental, Oral and Craniofacial Research adopted a policy statement on tobacco-industry-funded research in 2025 (Journal of Dental Research, May 2025) [title only]. - Inhaled medicines. PMI’s 2021 acquisition of the inhaler maker Vectura prompted respiratory clinicians to argue that its products and funding should be treated as tobacco industry (e.g. Hopkinson, Thorax, 2021/22; Capstick and Hopkinson, 2021) [titles only]. - Treaty. Article 5.3 obliges Parties to “protect these policies from commercial and other vested interests of the tobacco industry”. Its guidelines were adopted at the third Conference of the Parties in November 2008. This institutionalises separation at government level, which is the chapter’s logic but broader. - STOP’s Global Tobacco Industry Interference Index 2025 (100 countries; STOP is an advocacy network) reports persistent interference, including a “revolving door between the government and the industry” and partnerships on corporate social responsibility despite 32 countries banning them.

Sequestration tested: the Foundation for a Smoke-Free World. - Launch. PMI launched FSFW in 2017 with a pledged US$1 billion over 12 years, presented as independent. Van der Eijk, Bero and Malone (2019) found that its governing documents had “multiple loopholes, particularly regarding conflicts of interest”, and concluded “it cannot be regarded as independent”. Its then president, Derek Yach, defended its independence (Lancet, 2019) [title only]. - Research practice. Legg et al. (2021, 2024) documented publication-ethics problems and industry-like research practices (see Claim 1). - End of PMI funding. In September 2023 FSFW’s pledge agreement was terminated with a “final charitable gift” of US$140 million. A new CEO said FSFW would accept no tobacco or non-medicinal-nicotine funding (4 December 2023). Cohen, Eissenberg and Zeller (Tobacco Control, 2025) could not determine whether the reconstituted foundation met their 2009 criteria for acceptable industry-funded research models, because “critical details related to FSFW governance, decision-making and process have not been announced”. - Assessment. As a test of sequestration, FSFW failed the independence criteria that its critics set in its first phase. It shows why sequestration requires structural independence (control of agenda, governance and funding level) rather than a pledge.

A working sequestration model. US tobacco manufacturers and importers pay user fees under s. 919 of the Food, Drug, and Cosmetic Act that “provide funding for FDA’s tobacco regulatory activities”. Those activities include the joint FDA–NIH Tobacco Centers of Regulatory Science (Perry et al., 2020). Industry money reaches research through a public intermediary, with no industry say over the agenda.

This is my inference from the FDA user-fee page and the program’s structure. I did not find an evaluation of its independence.

Why continued covert funding matters. Braznell et al. (2025) show covert Philip Morris Japan funding of university research in 2014–2019. This supports the chapter’s premise that disclosure alone does not stop the practice.

Counter-arguments. - Harm-reduction researchers argue there are “scientific and ethical obligations to publish tobacco industry-funded research on nicotine delivery systems of reduced risk” (Polosa and Crawley, Toxicology, 2017) [title only]. - Regulators require manufacturers’ own studies in premarket applications under the US Tobacco Control Act [not re-fetched]. So industry science cannot be excluded from regulatory decisions, only discounted and checked. - ERIF argues against exclusion of experts “because of its origin” (Claim 5). - No study has measured whether funding bans improve the evidence base or policy outcomes.

Verdict: held up for tobacco. Bans spread, and the main attempt at “independent” industry-funded research (FSFW) and recent leaked documents vindicate the chapter’s scepticism. Sequestration worked only where public bodies control the money (FDA user fees). It failed where the company created the “independent” body. Whether bans work is unmeasured.

Weight. Strong for the lesson that structural separation, not disclosure, is the tested response to an industry with a documented record of deception (digest insight 12). Weak evidence on the costs and benefits of bans. The lesson generalises to other sectors only with care, because the chapter justifies bans by tobacco’s unique history (p. 167).

Claim 10. Five diagnostic questions, which trace a controversy to interest-group funding, interest-group publications, low-quality research, affiliated press sources and interest-group-only presentation to policymakers, should be applied “to all situations in which a company has an interest in creating controversy” (p. 167)#

Original claim. The questions are: 1. Does the controversy exist only because interest-group-funded findings differ from others’? 2. Is it supported mainly by interest-group-supported publications? 3. Is it supported mainly by low-quality research? 4. Is it perpetuated in the lay press through interest-group-affiliated individuals? 5. Are the data suggesting controversy presented to policymakers only by the interest group?

Policymakers “should apply these questions to all situations in which a company has an interest in creating controversy about the risks of its products” (p. 167).

What happened since. - No validation. I found no study that validated these five questions as a diagnostic tool (sensitivity, specificity, inter-rater reliability), and no regulator that formally adopted them. - Richer typologies developed and applied: - the Science for Profit Model (Legg et al., 2021), used as an analytic framework for FSFW (Legg et al., 2024) and Philip Morris Japan (Braznell et al., 2025); - Goldberg and Vandenberg’s 28 tactics, offered so that “others … build a case that an industry or group is deliberately manipulating information”; - the policy dystopia model (Ulucanlar et al., 2016).

These are used retrospectively by researchers. None has been prospectively tested for prediction. - Institutional analogues: - FCTC Article 5.3 (tobacco only); - the EU Transparency Regulation’s study-notification register (tests question 1 indirectly, by exposing unpublished sponsor studies); - ICMJE sponsor-role disclosure (question 2). - Question 3 remains the hardest. The chapter’s own evidence (Claim 2) and Cochrane 2017 show that industry bias is often not visible as low quality. A heuristic keyed to “low scientific quality” can miss the more sophisticated cases, such as industry-sponsored drug trials with good risk-of-bias scores. - Risk of misuse. Heuristics about who funds a claim can slide into dismissing valid criticism by origin. The chapter’s own account shows some industry criticisms of the EPA assessment were “valid” (p. 153). The breast-cancer dispute (Claim 3) shows public-interest bodies can also overreach. The questions test the provenance of a controversy, not its scientific merit, and need pairing with independent substantive review.

Verdict: unclear. The questions are consistent with a much larger later literature and with institutional practice in tobacco control, but they have not been validated or adopted as such. Their reliance on “low quality” as a marker is partly undercut by later evidence that sponsorship bias often survives quality screening.

Weight. Moderate as a first-pass screen for provenance, especially questions 1, 2, 4 and 5. Low as a stand-alone test of whether a controversy is manufactured.

Minor factual checks#

• “Cloud until 2002” (p. 164): consistent with the Fourth Circuit’s 11 December 2002 vacatur (313 F.3d 852).
• IARC 1998 study (p. 153): the reported 16% and 17% increases had 95% intervals including 1 (digest caveat). IARC classified SHS as Group 1 in 2004 and 2012 on the wider evidence.
• “90 times more likely” (digest caveat): an odds ratio for affiliation from a small sample (CI 16.4–476.5). It should not be restated as a probability ratio.
• Nasal sinus cancer (CalEPA 2005, p. 153): IARC 2012 found only “some evidence” from a cohort and case-control studies. It did not list nasal sinus cancer as caused by SHS.
• Table 7.2 transcription errors (digest caveat): not re-checked here.
• Omissions the chapter could have included in 2013: the FCTC (2003; Article 5.3 guidelines 2008); IARC 2004 Group 1; the 2006 US RICO judgment. All existed before publication.

Implications for the section’s transferable insights#

Grades refer to the digest’s list (LL2-07 digest, “Transferable insights”).

1. Where revenue depends on non-regulation, sustaining uncertainty can be deliberate and aimed at delay (pp. 153–156). Strengthened. Court findings (2006, 2009) and further document studies across sugar, PFAS and fossil fuels, with sector-specific gaps.
2. Sponsorship shapes conclusions without detectable method flaws (pp. 154, 160–161). Strong; corroborated by Cochrane 2017 for primary studies. For systematic reviews, conflicts also co-occur with lower quality (Cochrane 2019). Effect sizes vary by sector.
3. Peer-review status did not predict conclusions; affiliation did (p. 161). Moderate, unchanged. Not directly re-tested.
4. Controlling which questions get funded matters as much as controlling results (pp. 155–157). Strengthened. Fabbri et al. (AJPH, 2018; 36 studies): industry “tends to prioritize lines of inquiry that focus on products … that can be commercialized”. Document studies show topic prioritisation used to support “policy and legal positions”.
5. Legal structures hide knowledge; forced disclosure is the main window, which creates observability bias (pp. 152, 169). Strong, and the bias is now acknowledged by the synthesisers themselves (Legg et al., 2021). The EU notification register for commissioned studies (2019/1381) is a non-litigation window.
6. Changing evidentiary rules outlasts attacking studies (pp. 162–165). Strengthened on recurrence (EPA 2018–21; EO 14303; EU Better Regulation and innovation principle). Unclear on realised effect.
7. Transparency can be applied asymmetrically and used as a weapon (p. 163). Strengthened as a design point. The US continues to protect confidential business information; the EU chose symmetrical transparency for food-chain risk assessment.
8. Procedural safeguards favour the well-resourced (pp. 153, 166). Strong for these cases (EU Tobacco Products Directive lobby: 160+ PMI lobbyists; three-year delay). Procedural participation also caught valid errors.
9. Measurement and design standards are quiet battlegrounds (pp. 157, 161–162). Moderate, unchanged. Not re-tested here.
10. News practice can sustain apparent controversy after consensus forms (p. 166). Moderate to strong. Experimental evidence on false balance, plus institutional acknowledgement (BBC Trust 2014).
11. Influence toolkits are portable and reinvented (pp. 154, 157, 164–165). Strengthened. FSFW (2017–23) is a textbook reinvention. The European Risk Forum’s rebranding as ERIF, and its self-declared role in the innovation principle, show continuity of a tobacco-initiated vehicle.
12. Disclosure is insufficient; structural separation may be needed (pp. 169–170). Strengthened for the first half (systematic evidence). Supported but unmeasured for the second (bans spread; sequestration works only under public control).

Lessons the hindsight record adds: - Symmetry of scrutiny. The breast-cancer dispute shows that bodies aligned with public health can also overreach on uncertain evidence. The chapter’s diagnostic questions test provenance, not truth, so they should be paired with independent weight-of-evidence review applied to all sides. - Evidence tools cut both ways. Impact assessment helped industry delay the Tobacco Products Directive, and it helped the CJEU uphold it. Who controls the tool, and whether it is applied symmetrically, matters more than the tool. - Disclosure is the remedy that influence-seeking actors prefer. Where disclosure is offered as the alternative to structural separation, the chapter’s warning applies with force.

Method and access notes#

• The web-search quota (200 calls) was already exhausted when this check began, so no general web searches were run.
• Literature was found and verified through the Europe PMC REST API (titles and abstracts; some full texts via PMC), Crossref and OpenAlex.
• Primary legal and regulatory texts came from the Federal Register API (EPA rules, EO 14303, OSHA withdrawal), CourtListener (court opinions and snippets) and the EU Publications Office (COM(2015) 215, COM(2021) 219, Reg. 2021/695, Protocol No 2, C-547/14, Reg. 2019/1381).
• Organisational pages were fetched directly: CDC, WHO, IARC (Monograph 100E PDF), CMS, ICMJE (PDF), FDA, European Commission, ERIF, LobbyFacts, STOP, ANRF (PDF), BBC Trust (PDF).
• Blocked and not bypassed:
• NCBI Bookshelf (reCAPTCHA), so the 2006 and 2014 Surgeon General reports are cited via IARC 100E, CDC and secondary literature;
• HHS and CDC Stacks (403);
• bmj.com, ScienceDirect, Nature and ATS journal pages (403, redirect or login);
• the EU Transparency Register’s own interface (JavaScript-only), so LobbyFacts’ mirror of the registration was used.
• Items marked “[title only]” or “[not re-fetched]” are used only for existence or direction, not for detailed claims.
• Advocacy sources (ANRF, STOP, ERIF) are identified as such. ERIF’s claims about its own influence are self-reports.
• Independence of the evidence. Much of the corroborating literature for Claims 1, 2, 5 and 9 involves the chapter’s author or panel authors, or their close collaborators (Bero, Gilmore, Glantz, Malone). Independent corroboration comes mainly from courts (Claim 6), Cochrane methodology reviews with pre-specified methods (Claim 2), regulators and agencies (Claim 3), and the EU and US legal record (Claims 4 and 5).
• Not checked:
• Table 7.2’s historical entries;
• the Center for Indoor Air Research and ventilation-standard details (pp. 156–157, 162);
• the ISO and CORESTA tar-yield story beyond the chapter;
• the current membership of ERIF;
• national smoke-free laws outside Ireland, the EU and the US.

Sources#

Chapter - Bero, L.A. (2013). Tobacco industry manipulation of research; with Panel 7.1 by Smith, K., Gilmore, A. and Fooks, G. In EEA, Late lessons from early warnings: science, precaution, innovation, EEA Report No 1/2013, pp. 151–178. https://www.eea.europa.eu/publications/late-lessons-2

Claim 1: constancy and cross-sector strategies - Legg, T., Hatchard, J. and Gilmore, A.B. (23 June 2021). The Science for Profit Model. PLoS One 16(6):e0253272. https://doi.org/10.1371/journal.pone.0253272 - Kearns, C.E., Schmidt, L.A. and Glantz, S.A. (2016; issue 1 Nov 2016). Sugar industry and coronary heart disease research. JAMA Internal Medicine 176(11):1680–1685. https://doi.org/10.1001/jamainternmed.2016.5394 - Kearns, C.E., Apollonio, D. and Glantz, S.A. (21 Nov 2017). Sugar industry sponsorship of germ-free rodent studies. PLoS Biology 15(11):e2003460. https://doi.org/10.1371/journal.pbio.2003460 - Gaber, N., Bero, L. and Woodruff, T.J. (1 June 2023). The Devil they Knew: chemical documents analysis of industry influence on PFAS science. Annals of Global Health 89(1):37. https://doi.org/10.5334/aogh.4013 - Goldberg, R.F. and Vandenberg, L.N. (26 March 2021). The science of spin. Environmental Health 20:33. https://doi.org/10.1186/s12940-021-00723-0 - Supran, G. and Oreskes, N. (1 Aug 2017). Assessing ExxonMobil’s climate change communications (1977–2014). Environmental Research Letters 12:084019. https://doi.org/10.1088/1748-9326/aa815f - Supran, G., Rahmstorf, S. and Oreskes, N. (13 Jan 2023). Assessing ExxonMobil’s global warming projections. Science. https://doi.org/10.1126/science.abk0063 - Legg, T., Clift, B. and Gilmore, A.B. (2024). Document analysis of the Foundation for a Smoke-Free World’s scientific outputs and activities. Tobacco Control 33(4):525–534. https://doi.org/10.1136/tc-2022-057667 - Braznell, S., Laurence, L., Fitzpatrick, I. and Gilmore, A.B. (2025). “Keep it a secret”: leaked documents suggest Philip Morris International, and its Japanese affiliate, continue to exploit science for profit. Nicotine & Tobacco Research 27(5):794–804. https://doi.org/10.1093/ntr/ntae101 - Ulucanlar, S., Fooks, G.J. and Gilmore, A.B. (Sept 2016). The policy dystopia model. PLoS Medicine 13(9):e1002125. https://doi.org/10.1371/journal.pmed.1002125 - Gilmore, A.B. et al. (23 March 2023). Defining and conceptualising the commercial determinants of health. Lancet 401:1194–1213. https://doi.org/10.1016/S0140-6736(23)00013-2 - Consortium of the Center to End Corporate Harm, UCSF (March 2026). Corporate vectors of chronic disease: using internal industry documents to craft counterstrategies. NEJM 394(12):1231–1237. https://doi.org/10.1056/NEJMms2507028 [title only]

Claim 2: sponsorship bias - Barnes, D.E. and Bero, L.A. (1998). Why review articles on the health effects of passive smoking reach different conclusions. JAMA 279:1566–1570. https://doi.org/10.1001/jama.279.19.1566 [cited by the chapter; not re-read] - Lundh, A., Lexchin, J., Mintzes, B., Schroll, J.B. and Bero, L. (16 Feb 2017). Industry sponsorship and research outcome. Cochrane Database of Systematic Reviews MR000033.pub3. https://doi.org/10.1002/14651858.MR000033.pub3 - Hansen, C., Lundh, A., Rasmussen, K. and Hróbjartsson, A. (5 Aug 2019). Financial conflicts of interest in systematic reviews. Cochrane Database of Systematic Reviews MR000047.pub2. https://doi.org/10.1002/14651858.MR000047.pub2 - Nejstgaard, C.H. et al. (9 Dec 2020). Association between conflicts of interest and favourable recommendations. BMJ 371:m4234. https://doi.org/10.1136/bmj.m4234 - Pisinger, C., Godtfredsen, N. and Bender, A.M. (2019; online 18 Dec 2018). A conflict of interest is strongly associated with tobacco industry-favourable results, indicating no harm of e-cigarettes. Preventive Medicine 119:124–131. https://doi.org/10.1016/j.ypmed.2018.12.011 - Bes-Rastrollo, M. et al. (31 Dec 2013). Financial conflicts of interest and reporting bias regarding sugar-sweetened beverages and weight gain. PLoS Medicine 10(12):e1001578. https://doi.org/10.1371/journal.pmed.1001578 - Mandrioli, D., Kearns, C.E. and Bero, L.A. (8 Sept 2016). Relationship between research outcomes and risk of bias, study sponsorship, and author financial conflicts of interest in reviews of artificially sweetened beverages. PLoS One 11(9):e0162198. https://doi.org/10.1371/journal.pone.0162198 - Chartres, N., Fabbri, A. and Bero, L.A. (Dec 2016). Association of industry sponsorship with outcomes of nutrition studies. JAMA Internal Medicine 176(12):1769–1777. https://doi.org/10.1001/jamainternmed.2016.6721 - Chartres, N. et al. (4 Dec 2020). Association of food industry ties with findings of studies on dairy intake and cardiovascular disease. BMJ Open 10:e039036. https://doi.org/10.1136/bmjopen-2020-039036 - Fabbri, A., Lai, A., Grundy, Q. and Bero, L.A. (Nov 2018). The influence of industry sponsorship on the research agenda: a scoping review. American Journal of Public Health 108(11):e9–e16. https://doi.org/10.2105/AJPH.2018.304677

Claim 3: health effects of SHS - IARC (2012). Monographs Vol. 100E, Personal habits and indoor combustions, “Second-hand tobacco smoke” (evaluation, s. 5; breast cancer, s. 2.2). https://publications.iarc.who.int/download/mono100E-7_new.pdf ; volume page: https://publications.iarc.who.int/Book-And-Report-Series/Iarc-Monographs-On-The-Identification-Of-Carcinogenic-Hazards-To-Humans/Personal-Habits-And-Indoor-Combustions-2012 - US CDC. Health problems caused by secondhand smoke (page last updated 31 Jan 2025; accessed 25 Sept 2026). https://www.cdc.gov/tobacco/secondhand-smoke/health.html - US DHHS (2014). The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. https://www.ncbi.nlm.nih.gov/books/NBK179276/ [not fetchable: reCAPTCHA] - WHO. Tobacco fact sheet (version dated 26 June 2026; accessed 25 Sept 2026). https://www.who.int/news-room/fact-sheets/detail/tobacco - Öberg, M. et al. (2011). Worldwide burden of disease from exposure to second-hand smoke. Lancet 377:139–146. https://doi.org/10.1016/S0140-6736(10)61388-8 - Frazer, K. et al. (2016). Legislative smoking bans for reducing harms from secondhand smoke exposure. Cochrane Database of Systematic Reviews CD005992.pub3. https://doi.org/10.1002/14651858.CD005992.pub3 - Stallings-Smith, S. et al. (24 April 2013). Reductions in mortality following the national Irish smoking ban. PLoS One 8(4):e62063. https://doi.org/10.1371/journal.pone.0062063 - Shetty, K.D. et al. (2011; online Nov 2010). Changes in U.S. hospitalization and mortality rates following smoking bans. Journal of Policy Analysis and Management 30(1):6–28. https://doi.org/10.1002/pam.20548 - Glantz, S.A. and Johnson, K.C. (Jan 2014). The Surgeon General report on smoking and health 50 years later: breast cancer and the cost of increasing caution. Cancer Epidemiology, Biomarkers & Prevention 23(1):37–46. https://doi.org/10.1158/1055-9965.EPI-13-1081 - Johnson, K.C. and Glantz, S.A. (2008). Evidence secondhand smoke causes breast cancer in 2005 stronger than for lung cancer in 1986. Preventive Medicine 46(6):492–496. https://doi.org/10.1016/j.ypmed.2007.11.016 - Possenti, I. et al. (28 June 2024). Exposure to second-hand smoke and breast cancer risk in non-smoking women. British Journal of Cancer 131:1116–1125. https://doi.org/10.1038/s41416-024-02732-5 - Macacu, A. et al. (Nov 2015). Active and passive smoking and risk of breast cancer: a meta-analysis. Breast Cancer Research and Treatment 154:213–224. https://doi.org/10.1007/s10549-015-3628-4 - Yang, Y. et al. (18 Oct 2013). Lack of an association between passive smoking and incidence of female breast cancer in non-smokers. PLoS One 8(10):e77029. https://doi.org/10.1371/journal.pone.0077029 - Wada, K. et al. (April 2024). Active and passive smoking and breast cancer in Japan: pooled analysis of nine cohorts. International Journal of Epidemiology 53(3):dyae047. https://doi.org/10.1093/ije/dyae047

Claim 4: US “sound science” rules and the EU contrast - US EPA (30 April 2018). Strengthening Transparency in Regulatory Science, proposed rule, 83 FR 18768. https://www.federalregister.gov/documents/2018/04/30/2018-09078/strengthening-transparency-in-regulatory-science - US EPA (6 Jan 2021). Strengthening Transparency in Pivotal Science Underlying Significant Regulatory Actions and Influential Scientific Information, final rule, 86 FR 469. https://www.federalregister.gov/documents/2021/01/06/2020-29179/strengthening-transparency-in-pivotal-science-underlying-significant-regulatory-actions-and - US EPA (2 June 2021). Removal of the vacated rule, 86 FR 29515 (recounts EDF v. EPA, D. Mont., 27 Jan and 1 Feb 2021). https://www.federalregister.gov/documents/2021/06/02/2021-11317/strengthening-transparency-in-pivotal-science-underlying-significant-regulatory-actions-and - Executive Order 14303, Restoring Gold Standard Science (signed 23 May 2025; published 29 May 2025), 90 FR 22601. https://www.federalregister.gov/documents/2025/05/29/2025-09802/restoring-gold-standard-science - US EPA (25 Aug 2026). Glyphosate open literature search, notice of availability, 91 FR 54870. https://www.federalregister.gov/documents/2026/08/25/2026-17301/glyphosate-open-literature-search-to-inform-human-health-risk-assessment-notice-of-availability - Goobie, G.C., Rice, M.B. and Carlsten, C. (Oct 2021). The EPA’s “Strengthening Transparency in Pivotal Science” rule: don’t let history repeat itself. Annals of the American Thoracic Society 18(10):1614–1617. https://doi.org/10.1513/AnnalsATS.202103-259VP - Mansfield, B. (2021; online Nov 2020). Deregulatory science: chemical risk analysis in Trump’s EPA. Social Studies of Science 51(1):28–50. https://doi.org/10.1177/0306312720970284 - Thorp, H.H. et al. (26 Nov 2019). Joint statement on EPA proposed rule and public availability of data. PLoS Medicine 16(11):e1003014 (also in Science, Nature, Cell, PNAS, Lancet). https://doi.org/10.1371/journal.pmed.1003014 [title and bibliographic record] - Carter, J., Michaels, D. and Wagner, W. (10 June 2026). Trump’s gold standard science is harming US science and health. BMJ 393:e555297. https://doi.org/10.1136/bmj-2026-555297 [title only] - Thorp, H.H. (18 Sept 2025). Gold standard science requires gold standard scholarship. Science 389:1165. https://doi.org/10.1126/science.aec2360 - Regulation (EU) 2019/1381 of 20 June 2019 on the transparency and sustainability of the EU risk assessment in the food chain (Art. 32b; recital 28; applies from 27 March 2021). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019R1381

Claim 5: EU Better Regulation and the European Risk Forum - Smith, K.E., Fooks, G., Gilmore, A.B., Collin, J. and Weishaar, H. (Feb 2015). Corporate coalitions and policy making in the European Union: how and why British American Tobacco promoted “Better Regulation”. Journal of Health Politics, Policy and Law 40(2):325–372. https://doi.org/10.1215/03616878-2882231 - Peeters, S., Costa, H., Stuckler, D., McKee, M. and Gilmore, A.B. (2016; online 24 Feb 2015). The revision of the 2014 European tobacco products directive: the tobacco industry’s attempts to “break the health silo”. Tobacco Control 25(1):108–117. https://doi.org/10.1136/tobaccocontrol-2014-051919 - Costa, H. et al. (2014; online 13 Aug 2014). Quantifying the influence of the tobacco industry on EU governance. Tobacco Control 23(6):473–478. https://doi.org/10.1136/tobaccocontrol-2014-051822 - Lauber, K. and Brooks, E. (14 Sept 2023). Why meta-regulation matters for public health: the case of the EU better regulation agenda. Globalization and Health 19:70. https://doi.org/10.1186/s12992-023-00971-4 - Radaelli, C.M. (2023). Occupy the semantic space! Opening up the language of better regulation. Journal of European Public Policy 30(9):1860–1883. https://doi.org/10.1080/13501763.2023.2181852 - European Commission (19 May 2015). Better regulation for better results, COM(2015) 215 final. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52015DC0215 - European Commission (29 April 2021). Better regulation: joining forces to make better laws, COM(2021) 219 final. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021DC0219 - Regulation (EU) 2021/695 establishing Horizon Europe, recital 6. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32021R0695 - Protocol (No 2) on the application of the principles of subsidiarity and proportionality, Art. 5 (consolidated 2016). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:12016E/PRO/02 - CJEU, Case C-547/14 Philip Morris Brands SARL and Others, judgment of 4 May 2016. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62014CJ0547 - European Regulation and Innovation Forum. About us; Our governance; Key messages 2024–2025 (accessed 25 Sept 2026). https://www.eriforum.eu/about-us.html ; https://www.eriforum.eu/our-governance.html ; https://www.eriforum.eu/keymessages.html - LobbyFacts. European Regulation and Innovation Forum, EU Transparency Register ID 52996964558-47 (registration as of 12 Jan 2026; accessed 25 Sept 2026). https://www.lobbyfacts.eu/datacard/european-regulation-and-innovation-forum?rid=52996964558-47

Claim 6: delay, courts, smoke-free laws - Flue-Cured Tobacco Cooperative Stabilization Corp. v. EPA, 4 F. Supp. 2d 435 (M.D.N.C., 17 July 1998). https://www.courtlistener.com/opinion/2349856/flue-cured-tobacco-cooperative-stabilization-corp-v-united-states/ - Flue-Cured Tobacco Cooperative Stabilization Corp. v. EPA, 313 F.3d 852 (4th Cir., 11 Dec 2002). https://www.courtlistener.com/opinion/2967697/flue-cured-tobacco-cooperative-stabilization-corp-v-united-states/ - United States v. Philip Morris USA Inc., 566 F.3d 1095 (D.C. Cir., 22 May 2009). https://www.courtlistener.com/opinion/187362/united-states-v-philip-morris-usa-inc/ - United States v. Philip Morris USA Inc., 907 F. Supp. 2d 1 (D.D.C., 27 Nov 2012) (corrective statement topics). https://www.courtlistener.com/opinion/2661478/united-states-v-philip-morris-usa-inc/ - United States v. Philip Morris USA Inc., 801 F.3d 250 (D.C. Cir., 22 May 2015). https://www.courtlistener.com/opinion/2802952/united-states-v-philip-morris-usa-inc/ - United States v. Philip Morris USA Inc., 257 F. Supp. 3d 1 (D.D.C., 27 June 2017). https://www.courtlistener.com/opinion/4404055/united-states-v-philip-morris-usa-inc/ - Smith, P. et al. (June 2011). Correcting over 50 years of tobacco industry misinformation. American Journal of Preventive Medicine 40(6):690–698. https://doi.org/10.1016/j.amepre.2011.01.020 - Blake, K.D., Willis, G. and Kaufman, A. (2020; online Sept 2019). Population prevalence of exposure to court-ordered corrective statements. Tobacco Control 29(5):516–521. https://doi.org/10.1136/tobaccocontrol-2019-055095 - Gibson, L.P., Kaufman, A.R. and Blake, K.D. (2 April 2026). Point-of-sale, court-ordered corrective statements: exposure and reactions. Tobacco Control. https://doi.org/10.1136/tc-2025-059590 - OSHA (17 Dec 2001). Indoor Air Quality, withdrawal of proposal, 66 FR 64946. https://www.federalregister.gov/documents/2001/12/17/01-31165/indoor-air-quality - American Nonsmokers’ Rights Foundation (1 July 2026). Summary of 100% smokefree state laws and population protected (advocacy source). https://no-smoke.org/wp-content/uploads/pdf/SummaryUSPopList.pdf - European Commission. Smoke-free environments (2009 and 2024 Council Recommendations; accessed 25 Sept 2026). https://health.ec.europa.eu/tobacco/smoke-free-environments_en ; Council Recommendation on smoke- and aerosol-free environments (3 Dec 2024): https://health.ec.europa.eu/publications/council-recommendation-smoke-and-aerosol-free-environments_en

Claim 7: media - Koehler, D.J. (2016). Can journalistic “false balance” distort public perception of consensus in expert opinion? Journal of Experimental Psychology: Applied 22(1):24–38. https://doi.org/10.1037/xap0000073 - Dixon, G.N. and Clarke, C.E. (2013; online 13 Sept 2012). Heightening uncertainty around certain science. Science Communication 35(3):358–382. https://doi.org/10.1177/1075547012458290 [bibliographic record only] - Imundo, M.N. and Rapp, D.N. (June 2022). When fairness is flawed: effects of false balance reporting and weight-of-evidence statements. Journal of Applied Research in Memory and Cognition. https://doi.org/10.1016/j.jarmac.2021.10.002 [bibliographic record only] - Brüggemann, M. and Engesser, S. (Jan 2017). Beyond false balance: how interpretive journalism shapes media coverage of climate change. Global Environmental Change. https://doi.org/10.1016/j.gloenvcha.2016.11.004 [title only] - BBC Trust (July 2014). Trust conclusions on the Executive report on science impartiality review actions. http://downloads.bbc.co.uk/bbctrust/assets/files/pdf/our_work/science_impartiality/trust_conclusions.pdf

Claim 8: disclosure - ICMJE. Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals (updated January 2026). https://www.icmje.org/icmje-recommendations.pdf ; Disclosure of Interest form (updated February 2021): https://www.icmje.org/disclosure-of-interest/ - CMS. Open Payments (program year 2025 data; accessed 25 Sept 2026). https://www.cms.gov/priorities/key-initiatives/open-payments - Taheri, C. et al. (2021). Discrepancies in self-reported financial conflicts of interest disclosures by physicians: a systematic review. BMJ Open 11:e045306. https://doi.org/10.1136/bmjopen-2020-045306 - Cherla, D.V. et al. (March 2019). The effect of financial conflict of interest, disclosure status, and relevance on medical research. Journal of General Internal Medicine 34(3):429–434. https://doi.org/10.1007/s11606-018-4784-0 - Graham, S.S. et al. (14 May 2025). Research on policy mechanisms to address funding bias and conflicts of interest in biomedical research: a scoping review. Research Integrity and Peer Review 10:6. https://doi.org/10.1186/s41073-025-00164-0 - Loewenstein, G., Sah, S. and Cain, D.M. (2012). The unintended consequences of conflict of interest disclosure. JAMA 307(7):669–670. https://doi.org/10.1001/jama.2012.154 [title only] - Sah, S., Loewenstein, G. and Cain, D.M. (2013). The burden of disclosure. Journal of Personality and Social Psychology 104(2):289–304. https://doi.org/10.1037/a0030527 - Sah, S. and Loewenstein, G. (2014). Nothing to declare. Psychological Science 25(2):575–584. https://doi.org/10.1177/0956797613511824 - Rosenbaum, L. (May 2015). Conflicts of interest, part 1: reconnecting the dots. NEJM 372:1860–1864. https://doi.org/10.1056/NEJMms1502493 [title only] - Babor, T.F. and Miller, P.G. (2014). McCarthyism, conflict of interest and Addiction’s new transparency declaration procedures. Addiction 109:341–344. https://doi.org/10.1111/add.12384 [title only]

Claim 9: funding bans and sequestration - Godlee, F. et al. (15 Oct 2013). Journal policy on research funded by the tobacco industry. BMJ 347:f5193. https://doi.org/10.1136/bmj.f5193 [title and record; text blocked] - Malone, R.E. (Jan 2013). Changing Tobacco Control’s policy on tobacco industry-funded research. Tobacco Control 22:1–2. https://doi.org/10.1136/tobaccocontrol-2012-050874 [title only] - Arany, P. et al. (May 2025). The IADR and AADOCR policy statement on tobacco industry-funded research. Journal of Dental Research 104(8):817–819. https://doi.org/10.1177/00220345251326085 [title only] - Hopkinson, N.S. (online 23 Dec 2021). Vectura and Philip Morris: the leopard has not changed its spots. Thorax 77(6):537–538. https://doi.org/10.1136/thoraxjnl-2021-218328 [title only] - Capstick, T.G.D. and Hopkinson, N.S. (2021). Adapting inhaled medication practice in COPD and asthma to avoid funding the tobacco industry. International Journal of COPD. https://doi.org/10.2147/COPD.S337066 [title only] - van der Eijk, Y., Bero, L.A. and Malone, R.E. (2019). Philip Morris International-funded “Foundation for a Smoke-Free World”: analysing its claims of independence. Tobacco Control 28(6):712–718. https://doi.org/10.1136/tobaccocontrol-2018-054278 - Yach, D. (2019). Foundation for a Smoke-Free World: independent and making progress. Lancet 394:1008. https://doi.org/10.1016/S0140-6736(19)31892-6 [title only] - Legg, T., Legendre, M. and Gilmore, A.B. (online 28 April 2021). Paying lip service to publication ethics: scientific publishing practices and the Foundation for a Smoke-Free World. Tobacco Control 30(e1):e65–e72. https://doi.org/10.1136/tobaccocontrol-2020-056003 [title only] - Cohen, J.E., Eissenberg, T. and Zeller, M. (2025). PMI-funded Foundation for a Smoke-Free World: application of criteria to assess funding models. Tobacco Control 34(2):248–250. https://doi.org/10.1136/tc-2023-058474 - Polosa, R. and Crawley, F.P. (2017). Scientific and ethical obligations to publish tobacco industry-funded research on nicotine delivery systems of reduced risk. Toxicology 390:61–62. https://doi.org/10.1016/j.tox.2017.09.003 [title only] - WHO FCTC. Guidelines for implementation of Article 5.3 (adopted November 2008). https://fctc.who.int/resources/publications/m/item/guidelines-for-implementation-of-article-5.3 - STOP (2025). Global Tobacco Industry Interference Index 2025 (advocacy source). https://exposetobacco.org/global-index/ - US FDA. Tobacco user fees (accessed 25 Sept 2026). https://www.fda.gov/tobacco-products/manufacturing/tobacco-user-fees - Perry, C.L. et al. (2020; online 23 April 2019). Research on youth and young adult tobacco use, 2013–2018, from the FDA–NIH Tobacco Centers of Regulatory Science. Nicotine & Tobacco Research 22(7):1063–1076. https://doi.org/10.1093/ntr/ntz059 [title only]