Hindsight check: LL1-00 (Preface, acknowledgements, contents, Ch1 Introduction)#

Source section: EEA, Late lessons from early warnings: the precautionary principle 1896–2000, Environmental Issue Report No 22 (2001), report pp. 1–16. Check window: publication (2001) to late September 2026. Checked: 25 September 2026.

Method note. This check was done in two passes. - First pass. General web search was unavailable (the session’s search budget was exhausted). Sources were retrieved by fetching known primary repositories directly: HSE official statistics, Europe PMC and PubMed records, Crossref, EUR-Lex, the WTO dispute pages, the US Federal Register, APHIS, FDA, EPA, WHO, the UNEP Ozone Secretariat, NOAA/WMO, and the EEA’s own site. Several pages refused automated retrieval: parts of EUR-Lex (a bot challenge), CURIA, OECD, UNECE and the UK vCJD surveillance unit. - Second pass (same day). Web search was available again. I used it to close four items the first pass had left open: - the post-2024 status of the US chrysotile rule - the Rotterdam Convention’s chrysotile listing history - whether Hansen et al.’s 88 alleged false positives include the two candidates named in 2001 - what the 2013 volume’s annex says about the 2001 asbestos and CFC cases

The second pass also found contemporaneous (2001) and 2005 Eurobarometer trust data. I re-checked every HSE figure against the MESO01 spreadsheet (published 1 July 2026).

Where a point rests on a document I could not open, or on a secondary source, I say so.

The 2013 volume. The EEA’s own links to the full 2013 report now return 404 or 410. I used the 48-page summary from the EEA site, the Chapter 2 PDF (still downloadable from the EEA), and a copy of the full report hosted by DTU’s research repository (orbit.dtu.dk). In the DTU copy, the annex that summarises the fourteen 2001 cases is Annex 2, “An overview of Late lessons from early warnings … vol. 1” (report pp. 731–746). This brief calls it Annex 3. The DTU copy has no Annex 3, so the numbering may differ between versions.

Overview#

This section is front matter, but it makes empirical claims, forecasts and framework claims that later chapters build on. Twenty-five years on, the picture is mixed.

The forecasts about harm held, and some understated it. - “Many thousands” of European mesothelioma deaths was an understatement. Great Britain alone recorded 54,900 mesothelioma deaths from 2001 to 2024 (HSE MESO01, July 2026). Europe carried about half of the world’s roughly 30,000 annual mesothelioma cases and deaths in 2022 (Fu et al. 2026). - The ozone-related skin cancer “pipeline” is real and still running. Later modelling puts its size above “some thousands”, although most of the excess is in non-fatal skin cancers.

The specific numbers the section cites were loose. - The 250,000–400,000 range is attributed to Peto (1999), but Peto’s abstract gives about 250,000 male mesothelioma deaths in 1995–2029. - Early projections of the size and timing of the peak overshot. Pelucchi et al. (2004) found observed deaths in France, Germany and Italy below Peto’s projections. Peto’s 1995 GB projection of a male peak of 2,700–3,300 deaths around 2020 compares with an actual male average of 2,091 a year in 2011–2020. - The EEA did not correct the number. Its 2013 volume restated it as “some 400 000 expected deaths in Europe over the next three decades” from mesothelioma alone (Annex 2, p. 736).

The transatlantic framing held in substance, but Table 1.1 contains factual errors. - Later systematic work supports the claim that the US practises precaution under other names (Hammitt et al. 2005; Wiener & Rogers 2002, as summarised by the European Parliament’s research service, EPRS, in 2015). The Delaney principle was still being applied in January 2025. - The EU stilbene (DES) ban dates from 1981 (Directive 81/602/EEC), not 1987. - The Delaney Clause dates from 1958 (1960 for colour additives). It was narrowed in 1996, not ended. - The claim that a 1970s scrapie measure “may have helped” the US avoid BSE has no support in FDA’s own account, which starts with the 1997 feed ban. Its mechanism was also doubtful: the UK BSE Inquiry (October 2000) had favoured a new prion mutation “in cattle, or possibly sheep” in the early 1970s, rather than scrapie, as the origin of BSE.

The claim that the 2000 Commission Communication helped avoid trade disputes did not hold. - The US opened the GMO (“biotech”) dispute in 2003, and the EC lost it in 2006. - The hormone-beef dispute was never resolved on the science. It was managed through a 2009 memorandum of understanding (MOU) and a 2019 quota deal, and the US still holds its WTO retaliation authorisation. - The Communication did become the reference point for precaution in EU law and case law.

The framework’s more contestable claims remain contested. - False positives are “the smaller” risk. A 2007 systematic search found only 4 genuine false positives among 88 claimed cases, and the 2013 volume repeated this. - That search required “high confidence” that a risk did not exist before counting a case as a false positive. Cox (2007) argued its criteria were biased towards calling risks “real”. - Neither of the report’s own two candidates (North Sea sludge dumping, Y2K) was among the 88. - No independent test settles the matter. - Precaution stimulates innovation. The Porter-hypothesis literature is mixed. The most comprehensive meta-analysis found statistical insignificance to be the most likely result, with positive effects more likely at country level (Cohen & Tubb 2018). The EEA’s own 2013 summary of the ozone case concedes that the phase-out produced incremental chemical substitution, and that the political process “failed to stimulate more radical changes”. - Political will matters more than information. This is supported for asbestos by the long gap between scientific consensus and bans. - Twenty years of blocked chrysotile listing under the Rotterdam Convention (2005–2025) add further support. - So does a US ban that EPA moved to reconsider within fifteen months and that is still in litigation. - The claim is complicated by evidence that countries ban asbestos mainly once they identify mesothelioma in their own mortality data (Chimed-Ochir et al. 2022). That points to locally owned information being decisive. - Trust. The EU’s own survey data from 2001 show politicians trusted very little (6.6% esteem) but scientists relatively highly (44.9% esteem; the most trusted source after a local disaster). So the Preface’s “very low” trust in scientists holds only if it is read as referring to scientists in regulatory roles.

Implication for weight. The section’s structural lessons have strengthened: - long latency creates harm “pipelines” (p. 3) - action can precede mechanism (pp. 14–15) - one label covers very different commitments (Table 1.2) - the costs of error must be weighed both ways (pp. 13–14)

Its comparative and quantitative assertions should be cited with the corrections below, not at face value. The 2013 follow-up volume carried some of them forward without correction, so it cannot be relied on to have fixed them.

Claim-by-claim#

Claim 1: UK asbestos death rate and western European projection (p. 11)#

Original claim (p. 11). “The current asbestos-induced death rate in the United Kingdom is about 3 000 deaths per year, and some 250 000–400 000 asbestos cancers are expected in western Europe over the next 35 years, due to past exposures (Peto, 1999).”

Subsequent developments

What Peto 1999 actually projected. The abstract projects that male mesothelioma deaths in western Europe would “almost double over the next 20 years, from 5000 in 1998 to about 9000 around 2018, and then decline, with a total of about a quarter of a million deaths over the next 35 years” (Peto et al. 1999, Br J Cancer 79:666–672). - The figure is male mesothelioma only, for 1995–2029. - The report’s upper bound of 400,000, and its wording “asbestos cancers”, are not in the abstract. - The range may add asbestos-related lung cancers. HSE now treats the lung-cancer-to-mesothelioma ratio as uncertain: estimates run from 0.55 to 1.3, with about 1 as the working assumption (HSE 2026b). On that basis a combined total would be higher, not lower. - So the citation is imprecise, but the order of magnitude is not obviously inflated.

The EEA did not correct the figure in 2013; it hardened it. The 2013 volume’s annex summarising the 2001 asbestos case says mesothelioma “is now causing the bulk of current and future harm from asbestos, including some 400 000 expected deaths in Europe over the next three decades” (EEA 2013, Annex 2, p. 736). - The 2001 range was “250 000–400 000 asbestos cancers” over 35 years. The 2013 annex keeps only the upper bound, calls it mesothelioma deaths specifically, and shortens the period to three decades. - It gives no new source, and it does not mention the downward revisions below, although Pelucchi et al. had been published nine years earlier.

Observed deaths came in below projection. Pelucchi et al. (2004) is in the British Journal of Cancer (90:1022–1024), not Annals of Oncology as the reader’s flag says. They found that “for the period 1995–1999, the observed male deaths for all three countries considered were lower than predicted”: Germany by 2.3%, Italy by 6.4% and France by 11.4%. They concluded that the 1995–2029 European total “is likely to be lower than the 250 000 previously estimated”.

The GB experience shows the same pattern. - Peto et al. (1995, Lancet 345:535–539) projected a male peak “in about the year 2020 of between 2700 and 3300 deaths”. - Hodgson et al. (2005, Br J Cancer 92:587–593) revised this to a peak of “around 1950 to 2450 deaths per year between 2011 and 2015”. - The actual male average was 2,091 a year in 2011–2020, and total GB deaths peaked at 2,606 in 2016 (HSE MESO01, published 1 July 2026). The refined 2005 projection was accurate; the 1995 projection was not.

The UK “3,000 deaths” figure versus today’s “about 5,000”. - GB mesothelioma deaths were 1,541 (1998), 1,615 (1999) and 1,633 (2000) (HSE MESO01). So the report’s 3,000 must include lung cancer and asbestosis on some ratio. - HSE’s current estimate is “around 5,000 asbestos-related disease deaths per year, including mesothelioma, lung cancer and asbestosis”: 2,146 mesothelioma deaths in 2024, a similar estimated number of lung cancers, and 503 deaths mentioning asbestosis (HSE 2026b). - The rise from about 3,000 to about 5,000 is therefore partly real (mesothelioma rose by about 60% from 2000 to the 2016 peak) and partly a change of method. The two figures should not be compared directly.

Where the epidemic is now. - HSE says deaths “will continue to decline during the 2020s and 2030s”. It attributes most current deaths to “widespread industrial use of asbestos during 1950–1980” and cites lung-burden research showing exposures “reduced rapidly during the 1970s” (HSE 2026a). - Globally, the WHO estimates more than 200,000 deaths a year from occupational asbestos exposure (2016 estimates; WHO fact sheet, 11 August 2026). - Using the Global Burden of Disease 2021 data, Li et al. (2025) estimate 216,535 deaths from asbestos-related thoracic cancers in 2021, of which 29,619 were mesothelioma.

Verdict: partly held up. The scale and the “pipeline” logic held. The citation is imprecise: the range goes beyond Peto’s abstract and mixes endpoints. The best projections available in 2001 overstated how large and how late the peak would be. The UK figure used a different basis from today’s HSE estimate. The EEA’s 2013 restatement moved further from the evidence, not closer.

Implications for weight. The qualitative lesson is strengthened: long latency means decades of committed harm after exposure falls (pp. 3, 11). Anyone reusing the numbers should cite Peto correctly (about 250,000 male mesotheliomas, 1995–2029) and note that observed outcomes came in lower. The 2013 annex’s “some 400 000” should not be cited.

The episode also shows something about the reports themselves. Headline numbers were carried forward without being updated against later evidence, even by the organisation that published them. Each volume’s quantitative claims therefore need checking against the primary literature.

A general methodological lesson also follows: early projections of latent-harm epidemics can be wrong in both size and timing, even when the direction is right. That cuts against overconfident forecasting on either side.

Claim 2: Forward projections of skin cancers and mesothelioma deaths (p. 3)#

Original claim (Preface, p. 3). “Over the next 50 years we will see some thousands of extra skin cancers as today’s children grow up exposed to the higher levels of ultraviolet radiation … Over the same period many thousands of Europeans will die from … mesothelioma.” Both hazards had “such long latent periods” that “‘pipelines’ of unstoppable consequences, decades long, were set in place”.

Subsequent developments

Mesothelioma: the projection has been exceeded well within the 50 years. - GB alone recorded 54,900 mesothelioma deaths from 2001 to 2024 (HSE MESO01). HSE also reports 45,330 male and 9,517 female mesothelioma cancer registrations in GB for 2001–2022 (HSE 2026a, Annex 2). - Using GLOBOCAN 2022, Fu et al. (2026) estimate 30,633 new mesothelioma cases worldwide in 2022. “Europe bore the highest burden, accounting for 48.1% of global cases and 48.4% of deaths.” - Incidence is now falling in the UK, Germany, the Netherlands, Norway and others, but rose in 2003–2017 among men in Croatia and Slovenia (Fu et al. 2026).

Skin cancer: the direction held; “some thousands” looks low for incidence. - van Dijk et al. (2013, Photochem Photobiol 89:234–246) modelled the skin cancer risk “inevitably committed once ozone depletion was recognized”. Excess incidence “will peak mid 21st century and then recover”. Under full compliance with the Montreal Protocol, the excess is “30–40” extra cases per million people per year in western Europe, and “90–100” around the Mediterranean. - My rough inference: applied to a western European population of several hundred million, this implies excess cases on the order of ten thousand a year at the peak. Most would be keratinocyte (non-melanoma) cancers with low fatality. - Lee-Taylor et al. (2026, GeoHealth 10:e2024GH001283) estimate for the US and Australia that ozone-depleting-substance impacts “peak several decades later at ca. … 1% for melanoma incidence, and 5% for keratinocyte cancer incidence (relative to 1980)”. They also find that differences in UV “can only partly explain” differences in national skin cancer rates, with behaviour and skin type mattering. - That makes the ozone component real but small against behavioural drivers, and hard to see directly in registry trends.

The pipeline is finite, and the counterfactual was far worse. - The 2022 WMO/UNEP assessment projects total column ozone returning to 1980 values around 2040 (60°N–60°S), 2045 (Arctic) and 2065 (Antarctic). - The US EPA estimates that full implementation of the Montreal Protocol will prevent “approximately 443 million cases of skin cancer, 2.3 million skin cancer deaths, and 63 million cases of cataracts” for people in the US born 1890–2100. For the world as a whole, van Dijk et al. (2013) estimate that by 2030 about 2 million skin cancer cases a year are prevented.

The regime was also tested after 2001. The 2022 assessment reports unexpected CFC-11 emissions rising from about 57 Gg/yr in 2012 to about 78 Gg/yr in 2017. A large fraction came from eastern China, and emissions fell back to about 45 Gg/yr in 2019–2020 after detection. Monitoring worked as the early-warning mechanism the report calls for.

Verdict: held up. The mesothelioma projection is strengthened: “many thousands” understates it. The skin cancer projection held in direction and is probably understated for incidence, though not for mortality.

Implications for weight. The “pipeline” insight (p. 3) is one of the section’s most robust transferable lessons: once exposure has occurred, harm is committed regardless of later action. The ozone case adds two points: - Decisive action, even late, can cap the pipeline. Recovery dates are now in view. - Continued monitoring is needed to detect non-compliance. This ties to Lesson 2’s call for long-term monitoring (p. 4 of the Preface; Ch 16).

Claim 3: Misplaced “certainty” about the absence of harm delayed action “in most of the case studies” (p. 4)#

Original claim (Preface, p. 4). “Misplaced ‘certainty’ about the absence of harm played a key role in delaying preventive actions in most of the case studies.”

Subsequent developments

A chapter-by-chapter check is outside this section’s hindsight remit and needs the chapter digests at synthesis. The report’s own contents list (pp. 7–10) is suggestive but mixed: - At least one case is explicitly framed around reassurance: Ch 15, “‘Mad cow disease’ 1980s–2000: how reassurances undermined precaution”. - At least one is framed around a contested precautionary action rather than misplaced certainty: Ch 14, “Hormones as growth promoters: the precautionary principle or a political risk assessment?”. - So “most”, not “all”, is the right reading, and it still needs checking.

The 2013 volume reframed the mechanism. Its summary emphasises “manufacturing doubt about the science supporting such warnings” rather than misplaced certainty (EEA 2013, pp. 10, 12). It reports that the nine new false-negative cases share “more than sufficient evidence for much earlier action” and “slow and sometimes obstructive behaviour by businesses”. Misplaced certainty about safety and manufactured doubt about harm are two faces of the same evidential asymmetry. The 2013 cases are again EEA-selected, with the same selection-on-outcome limit.

Independent support for the broader pattern. In a 198-country analysis, Chimed-Ochir et al. (2022, BMJ Global Health 7:e010553) found that “the odds of a country banning asbestos increased 14.1-fold” with national identification of mesothelioma, and by 26% per unit rise in the mesothelioma mortality rate in the 1–5 years before a ban. Countries have largely acted on their own demonstrated harm, not on evidence from elsewhere. That is consistent with the absence of local evidence being treated as a reason to wait.

A caution from the other side. Cox (2007, Risk Analysis 27:1083–1086) argues that precaution advocates can make the mirror-image error. On his account, they treat “highly uncertain risks as ‘real’ even when no real or potential harm has actually been demonstrated”. Misplaced certainty can run in either direction.

Verdict: unclear at this level; partly held up as a general pattern. Later evidence supports the broad pattern of delay while awaiting local or definitive proof. The specific “most of the case studies” count has not been independently verified.

Implications for weight. Treat this as a well-supported mechanism: institutions treat absence of evidence as evidence of absence, and demand local proof. It is not yet a verified frequency. The transferable form is technology-neutral: when evidence of harm is not yet local, is slow to accrue, or is actively contested, institutions default to the status quo (pp. 4, 15; cf. p. 13 on data without understanding).

Claim 4: Lack of political will mattered more than lack of trusted information (p. 4)#

Original claim (Preface, p. 4). “The absence of political will to take action to reduce hazards, in the face of conflicting costs and benefits, seems to be an even more important factor in these histories than is the availability of trusted information.” The editors “largely leave [this] to the reader”.

Subsequent developments

Asbestos supports it strongly. - The WHO states that “all forms of asbestos, including chrysotile, are carcinogenic to humans”. It recommends stopping all use, yet only “more than 50 WHO Member States” have legislated bans (WHO fact sheet, 11 August 2026). - The US banned chrysotile under TSCA only in 2024: final rule published 28 March 2024, effective 28 May 2024 (89 FR, FR Doc 2024-05972). That is decades after the health evidence was settled. - The ban has stayed politically fragile since 2024: - On 16 June 2025, EPA asked the Fifth Circuit to hold the industry challenge (Texas Chemistry Council v. EPA, No. 24-60193) in abeyance while it reconsidered the rule through new rulemaking. - On 7 July 2025, EPA withdrew that motion, saying it “no longer intends to conduct notice-and-comment rulemaking” and would instead “explore whether guidance could provide further clarity” (Bergeson & Campbell, 10 July 2025, reporting the court filing). - Before the hearing, two sets of industry challengers pulled out. Olin Corporation withdrew its challenge on 6 May 2026. On 20 May, the American Chemistry Council and allied petitioners withdrew their challenge to EPA’s risk evaluation and its scientific basis. - Oral argument was heard on 1 June 2026. Just before it, the court asked the parties to address whether the remaining petitioners had standing. The Justice Department attorney defending the rule told the court: “If EPA cannot ban asbestos, it’s unclear under what circumstances EPA could ever ban a chemical” (ADAO, 2 June 2026). - That court is the one whose ruling in Corrosion Proof Fittings v. EPA (1991) overturned most of EPA’s 1989 asbestos ban. - I found no decision as of late September 2026. The sources for this sequence are law-firm and advocacy (ADAO) reports, not the court docket, which I did not access. The oral-argument recording is posted by the court. - EPA’s rule page, last updated 5 May 2026, still describes the ban as in force, with compliance dates running to 2037. - The Rotterdam Convention shows the pattern internationally: - Its Chemical Review Committee recommended listing chrysotile in Annex III (the prior-informed-consent list) in 2005. - The Secretariat’s page records that the Conference of the Parties deferred or failed to reach consensus at every meeting from COP-3 (2006) to COP-11 (2023). COP-12 (April–May 2025) deferred it again (Ban Asbestos Network of India, May 2025, secondary). - Listing needs consensus. In 2017 the listing was blocked by India, Kazakhstan, Kyrgyzstan, Russia, Syria and Zimbabwe (IndustriALL, 3 May 2017, secondary). - According to Ban Asbestos India, the COP has accepted that chrysotile meets the listing criteria. The Secretariat’s page, as retrieved, records only the lack of consensus. - My reading: the obstacle for twenty years has been the economic interests of a few producing and using countries, not a shortage of trusted information. That is close to a textbook case of the Preface’s claim. - In fairness, the blocking governments present their position as a dispute about evidence, holding that chrysotile can be used safely under controls. For example, India has argued since 2005 that chrysotile “is not a hazardous chemical” (as reported by Ban Asbestos India). WHO rejects this: “all forms of asbestos, including chrysotile, are carcinogenic to humans” (WHO fact sheet, 11 August 2026), and IARC classifies all forms of asbestos as Group 1 carcinogens (IARC Monographs vol. 100C, 2012). My interpretation: contesting settled information can itself be how a lack of political will is expressed, which makes the Preface’s two factors hard to separate in practice.

The 2013 volume makes the point more systematically, but from the same stable. - Its chapter on business behaviour reports that “economic motives often drive non-precautionary business decisions. In virtually all reviewed cases it was perceived to be profitable for industries to continue” (EEA 2013 summary, ch. 25). It also stresses “political actions” to “deny or even suppress early warning signals”. - Gee’s chapter concludes that the capacity to forestall disasters “especially when such action is opposed by powerful economic and political interests, appears to be limited” (ch. 27).

Information still matters, at the right scale. Chimed-Ochir et al. (2022) found that identifying mesothelioma in national mortality data was the strongest predictor of an asbestos ban (odds up 14.1-fold). That points to locally owned information as a key trigger for political will. It fits the Preface’s own follow-on questions: “whose information is received? Is it ‘true, fair and independent’?” (p. 4).

Where information genuinely was the constraint. In some of the report’s cases, key hazards were not predicted by the science of the day. The Preface itself notes that ozone holes were a “surprise” (p. 4). In such cases the binding constraint was knowledge, not will.

Verdict: partly held up. Strong for asbestos, and consistent with the 2013 volume. Post-2001 asbestos governance (the Rotterdam deadlock and the fragile US ban) adds fresh examples where the evidence was settled and action still stalled. But “even more important” was never tested comparatively. The best later evidence suggests political will and information interact: will tends to follow locally owned, trusted evidence of harm.

Implications for weight. Use this as a hypothesis with good case support, not an established ranking. The durable, technology-neutral lesson: - Concentrated, near-term costs of action versus diffuse, deferred costs of inaction bias decisions towards delay (pp. 3–4). - Evidence generated or owned by the affected jurisdiction is a key lever for overcoming that bias.

Claim 5: Wider use of precaution can stimulate innovation and science (pp. 4, 5, 11)#

Original claim (Preface, p. 4). “Many of the case studies suggest that wider use of the precautionary principle can help stimulate both innovation and science, replacing the 19th century technologies … with the ‘eco-efficient’ technologies and systems science of the third [industrial revolution].” Related claims: the aim of preventing harm “without stifling innovation” (p. 11), and “maximise innovation” (p. 5).

Subsequent developments

The closest empirical literature is the Porter hypothesis: mixed. - Ambec, Cohen, Elgie and Lanoie (2013, Review of Environmental Economics and Policy 7(1):2–22) report “mixed empirical findings” after two decades of work. - Cohen and Tubb’s meta-analysis (2018, Journal of the Association of Environmental and Resource Economists 5(2):371–399) covers 103 publications and over 2,000 effect sizes. It finds “considerable heterogeneity”, and a positive effect “more likely at the state, region, or country level … although in both cases the most likely scenario is statistical insignificance”. The authors call this “consistent with the strong version of the Porter hypothesis”. - Environmental regulation is not the same thing as precaution, so this is the nearest test available, not a direct one.

The EEA doubled down in 2013, largely on case evidence. The 2013 conclusion states there is “growing evidence that precautionary measures do not stifle innovation, but instead can encourage it, in particular when supported by smart regulation or well-designed tax changes” (EEA 2013 summary, ch. 28). It does not cite a systematic test in the summary.

The claim was actively contested in EU policy. In 2015 the European Risk Forum, BusinessEurope and the European Round Table of Industrialists proposed an “innovation principle” to complement precaution. They argued that whenever EU institutions consider proposals, “impact on innovation should be fully assessed and addressed” (as reported by EPRS 2015, section 4.3). - I could not verify from EUR-Lex how far that principle was later written into EU instruments. - The Commission’s 2021 Better Regulation Communication (COM(2021) 219) mentions neither principle.

A cautionary example from within the report’s own cases: regrettable substitution. - Controls on ozone-depleting substances did drive substitution. But hydrofluorocarbons (HFCs), “originally developed as alternatives to ozone-depleting CFCs and HCFCs”, turned out to be potent greenhouse gases (HFC-134a has a 100-year global warming potential of 1,430). This required the Kigali Amendment (adopted October 2016, in force 1 January 2019) to phase them down. - The report itself includes a similar case: MTBE as a substitute for lead in petrol (Ch 11, contents p. 8). - Innovation induced by precaution is not automatically benign. The report’s own Lesson 7 on evaluating alternatives (reproduced in EEA 2013 summary, p. 11) is the needed qualifier.

The EEA’s own case author conceded the limits of the innovation stimulus. The 2013 annex summarising the 2001 halocarbons case says CFCs “were largely replaced by HCFCs (ozone depleting substances, but less harmful that [sic] CFCs) and HFCs (with zero ozone depletion but powerful greenhouse gases)”. It adds: “The political process failed to stimulate more radical changes towards halocarbon‑free and energy‑efficient technology” (EEA 2013, Annex 2, p. 738, summarising Farman’s 2001 chapter). So even the ozone case, widely cited as precaution’s success story, produced incremental chemical substitution on the EEA’s own account, not the shift to “eco-efficient” technologies the Preface describes (p. 4).

Verdict: contested. It is plausible and case-supported. Systematic evidence neither confirms nor refutes it at firm level, and it is somewhat favourable at country level. It remains a live policy fight.

Implications for weight. Present it as a conditional hypothesis: well-designed, flexible precautionary or regulatory pressure can induce innovation, especially at system level, but the average effect is often not statistically distinguishable from zero. The more robust lesson is technology-neutral: precaution works best when paired with active appraisal of alternatives, because substitution can create new hazards (pp. 4, 13; Lesson 7).

Claim 6: No robust false positives found; false positives are “the smaller but commonly feared risk” (pp. 12–13, 16)#

Original claim. - p. 12: despite invitations to industry, “no suitable examples emerged”. Facts versus fears (Lieberman and Kwon 1998) offered about 25 candidates, but these “turned out not to be robust enough for those who recommended them” to use. - p. 13: “possible candidates … include the ban on dumping sewage sludge in the North Sea, and the ‘Y2K millennium bug’”. - p. 16: the lessons “should also help reduce the smaller but commonly feared risk of ‘false positives’”.

Subsequent developments

The systematic follow-up. Hansen, Krayer von Krauss and Tickner (2007, Risk Analysis 27(1):255–269) defined regulatory false positives and reviewed 88 alleged cases. They found “only four cases that fit the definition”: Southern corn leaf blight, the 1976 US swine flu programme, saccharin, and food irradiation (for consumer health). - The 2013 volume repeats this: “88 cases were identified to be alleged false positives, however … most of them turned out to be either real risks, or cases where ‘the jury is still out’, or unregulated alarms, or risk-risk trade-offs” (EEA 2013 summary, ch. 2). - It concludes that “fear of false positives is misplaced”. It also acknowledges that the swine flu action “did lead to some unintended deaths and human suffering, and diverted resources”, so false positives are not only economic. - It suggests that the large number of “mistaken false positives” may reflect a deliberate strategy of “claiming over-regulation”. - EPRS (2015) summarised this finding for the European Parliament.

How high the bar was set. The 2013 chapter by Hansen and Tickner (EEA 2013, ch. 2, pp. 17–45) states the evidential standard used. - For a case to count as a false positive, “scientific evidence must exist showing that a perceived risk is actually non-existent”, and this must be “generally accepted”. The authors “argue that there should at least be a ‘high confidence’ (67–95 %) in the scientific evidence indicating no harm”. - This is a demanding evidential standard applied to absence of harm. The Preface warns (p. 4) against the mirror-image error, misplaced certainty about absence of harm, but no comparably strict test is applied before a risk counts as “real”. - The authors acknowledge “concerns about subjectivity”. They publish the full list of 88 cases (Table 2.3) “so that other researchers can repeat the analysis”, and cite Cox (2007) as the counter-view.

The critique. Cox (2007) argues that Hansen et al. apply “a diverse set of questionable criteria to label many highly uncertain risks as ‘real’”. His examples include “lack of a known exposure threshold”, “treating deliberately conservative (upper-bound) regulatory assumptions as if they were true values” and “treating claimed … epidemiological associations as if they were known to be true causal relations”. Such criteria “can classify even nonexistent and unknown risks as ‘real,’ providing an alternative possible explanation for why the authors failed to find more false positives, even if they exist”.

The structural problem remains. Once exposure is controlled, the evidence needed to show a risk was never real is rarely generated. A search of alleged cases can show false positives are rarely demonstrable. It cannot show they are rare, or cheaper than false negatives. No study since has estimated the relative costs of the two error types across a representative sample.

The two named candidates. Neither appears among the 88 cases in Hansen and Tickner’s Table 2.3 (EEA 2013, ch. 2, pp. 35–36). So neither of the two false-positive candidates the 2001 report named was ever assessed in the follow-up it prompted. - North Sea sewage sludge. Directive 91/271/EEC, Article 14, required that “by 31 December 1998 the disposal of sludge to surface waters by dumping from ships, by discharge from pipelines or by other means is phased out”. The measure was implemented. - My inference: its hindsight status looks like a risk-risk trade-off (sludge diverted to land and incineration) rather than a clean false positive. - Y2K. It remains contested. Quigley (2004, Public Administration 82(4):801–829) describes the UK response as “an uncreative, resource-heavy, centralized operation” vulnerable to “IT opportunists” and “an alarmist media”. He also quotes the Cabinet Office: “Things did not go right by accident”. Y2K was not an environmental or health measure, so it is a weak test of the principle either way.

Verdict: contested. The report’s finding that no robust false positive was found was followed by a systematic search that found few (4 of 88). That supports the claim’s direction. But the search’s classification criteria have been credibly challenged, and it set a high bar (“high confidence” of no harm) for counting a case as a false positive. Neither of the report’s own named candidates was assessed. “The smaller … risk” remains an asserted comparison, not a measured one.

Implications for weight. The methodological point is strong: a record of confirmed harms cannot estimate false-positive rates (pp. 12–13). The substantive claim that false positives are rare and cheaper should be cited with Cox’s critique alongside Hansen et al. The genuine false positives that were found (for example, 1976 swine flu) show they can carry human as well as economic costs. That supports the report’s broader call to weigh the costs of error in both directions (pp. 13–14, 16).

Claim 7: The US practises “precautionary prevention” without the label (Table 1.1, p. 12)#

Original claim (p. 12, Table 1.1). - Delaney Clause “1957–96”. - A scrapie-meat ban in the early 1970s “which may have helped the United States to avoid BSE”. - A 1977 aerosol CFC ban “several years before similar action in most of Europe”. - A DES growth-promoter ban 1972–79, “nearly 10 years before the EU ban in 1987”.

Subsequent developments

The general claim is supported by later systematic comparison. - Hammitt, Wiener, Swedlow, Kall and Zhou (2005, Risk Analysis 25(5):1215–1228) coded relative precaution for 100 randomly sampled risks, drawn from nearly 3,000, for each year from 1970 to 2004. They found “no significant difference in relative precaution over the period”, “a modest shift toward greater relative precaution of European regulation since about 1990”, and “a mixed and diverse pattern”. - EPRS (2015) summarises Wiener and Rogers (2002, Journal of Risk Research 5(4):317–349): the EU was more precautionary on growth hormones, while “the US has taken more precautions than the EU with respect to BSE”. - The fuller treatment is Wiener, Rogers, Hammitt and Sand (eds), The Reality of Precaution (RFF Press 2011).

The Delaney principle is still operating, and the table’s dates are wrong. - The Food Quality Protection Act (Public Law 104-170, enacted 3 August 1996) removed “a pesticide chemical residue in or on a raw agricultural commodity or processed food” from the definition of food additive, and so from Delaney. It did not repeal Delaney for additives or colour additives. - On 15 January 2025, FDA revoked authorisation of FD&C Red No. 3 under the colour-additive Delaney Clause (enacted 1960) because of cancer in male rats. It did so even though, on FDA’s own account, the mechanism seen in male rats does not occur in humans (paraphrased from the FDA notice). This is a live example of precautionary prevention under another name. - The table’s “1957–96” is wrong at both ends: the food-additive clause was enacted in 1958, and it did not end in 1996.

The EU DES date is wrong. Council Directive 81/602/EEC of 31 July 1981 prohibited administering to farm animals “substances having a thyrostatic action or substances having an oestrogenic, androgenic or gestagenic action”. It also prohibited “the placing on the market of stilbenes, stilbene derivatives”, the class that includes DES. - The US lead over the EU was therefore about two years from the final 1979 US action, or nine years from the first 1972 action. It was not “nearly 10 years before … 1987”. - A likely source of the 1987 date is the general growth-hormone ban (Directive 85/649/EEC, annulled by the Court and re-adopted as 88/146/EEC). I could not re-open that text in this check.

BSE: the outcome holds; the causal attribution is unsupported. - APHIS reports seven US BSE detections from 2003 to 2023: one classical case “imported from Canada” (2003) and six atypical cases. The US has WOAH negligible-risk status. - FDA’s official account of US BSE controls begins with the 1997 ruminant feed ban (21 CFR 589.2000), strengthened in 2008 (21 CFR 589.2001). It mentions no 1970s scrapie measure. - The 2003 case, a cow born in Canada, shows that the integrated North American cattle system was not immune. - A second-pass web search also failed to find a US federal source for a 1970s scrapie-meat ban. That does not show no such measure existed, only that I could not document one. The 2013 volume’s annex repeats the claim (“Mid-1970s The United States of America bans scrapie-infected sheep and goat meats from cattle food chain”, EEA 2013, Annex 2, Table A2.15, p. 746) but gives no regulatory citation. - The claim’s mechanism was already disputed when the report came out. The UK BSE Inquiry concluded that BSE most likely began as “a new prion mutation in cattle, or possibly sheep, in the early 1970s” (Minister’s statement to the House of Commons, 26 October 2000). If that is right, a measure aimed at scrapie-infected sheep would probably have done little to prevent BSE. The controls officially credited with stopping BSE are ruminant-feed bans: the UK ban in the EEA timeline (Annex 2, Table A2.15) and the 1997 US ban noted above. - Later experiments reopened a scrapie route, but a different one. Huor et al. (2019, PNAS 116:26853–26862) found that classical BSE prions emerged when atypical/Nor98 scrapie was passaged in mice carrying the bovine prion protein. Atypical scrapie is not the classical scrapie a 1970s measure would have targeted. - So the table’s “may have helped” is at best an untested hypothesis, and its mechanism is uncertain. The measures officially credited came later.

CFCs. Not flagged, and consistent with the record. I did not re-verify the European Community’s 1980 aerosol decision (80/372/EEC) in this check.

After 2001 the US added further precautionary-preventive measures, notably the 2024 chrysotile ban (see Claim 4).

Verdict: partly held up. The core point that the US practises precaution without the label is supported and strengthened by later systematic work. Three of the four table rows contain errors: Delaney’s dates, the EU DES date, and the unsupported BSE attribution.

Implications for weight. The transferable lesson is sound: the label “precautionary principle” is a poor guide to actual precautionary practice, and jurisdictions differ by issue, not wholesale (p. 12, Table 1.2). Table 1.1 should not be quoted without correction.

Claim 8: The 2000 Communication and Nice decision helped practical implementation, “especially concerning stakeholder involvement and the avoidance of trade disputes” (p. 13)#

Original claim (p. 13). The European Commission’s Communication on the Precautionary Principle (COM(2000) 1) and the Council’s Nice decision “have made significant contributions to the practical implementation of the precautionary principle, especially concerning stakeholder involvement and the avoidance of trade disputes”.

Subsequent developments

Implementation in EU law and courts: substantial. EPRS (2015) documents that precaution became embedded in, among others: - the General Food Law (Regulation (EC) No 178/2002) - REACH (Regulation (EC) No 1907/2006) - the GMO Directive 2001/18/EC

The EU courts applied the principle in Pfizer Animal Health (T-13/99, judgment of 11 September 2002, para. 444, upholding the ban on antibiotic growth promoters) and in Waddenzee (C-127/02, 2004). I could not re-open the judgments themselves on EUR-Lex or CURIA; this relies on EPRS’s account.

Avoiding trade disputes: not achieved. - EC–Biotech (DS291). The US requested consultations on 13 May 2003. The panel report (circulated 29 September 2006, adopted 21 November 2006) found “a general de facto moratorium on the approval of biotech products between June 1999 and August 2003”. It found undue delay for 24 of 27 products, and that member-state safeguard bans “were not based on risk assessments”. “The panel did not rule on the precautionary principle” (WTO DS291 summary). A mutually agreed framework followed in January 2008. - Hormones (DS26 and DS320). - The EU adopted Directive 2003/74/EC (22 September 2003). It relied on its scientific committee’s finding that oestradiol-17β “has to be considered as a complete carcinogen” and banned it permanently, while provisionally banning five other hormones “while the Community seeks more complete scientific information”. - The EC then challenged continued US retaliation (DS320). In October 2008 the Appellate Body reversed the panel on the provisional ban’s consistency with SPS Article 5.7, but “could not complete its analysis”. - The dispute was managed rather than resolved: a 2009 MOU (revised 2014), then an agreement signed 2 August 2019 to allocate “35,000 metric tons of the 45,000 metric tons” high-quality beef quota to the US from 1 January 2020. The USTR notes that “the United States continues to have an authorization to suspend concessions” (FR Doc 2019-26924, 13 December 2019). - Trade agreements. The EU has since tried to write precaution into bilateral agreements, but where it sits and whether it binds are disputed. - For the EU–Mercosur agreement, the Commission says the text “reaffirms the precautionary principle within the agreement, permitting the adoption of provisional SPS measures, including in cases where scientific information is insufficient” (DG Trade factsheet, undated, current 2025). SPS here means sanitary and phytosanitary (food-safety and animal- and plant-health) measures. - The Institute for Agriculture and Trade Policy (IATP; 7 December 2020, an NGO analysis of the 2019 agreement-in-principle text) says the only explicit reference is in the trade and sustainable development chapter. That chapter is “not enforceable” through the agreement’s dispute settlement, and the SPS chapter defers to WTO and Codex standards. - I did not verify the final treaty text. Either way, the principle has been negotiated clause by clause with each trading partner, not accepted as a shared baseline. - 2025. The EU–US Joint Statement of 21 August 2025 still refers to “longstanding concerns” about “non-tariff barriers affecting trade in food and agricultural products”. It is silent on hormones, GMOs and precaution.

One counter-example. In EC–Asbestos (DS135), the Appellate Body upheld France’s asbestos ban in March 2001. It confirmed that a country may act on “a divergent opinion coming from qualified and respected sources” (EPRS 2015). The WTO can accommodate precautionary health measures, but that ruling predates the Communication’s effects.

Stakeholder involvement: see Claim 10. EFSA was created in 2002. The 2019 Transparency Regulation expanded public consultation and disclosure of studies, but it was a response to continuing distrust, not evidence that the 2000 settlement had worked.

Verdict: partly held up on implementation; overturned on the avoidance of trade disputes. The Communication became the reference framework inside the EU. The major transatlantic disputes that precaution was invoked in continued for years after 2000. They were settled through market-access deals, not through shared understanding of precaution.

Implications for weight. The report’s hope for “improved transatlantic agreement” (p. 3) was not met. The transferable lesson is technology-neutral: a jurisdiction’s procedural codification of precaution does not by itself persuade trading partners who use different evidential triggers. Disputes tend to be resolved by compensation, not by convergence on the principle.

Claim 9: Snow-style precautionary prevention after the early asbestos warnings would have averted “much of the tragedy and the huge costs” (p. 15)#

Original claim (p. 15). “Had governments adopted a similar approach to precautionary prevention as Dr Snow, once the early warnings on asbestos had been published, much of the tragedy and the huge costs of asbestos exposure could have been averted.”

Subsequent developments

The exposure–outcome structure supports the counterfactual’s direction. - HSE attributes current GB deaths mainly to “widespread industrial use of asbestos during 1950–1980” (HSE 2026a). The early warnings predate that: the 1898 inspector’s report quoted on p. 11, and later warnings covered in Ch 5. - The birth cohort with the highest male mesothelioma rates is 1935–39, people “young during the time of the peak asbestos usage”. - GB recorded 23,861 mesothelioma deaths from 1968 to 2000 and 54,900 from 2001 to 2024 (HSE MESO01). Most of this toll traces to exposures after the early warnings.

Bans work, with a long lag. - Jin et al. (2025, Lung Cancer 203:108534) report that “most countries with asbestos bans continued to exhibit declining trends”, and that “global asbestos bans have demonstrated sustained effectiveness in controlling mesothelioma”. - Li et al. (2025, BMJ Public Health 3:e001717) find “a positive impact of asbestos ban policies … although the effects take decades to manifest”. - This supports the counterfactual that earlier restriction would have averted much of the later toll, and it shows why the benefits would have been invisible for decades.

What later evidence does not address. The feasibility and cost of such action in, say, the 1930s or 1960s remain untested: substitutes, economic dependence, the scale of use in construction. The Snow analogy understates this gap: Snow’s intervention was cheap, local, reversible and immediately testable (pp. 14–15, Box 1.1). The report itself notes that the latency was “hours in the case of cholera but decades in the case of asbestos” (p. 15). Chimed-Ochir et al. (2022) show that, in practice, countries have mostly waited for their own mesothelioma burden before banning.

Verdict: partly held up. The counterfactual’s direction and rough scale are well supported by later exposure-cohort and ban-effect evidence. Its implied ease (“similar approach … as Dr Snow”) is not; no later work shows that such action was feasible at acceptable cost at the dates in question.

Implications for weight. The logic of acting before mechanism or definitive proof is well founded (pp. 14–15). So is the point that the payoff is deferred and hard to see. The Snow analogy should be used to illustrate that logic, not as evidence that early action would have been easy. The key technology-neutral lesson is that the benefits of early restriction arrive decades later, which structurally weakens the political case for acting (pp. 3–4).

Claim 10: Public trust in politicians and scientists is very low, especially in Europe; the report aims to contribute to “democratising scientific expertise” (p. 16)#

Original claim (p. 16). “Public trust in the politicians and scientists who are trying to protect people and the planet from hazards is very low, especially in Europe”, citing BSE, Belgian dioxins and the French contaminated-blood affair. Governments are responding, for example through the EU White Paper on European Governance (July 2001). The report “aims to contribute to the debate on the emerging issue of democratising scientific expertise”.

Subsequent developments

Trust in politicians remains very low. In the UK, the Ipsos Veracity Index 2024 (published 20 November 2024) found “just eleven per cent” trust politicians to tell the truth, after a “40-year low of nine per cent” the previous year. Government ministers stood at 15%.

Trust in scientists is not “very low”, and Europe ranks comparatively high. - In the Wellcome Global Monitor 2018, “eleven out of the 19 countries where more than 30% of people express ‘high’ trust in scientists are in Europe”. Examples: Belgium 42% high trust, Spain 39%, Ireland 39%. The world average was 18%. - Cologna et al. (2025, Nature Human Behaviour 9:713–730; 71,922 respondents in 68 countries) found that “in most countries, most people trust scientists”, and that “there is no widespread lack of trust in scientists”. Contemporaneous EU data did not support “very low” trust in scientists either. Eurobarometer 55.2 was fielded in May–June 2001 in the EU15, while the report was being finalised, and published in December 2001. - Asked which professions they held in most esteem, 71.1% chose doctors, 44.9% scientists and 29.8% engineers. Politicians came last, at 6.6%. - Asked whom they would most trust to explain a local disaster, 62.6% chose scientists (cumulative across three choices), 59.8% environmental groups and 55.3% doctors. Government representatives drew 19.9% and businesses 4.4%. - The same survey shows where the damage was. For the “mad cow affair”, 74.3% blamed the agri-food industry, 68.6% politicians, 59.1% farmers and 50.6% scientists. And 80.3% agreed that “the authorities should formally oblige scientists to observe ethical rules” (European Commission 2001, pp. 7, 43–45).

Special Eurobarometer 224 (fieldwork 3 January–15 February 2005; EU25) found the same split by institution. - 52% named “scientists working at a university or government laboratory” among the best qualified to explain the impacts of science and technology, against 28% for scientists in industrial laboratories, 6% for industry, 6% for the government and 5% for politicians. - 73% agreed that “politicians should rely more on the advice of expert scientists” (European Commission 2005, pp. 49–50, 90).

So the report’s pairing of “politicians and scientists” as equally distrusted was inaccurate even at the time for scientists in general. It is fairer to read “scientists who are trying to protect people” as regulatory and advisory scientists. For them the data show real reputational damage (half of respondents blamed scientists for BSE) and a strong demand for accountability. But publicly employed scientists remained far more trusted than government or industry.

Where distrust persisted: regulatory science. Governance reforms did follow, and they were explicitly about trust. - EFSA was “set up in 2002” (EFSA). - Regulation (EU) 2019/1381 (20 June 2019, applicable 27 March 2021) cites the General Food Law fitness-check finding that risk communication was not effective enough overall, which affected public confidence (recital 3). It also cites the citizens’ initiative “Ban glyphosate and protect people and the environment from toxic pesticides”, which “confirmed concerns regarding transparency with respect to studies commissioned by industry” (recital 27). - It introduced notification of studies, proactive disclosure and consultation of third parties. - The 2013 volume reports that “the number of stakeholders involved in decision-making has become larger and more diverse”. It also says political and scientific “bureaucratic silos” remain (EEA 2013 summary, ch. 28).

Verdict: partly held up. Low trust in politicians held and persists. Low trust in scientists “especially in Europe” was not supported by the EU’s own survey data in 2001 or 2005, nor by later cross-national data, unless “scientists” is read narrowly as regulatory and advisory scientists. The “democratising expertise” agenda was taken up in EU risk governance, notably food and chemicals, but distrust of regulatory science resurfaced nearly two decades later and required further legislation.

Implications for weight. The link between trust erosion and participatory reform (p. 16) is borne out as a policy dynamic. The analytical lesson should be stated precisely: distrust concentrates on the institutions and evidence pipelines of regulatory science (who funds and controls the studies), not on scientists as such. Transparency about the provenance of evidence is the lever reforms actually pulled.

Reception and uptake of the section’s framework#

The section’s framing became a standard reference in EU policy discussion. - EPRS (2015) adapted its case list and cited both volumes in its briefing to MEPs. - The 2013 volume restated the twelve lessons unchanged and tried to fill the report’s acknowledged false-positive gap (EEA 2013 summary, pp. 9–11).

The main lines of critique map onto the digest’s caveats: - selection on the outcome (answered only partly by Hansen et al. 2007, and challenged by Cox 2007) - whether Europe is really more precautionary than the US (Hammitt et al. 2005: mixed) - the innovation claim (Porter-hypothesis literature: mixed; the industry “innovation principle” campaign)

I found no later work that overturns the section’s structural claims about latency, action before mechanism, the ambiguity of the term, or two-way weighing of error costs. Several later findings strengthen them: HSE cohort analyses, ban-lag studies, and ozone-recovery timelines.

Items to carry into synthesis#

Still open: 1. The chapter-by-chapter test of “misplaced certainty … in most of the case studies” (Claim 3). 2. The Fifth Circuit’s decision in Texas Chemistry Council v. EPA (No. 24-60193), argued 1 June 2026 and pending as far as I could find. 3. The source of Table 1.1’s “EU ban in 1987” for DES: probably the general hormone ban (Directive 85/649/EEC, re-adopted as 88/146/EEC), not re-opened here. Also whether a US federal scrapie-meat measure of the early or mid-1970s can be documented. 4. Whether the local copy of the 2013 report numbers its annex on the 2001 cases differently (this brief calls it Annex 3; the DTU-hosted copy calls it Annex 2).

Closed in the second pass: - The 2013 annex on the 2001 cases: it restates asbestos as “some 400 000” European mesothelioma deaths (Claim 1) and concedes that the CFC phase-out produced only incremental substitution (Claim 5). - The US chrysotile rule and the Rotterdam listing history (Claim 4). - North Sea sludge and Y2K are not among Hansen et al.’s 88 cases (Claim 6).

Sources#

All accessed 25 September 2026 unless noted. “Not re-fetched” marks documents I could not open during this check.

Official statistics and agency sources - HSE (2026a). Mesothelioma statistics for Great Britain, 2026. July 2026. https://www.hse.gov.uk/statistics/assets/docs/mesothelioma.pdf - HSE (2026b). Asbestos-related disease statistics, Great Britain 2026. July 2026. https://www.hse.gov.uk/statistics/assets/docs/asbestos-related-disease.pdf - HSE. Table MESO01: Death certificates mentioning mesothelioma, 1968–2024. Published 1 July 2026. https://www.hse.gov.uk/statistics/assets/docs/meso01.xlsx - WHO. Asbestos: elimination of asbestos-related diseases (fact sheet). 11 August 2026. https://www.who.int/news-room/fact-sheets/detail/asbestos-elimination-of-asbestos-related-diseases - US EPA. Asbestos Part 1; Chrysotile Asbestos; Regulation of Certain Conditions of Use Under TSCA. Final rule, Federal Register, 28 March 2024 (FR Doc 2024-05972), effective 28 May 2024. https://www.federalregister.gov/documents/2024/03/28/2024-05972/asbestos-part-1-chrysotile-asbestos-regulation-of-certain-conditions-of-use-under-the-toxic - US EPA. EPA actions to protect the public from exposure to asbestos. https://www.epa.gov/asbestos/epa-actions-protect-public-exposure-asbestos - US EPA. Health and environmental effects of ozone layer depletion. https://www.epa.gov/ozone-layer-protection/health-and-environmental-effects-ozone-layer-depletion - WMO/UNEP. Scientific Assessment of Ozone Depletion: 2022, Executive Summary (NOAA CSL host). 2022. https://csl.noaa.gov/assessments/ozone/2022/executivesummary/ - UNEP Ozone Secretariat. Kigali Amendment (2016). https://ozone.unep.org/treaties/montreal-protocol/amendments/kigali-amendment-2016-amendment-montreal-protocol-agreed - USDA APHIS. Bovine spongiform encephalopathy. https://www.aphis.usda.gov/livestock-poultry-disease/cattle/bse - US FDA. Bovine spongiform encephalopathy. https://www.fda.gov/animal-veterinary/compliance-enforcement/bovine-spongiform-encephalopathy - US FDA. FDA to revoke authorization for the use of Red No. 3 in food and ingested drugs. 15 January 2025. https://www.fda.gov/food/hfp-constituent-updates/fda-revoke-authorization-use-red-no-3-food-and-ingested-drugs - US EPA. Risk Management for Asbestos, Part 1: Chrysotile Asbestos. Last updated 5 May 2026. https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-asbestos-part-1-chrysotile-asbestos - US Court of Appeals for the Fifth Circuit. Texas Chemistry Council v. EPA, No. 24-60193, oral argument recording, 1 June 2026. https://www.ca5.uscourts.gov/OralArgRecordings/24/24-60193_6-1-2026.mp3 (not listened to; cited for the hearing date) - Rotterdam Convention Secretariat. Chrysotile asbestos (chemicals recommended for listing). https://www.pic.int/TheConvention/Chemicals/RecommendedtoCOP/Chrysotileasbestos/tabid/1186/language/en-US/Default.aspx - UK House of Commons. BSE Inquiry Report (statement by Nick Brown, Minister of Agriculture). Hansard, 26 October 2000. https://api.parliament.uk/historic-hansard/commons/2000/oct/26/bse-inquiry-report - European Commission, DG Research (2001). Eurobarometer 55.2: Europeans, science and technology. December 2001 (fieldwork May–June 2001). Archived copy: https://web.archive.org/web/2010/http://ec.europa.eu/public_opinion/archives/ebs/ebs_154_en.pdf (the original europa.eu URL returns 404) - European Commission (2005). Special Eurobarometer 224: Europeans, Science & Technology. June 2005 (fieldwork 3 January–15 February 2005). https://lms.lt/archyvas/files/active/0/ebs_224_report_en.pdf (third-party mirror of the Commission report) - European Commission, DG Trade. Factsheet: EU–Mercosur partnership agreement — respecting Europe’s health and safety standards (undated; accessed 25 September 2026). https://policy.trade.ec.europa.eu/eu-trade-relationships-country-and-region/countries-and-regions/mercosur/eu-mercosur-agreement/factsheet-eu-mercosur-partnership-agreement-respecting-europes-health-and-safety-standards_en - EFSA. About EFSA. https://www.efsa.europa.eu/en/about/about-efsa - EFSA. Transparency Regulation implementation. https://www.efsa.europa.eu/en/stakeholders/transparency-regulation-implementation - EU-OSHA. Directive 2009/148/EC on exposure to asbestos at work (as amended by Directive (EU) 2023/2668). https://osha.europa.eu/en/legislation/directive/directive-2009148ec-exposure-asbestos-work

Legislation, treaties and trade - Council Directive 81/602/EEC of 31 July 1981 (prohibition of certain substances having a hormonal action and of any substances having a thyrostatic action). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31981L0602 - Directive 2003/74/EC of 22 September 2003 amending Council Directive 96/22/EC. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32003L0074 - Council Directive 91/271/EEC of 21 May 1991 (urban waste water treatment), Article 14. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31991L0271 - Regulation (EU) 2019/1381 of 20 June 2019 (transparency and sustainability of EU risk assessment in the food chain). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32019R1381 - Commission Communication COM(2021) 219, Better regulation: Joining forces to make better laws, 29 April 2021. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021DC0219 - Council Directive 88/146/EEC (not re-fetched). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31988L0146 - Council Decision 80/372/EEC on CFCs (not re-fetched). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31980D0372 - Food Quality Protection Act of 1996, Public Law 104-170, 3 August 1996. https://www.govinfo.gov/content/pkg/PLAW-104publ170/html/PLAW-104publ170.htm - WTO. DS291: European Communities — Measures Affecting the Approval and Marketing of Biotech Products. https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds291_e.htm - WTO. DS26: European Communities — Measures Concerning Meat and Meat Products (Hormones). https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds26_e.htm - WTO. DS320: United States — Continued Suspension of Obligations in the EC–Hormones Dispute. https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds320_e.htm - USTR. Determination Not To Reinstate Action in Connection With the European Union’s Measures Concerning Meat and Meat Products. Federal Register 84:68286–68287, 13 December 2019 (FR Doc 2019-26924). https://www.federalregister.gov/documents/full_text/text/2019/12/13/2019-26924.txt - European Commission (DG Trade). Joint Statement on a United States–European Union framework on an Agreement on Reciprocal, Fair and Balanced Trade. 21 August 2025. https://policy.trade.ec.europa.eu/news/joint-statement-united-states-european-union-framework-agreement-reciprocal-fair-and-balanced-trade-2025-08-21_en

Peer-reviewed literature - Peto J, Decarli A, La Vecchia C, Levi F, Negri E (1999). The European mesothelioma epidemic. Br J Cancer 79(3–4):666–672. https://doi.org/10.1038/sj.bjc.6690105 (abstract via Europe PMC, PMC2362439) - Peto J, Hodgson JT, Matthews FE, Jones JR (1995). Continuing increase in mesothelioma mortality in Britain. Lancet 345(8949):535–539. PMID 7776771. https://pubmed.ncbi.nlm.nih.gov/7776771/ - Pelucchi C, Malvezzi M, La Vecchia C, Levi F, Decarli A, Negri E (2004). The mesothelioma epidemic in Western Europe: an update. Br J Cancer 90(5):1022–1024. https://doi.org/10.1038/sj.bjc.6601638 (full text: https://europepmc.org/articles/PMC2409631) - Hodgson JT, McElvenny DM, Darnton AJ, Price MJ, Peto J (2005). The expected burden of mesothelioma mortality in Great Britain from 2002 to 2050. Br J Cancer 92(3):587–593. PMID 15668716. https://pubmed.ncbi.nlm.nih.gov/15668716/ - Chimed-Ochir O et al. (2022). Must countries shoulder the burden of mesothelioma to ban asbestos? A global assessment. BMJ Glob Health 7(12):e010553. Published 1 December 2022. https://doi.org/10.1136/bmjgh-2022-010553 - Odgerel CO et al. (2017). Estimation of the global burden of mesothelioma deaths from incomplete national mortality data. Occup Environ Med 74(12):851–858. https://doi.org/10.1136/oemed-2017-104298 - Jin W et al. (2025). The global burden of mesothelioma and its association with asbestos bans, 1990–2021. Lung Cancer 203:108534. Published 9 April 2025. https://doi.org/10.1016/j.lungcan.2025.108534 - Li WZ et al. (2025). Global mortality burden of lung cancer and mesothelioma attributable to occupational asbestos exposure and the impact of national asbestos ban policies, 1990–2021. BMJ Public Health 3(2):e001717. Published 8 July 2025. https://doi.org/10.1136/bmjph-2024-001717 - Fu L et al. (2026). Global patterns and trends in mesothelioma incidence: a retrospective cross-sectional study. Lung Cancer 211:108878. Published 11 December 2025. https://doi.org/10.1016/j.lungcan.2025.108878 - van Dijk A et al. (2013). Skin cancer risks avoided by the Montreal Protocol — worldwide modeling integrating coupled climate-chemistry models with a risk model for UV. Photochem Photobiol 89(1):234–246. https://doi.org/10.1111/j.1751-1097.2012.01223.x - Lee-Taylor J et al. (2026). Effects of ozone-depleting substances on ultraviolet radiation and skin cancer rates in Australia and the United States of America. GeoHealth 10(5):e2024GH001283. https://doi.org/10.1029/2024gh001283 - Hansen SF, Krayer von Krauss MP, Tickner JA (2007). Categorizing mistaken false positives in regulation of human and environmental health. Risk Anal 27(1):255–269. https://doi.org/10.1111/j.1539-6924.2006.00874.x - Cox LA Jr (2007). Regulatory false positives: true, false, or uncertain? Risk Anal 27(5):1083–1086. https://doi.org/10.1111/j.1539-6924.2007.00975.x - Hammitt JK, Wiener JB, Swedlow B, Kall D, Zhou Z (2005). Precautionary regulation in Europe and the United States: a quantitative comparison. Risk Anal 25(5):1215–1228. https://doi.org/10.1111/j.1539-6924.2005.00662.x - Wiener JB, Rogers MD (2002). Comparing precaution in the United States and Europe. J Risk Res 5(4):317–349. https://doi.org/10.1080/13669870210153684 (content as summarised in EPRS 2015) - Wiener JB, Rogers MD, Hammitt JK, Sand PH (eds) (2011). The Reality of Precaution: Comparing Risk Regulation in the United States and Europe. RFF Press (Routledge reissue 2013). https://doi.org/10.4324/9781936331802 (not read in full) - Ambec S, Cohen MA, Elgie S, Lanoie P (2013). The Porter Hypothesis at 20: can environmental regulation enhance innovation and competitiveness? Rev Environ Econ Policy 7(1):2–22. https://ideas.repec.org/a/oup/renvpo/v7y2013i1p2-22.html - Cohen MA, Tubb A (2018). The impact of environmental regulation on firm and country competitiveness: a meta-analysis of the Porter hypothesis. J Assoc Environ Resour Econ 5(2):371–399. https://doi.org/10.1086/695613 - Quigley K (2004). The Emperor’s New Computers: Y2K (re)visited. Public Administration 82(4):801–829. https://doi.org/10.1111/j.0033-3298.2004.00420.x - IARC (2012). Arsenic, Metals, Fibres, and Dusts. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 100C (asbestos, all forms, Group 1). https://publications.iarc.who.int/120 - Huor A et al. (2019). The emergence of classical BSE from atypical/Nor98 scrapie. Proc Natl Acad Sci USA 116(52):26853–26862. Published 16 December 2019. https://doi.org/10.1073/pnas.1915737116 - Cologna V et al. (2025). Trust in scientists and their role in society across 68 countries. Nat Hum Behav 9(4):713–730. Published 20 January 2025. https://doi.org/10.1038/s41562-024-02090-5

Reports and surveys - EEA (2013). Late lessons from early warnings: science, precaution, innovation — Summary. EEA Report No 1/2013, 22 January 2013. https://www.eea.europa.eu/en/analysis/publications/late-lessons-2 (summary PDF via the page’s download link; full-report links returned 404/410) - EEA (2013). Late lessons from early warnings: science, precaution, innovation, full report, EEA Report No 1/2013. Copy hosted by DTU Orbit: https://backend.orbit.dtu.dk/ws/portalfiles/portal/52992783/Gee%20et%20al.%202013%20Late%20lessons%20from%20early%20warnings%20II%20-%20Full%20report-2.pdf (used for Annex 2, “An overview of Late lessons from early warnings … vol. 1”, pp. 731–746) - Hansen SF, Tickner JA (2013). The precautionary principle and false alarms — lessons learned. In EEA Report No 1/2013, ch. 2, pp. 17–45. https://www.eea.europa.eu/en/analysis/publications/late-lessons-2/late-lessons-chapters/late-lessons-ii-chapter-2/@@download/file

Secondary and advocacy sources (used only where no primary source was retrieved) - Bergeson & Campbell PC. EPA withdraws motion to hold asbestos case in abeyance, will explore using guidance to clarify workplace protection requirements. 10 July 2025. https://www.lawbc.com/epa-withdraws-motion-to-hold-asbestos-case-in-abeyance-will-explore-using-guidance-to-clarify-workplace-protection-requirements/ - Asbestos Disease Awareness Organization (ADAO). ADAO returns to the Fifth Circuit to defend the science and strengthen asbestos protections. 2 June 2026. https://www.asbestosdiseaseawareness.org/newsroom/blogs/adao-court-asbestos-recap/ (advocacy group and intervenor in the case) - Ban Asbestos Network of India. COP-12 of Rotterdam Convention keeps inclusion of chrysotile asbestos in Annex III pending. May 2025. https://www.asbestosfreeindia.org/2025/05/cop-12-of-rotterdam-convention-keeps.html (advocacy source) - IndustriALL Global Union. Rotterdam Convention discredited as chrysotile asbestos fails to be listed. 3 May 2017. https://www.industriall-union.org/rotterdam-convention-discredited-as-chrysotile-asbestos-fails-to-be-listed (trade-union source) - Institute for Agriculture and Trade Policy (IATP). Food safety and the EU–Mercosur Agreement: risking weaker standards on both sides of the Atlantic. 7 December 2020. https://www.iatp.org/documents/food-safety-and-eu-mercosur-agreement-risking-weaker-standards-both-sides-atlantic (NGO analysis) - Bourguignon D (2015). The precautionary principle: definitions, applications and governance. EPRS In-depth analysis PE 573.876, December 2015. https://www.europarl.europa.eu/RegData/etudes/IDAN/2015/573876/EPRS_IDA(2015)573876_EN.pdf - Wellcome (2019). Wellcome Global Monitor 2018, ch. 3: Trust in science and health professionals. https://wellcome.org/reports/wellcome-global-monitor/2018/chapter-3-trust-science-and-health-professionals - Ipsos (2024). Ipsos Veracity Index 2024. 20 November 2024. https://www.ipsos.com/en-uk/ipsos-veracity-index-2024