Late Lessons, Jensen Huang and AI
Markdown version

LL1-15 hindsight check: ‘Mad cow disease’ 1980s–2000 (van Zwanenberg and Millstone), Late lessons from early warnings (EEA, 2001), Ch. 15, pp. 157–167#

Checked 25 September 2026. The check covers what happened between 2001 and September 2026. Page numbers refer to the 2001 report. “Phillips” means the BSE Inquiry report (October 2000), and “para” means a paragraph of its Volume 1 unless stated otherwise.

The Inquiry reported shortly before the chapter was finalised, and the chapter responds to its headline conclusions in Box 15.4 (p. 165). Strictly, then, Phillips is not hindsight. But the chapter engages only with two sentences of a 16-volume report. The Inquiry’s detailed findings bear directly on claims 1, 2, 3, 4 and 10, so I use them alongside the genuinely later evidence.

Overview#

The chapter’s factual narrative has held up well, both against the official record and against 25 years of epidemiology. Its interpretive charge has not been settled: that government knowingly and “covertly” put agricultural sales before health. The official inquiry rejected it, and later scholarship is split. Several of the chapter’s forward-looking statements were borne out: the continental European epidemic, large and long-lived costs, and decades of uncertainty about the human toll. Its institutional prescriptions were largely adopted. What is hard to show is that they made policy “more precautionary”. The clearest effect is that decisions became more open, more explicitly costed and more reversible.

Vindicated or strengthened - The feed ban leaked and prolonged the epidemic (claim 3). - By mid-2000 there were 41,538 confirmed cases in cattle born after the 1988 feed ban (Phillips, para 338). That is roughly a quarter of all UK cases. - UK cases among cattle born in 1995 numbered 1,248. Among those born in 1996, the year feed containing mammalian meat and bone meal was banned for all farm animals, there were 111 (APHA). - Great Britain has since recorded 191 cases in cattle born after the reinforced ban of August 1996. They include a classical case confirmed on 18 August 2026 in a cow born in 2019. - Enforcement was poor, and the reassurance message helped cause it (claim 10). - Phillips found enforcement worse than the chapter’s 48% figure suggests. More than half of plants visited in June 1995 were not meeting the staining rules. - Once enforced, the amount of offal reaching renderers more than doubled. That implies roughly a third had been evading the controls. - Phillips attributes the lack of diligence partly to the reassurance campaign (para 1186), which is close to the chapter’s causal claim. - BSE spread across continental Europe, as the chapter expected (claim 9). - In today’s 27 EU member states, classical BSE cases rose from 481 in 2000 to a peak of 1,006 in 2002 (EFSA). They fell to zero by 2017. - The cumulative total is 5,426. Ireland had 1,656 cases, Portugal 1,079 and France 969. - France recorded 28 cases of variant CJD (vCJD), the human disease linked to BSE. - Costs (claim 6). - The forecast the chapter cites is accurate. - Actual UK public spending came to about £4.8bn from 1996–97 to 2002–03, of which the European Community (EC) reimbursed £1.77bn (Hansard, HL 2004). The UK scheme that kept cattle over 30 months old out of the food chain ran until November 2005. - EU restrictions on UK beef lasted until May 2006. Access to the US, Chinese and Japanese markets reopened only in 2018–2020. - The human toll was genuinely uncertain (claim 5). - The UK had 178 vCJD deaths, the last in 2016 (NCJDRSU, December 2024). That is at the low end of the range the chapter quoted. - But the long tail remains open. Surveys of stored appendix tissue found abnormal prion protein in about 1 in 2,000 people in the birth cohorts sampled. There were also three transmissions of clinical vCJD by blood transfusion.

Contested, weakened or unclear - Motive (claim 1): contested. - Contemporaneous documents support the charge for 1986–88. They show fiscal and export motives and a restriction on information. - For 1986–96 as a whole, Phillips found that officials sincerely believed the risk was remote. It also found that the Department of Health (DH), which had no conflict of interest over food, was as keen as the agriculture ministry (MAFF) to avoid alarm (para 1189). - “Knew to be unavailable” (claim 2): partly held up. - Hansard confirms the gap between the advisers’ caveats and the minister’s statement. - Phillips judged that the reassurance campaign set out to sedate, not to deceive. - Exposure numbers (claim 4): partly held up. - Later modelling puts infected cattle at 1–3 million, most of them slaughtered for food before they showed signs of disease. On that scale the 1988–89 window the chapter highlights is a small part of total exposure. Exposure continued until 1996, including through mechanically recovered meat. - The comparative claim that Ireland was less precautionary than Austria because of how its institutions were organised (claim 8): unclear. - I found no systematic test of this. - The comparison is confounded by exposure: Austria recorded 5 classical cases, Ireland 1,656. - The institutional remedies (claim 7): partly held up. - They were adopted, but their record is mixed. - The authors’ own 2005 book calls the reform of food policy-making “partial”.

What the chapter could not see - Atypical BSE. Great Britain has recorded 10 H-type and 10 L-type cases since 2005 (APHA). Low-level cases therefore continue regardless of feed controls, and “eradication” is not a stable end state. - Mass testing revealed hidden disease. - Active surveillance began in 2001, testing of the order of a million cattle a year in the EU. - It found BSE in countries that had believed themselves free of it. - Back-calculation suggests that Great Britain itself under-recorded clinical cases, and that two to four times more cattle were infected than earlier models estimated (Donnelly et al., 2002). - Precautionary measures were later wound back through open, costed decisions. - In 2005 the Over Thirty Months rule was replaced by testing. The Food Standards Agency’s review found that keeping the rule would have cost roughly £2bn for each death prevented (Krebs, 2005). - The EU relaxed its feed ban in 2013 and 2021. - In 2025 the World Organisation for Animal Health (WOAH) classed Great Britain as negligible risk. - Much of the post-1996 spending bought back credibility rather than reducing risk. MAFF told the Public Accounts Committee that the selective cull and the cattle traceability system were largely irrelevant to eradicating BSE, but essential to lifting the export ban (PAC, 1999).

Weight for the lens. The chapter’s strongest lessons concern how controls fail, and later evidence supports all four: - reassurance that makes further precaution self-incriminating; - partial controls that leak through adjacent routes; - rules that are not enforced when their purpose is played down; - “no evidence of harm” produced by not looking.

Its structural lesson is plausible but only moderately supported: that combining promotion and protection in one body subordinates protection. Phillips’s finding about DH shows that the urge to reassure was not confined to the sponsoring department. The chapter’s language of intent (“covertly subordinated”, “knew to be unavailable”) should not be carried into the lens as settled fact.

2013 update. Annex 3 of Late lessons II (EEA Report 1/2013) does not update the BSE case. Annex 2 (p. 716) repeats the 2001 summary and Table 15.1 unchanged, as Table A2.15. That includes the 1989 date for the feed ban and the “almost 50 %” non-compliance line. Table A2.1 (p. 702, after Gee, 2009) dates the early warnings to 1979–1986 and action to “1989 partial; 1996 total ban”, with 10–17 years of inaction. Elsewhere the 2013 volume uses the case in four places, none of them adding new evidence: - the lesson “maintain regulatory independence” (Ch. 22, pp. 545–546); - MAFF’s conflicting mandates as an example of regulatory capture (Ch. 25, p. 612); - the 1998 European Court of Justice ruling in National Farmers’ Union (C‑157/96) as a statement of precaution in EU law (Ch. 27, p. 649); - lost trust in science as a cost that cannot be quantified (Ch. 27, p. 658).

Claim 1. UK BSE policy from 1986 to March 1996 was not precautionary; its “primary objective” was to limit the short-term impact on food-industry profits and public spending, and it “covertly subordinated” public health to agricultural sales (pp. 159, 164)#

Original claim. - “The government’s policy was not precautionary”. Its primary objective was to diminish “the short-term adverse impact of BSE on the profitability of the food industry and the level of public expenditure” (p. 159). This follows the account of 1986–88: export fears, the Treasury-minded minute of February 1988, and restricted information. - In conclusion (p. 164), the government “claimed to be prudently protecting public health, while in practice it covertly subordinated the protection of public health to the support of agricultural sales”, with a view also to minimising intervention and spending. - Box 15.4 (p. 165) quotes Phillips’s conclusions that the government “was anxious to act in the best interests of human health” and that MAFF did not “lean in favour of the agricultural producers”. The authors find these “difficult to see” as consistent with the evidence.

What happened since. - What Phillips actually concluded. The key conclusions (Vol. 1, pp. xvii–xviii) are more mixed than Box 15.4 suggests. They say: - the measures were sensible “but they were not always timely nor adequately implemented and enforced”; - the government “was anxious to act in the best interests of human and animal health”; - it was not MAFF’s policy to lean towards producers “to the detriment of the consumer”; - the government “did not lie”; - its “campaign of reassurance was a mistake”. - Phillips findings that support the chapter’s narrower claim (1986–88, and industry influence): - An internal embargo on information in the first half of 1987 “should not have occurred” (Executive Summary §3). - The minister’s reservations about compensation were anticipated because he had come from the Treasury. He worried about a “read across” to sugar-beet compensation (para 235). - Concern for the industry made officials and ministers “particularly concerned about how the public would perceive the risk” (para 1173). - In 1994–95, MAFF and the feed-trade association shared “reluctance” to push for measures that might lead compounders to stop using animal protein (para 350). - In 1994 the chief veterinary officer told the minister that banning meat and bone meal from all animal feed would be economically “devastating” (para 346). - Phillips findings that cut against the broader claim: - “MAFF never did less, and on occasion did more, than SEAC recommended” (para 1173). - Most of those involved believed, “in their heart of hearts”, that BSE was no threat to humans (para 1176). The Inquiry asked more than 270 witnesses whether BSE had changed their diet, and all but a handful said it had not (para 1177). - DH officials, who had no producer interest in food, were “as eager as MAFF” to present information so as to cause least alarm (para 1189). - On the February 1988 decision the chapter quotes (p. 159), Phillips judged it unfair to criticise the minister, because his reservations “did not lead him to reject the policy”. He referred the question to the Chief Medical Officer (paras 236–237). - Later historiography remains split. - The authors restated their thesis in “BSE: a paradigm of policy failure” (Political Quarterly, 2003) and in a 2005 monograph (BSE: risk, science and governance, OUP). I did not read either. - Forbes (2004, Political Studies) argues the opposite. In his reading the episode was neither a crisis nor a policy failure or disaster, and is better understood as an intractable policy problem. - A York working paper comparing the UK and Germany (Beck, Kewell and Asenova, 2007) ranks the UK episode among the worst UK policy disasters of recent decades. It finds similar patterns of risk-management failure in Germany, where agriculture and health responsibilities were divided. - No later disclosure I found overturns the documentary record on which both Phillips and the chapter rely. The dispute is about interpretation, not facts.

Verdict: contested. - The narrower claim is well supported, including by Phillips’s own findings: fiscal and export concerns and restricted information shaped policy in 1986–88, and industry comfort shaped the tempo of implementation later. - The broader claim of “covert subordination” of health across 1986–96 is contested. It imputes an intent that the official inquiry, having heard the witnesses, explicitly rejected. The historiography has not converged.

Weight. - Use the mechanism in a form the evidence supports: a department that is responsible for an industry’s prosperity will be especially anxious about how risk is perceived, and that anxiety shapes communication and the rigour of implementation (Phillips paras 1173, 1189). - Do not use the stronger form, that regulators knowingly traded lives for sales. - Phillips’s DH finding matters for the lens. The urge to reassure appeared without a sponsorship conflict too. Removing the conflict of interest is therefore probably necessary but not sufficient.

Claim 2. Ministers publicly claimed a certainty they knew was unavailable: in May 1990 SEAC advised that “no risk” could not be stated categorically, yet on 7 June 1990 the Agriculture Minister told the Commons there was “clear scientific evidence that British beef is perfectly safe” (p. 161)#

Original claim. - In May 1990 the government’s Spongiform Encephalopathy Advisory Committee (SEAC) advised that “it would not be justified to state categorically that there was no risk to humans”. - On 7 June 1990 the Agriculture Minister told the House of Commons there was “clear scientific evidence that British beef is perfectly safe”. - “Policy-makers repeatedly claimed a certainty which was unavailable and which they knew to be unavailable.” - They also claimed that the November 1989 controls kept all potentially contaminated material out of the food chain (p. 161).

What happened since. - The statement is confirmed. - It was made in the Commons on 7 June 1990, in exchanges on the minister’s statement about the EC Agriculture Council. - The minister said the common regime must be based on “the clear scientific evidence that British beef is perfectly safe” (Hansard, HC Deb 7 June 1990, around cols 906–907). - The context was trade. The Council had just overturned unilateral French, German and Italian restrictions on UK beef. - The chapter’s reference list gives the date as 8 June (p. 167). That is a minor slip; its text is correct. - Phillips on belief and intent. - From December 1986 to March 1996, ministers and officials followed an approach “whose object was sedation”. After the first half of 1987 there was no attempt to conceal facts, and the approach “did not set out to deceive”. One witness called it “leaning into the wind” (para 1179). - The Executive Summary concludes that the government “did not lie”, and that it “believed that the risks posed by BSE to humans were remote”. - Phillips on what the public was told. - Public statements “conveyed the message not merely that beef was safe but that BSE was not transmissible”. - The Chief Medical Officers’ statements that beef was safe rested on the assumption that the infective parts of the carcass were kept out of the food chain. That assumption “was not made clear to the public” (para 1181). - The public was not told that the perceived likelihood of transmission had changed (Executive Summary §5). - SEAC’s 1994 reassessment “vanished without trace” as far as the public was concerned (para 1185). - The reassuring premise was false in practice. - Mechanically recovered meat, taken from the spinal column, remained a possible route to human infection until December 1995 (Executive Summary §7). It could include pieces of spinal cord and dorsal root ganglia. Those nerve tissues were later shown to be infectious late in incubation. - In October 1995 spinal cord was found in four carcasses that had passed meat inspection (para 462). - So the implied claim that the 1989 controls kept all infective material out of food was untrue in fact, whatever ministers believed. - The harm materialised. There were 178 vCJD deaths in the UK (see claim 5). When a probable link was announced on 20 March 1996, Phillips found, “the public felt that they had been betrayed”.

Verdict: partly held up. - The core fact holds and is now better documented: public statements were categorical while the advice was not. The “total protection” premise was false in practice. - The phrase “knew to be unavailable” is literally supported, since ministers had been told certainty was unavailable. But in context it implies deliberate misrepresentation, and Phillips found sincere belief that the risk was remote. That part is contested.

Weight. The mechanism is strong: categorical public reassurance ran ahead of the advisers’ caveats and depended on controls that were leaking. Frame it as over-reassurance and failure to communicate how uncertain and conditional the safety claims were, not as proven lying.

Claim 3. The June 1988 feed ban covered only ruminants because an all-species ban would have cost renderers their main pig and poultry market; for “six or so years” cross-contamination occurred, “greatly prolonging the BSE epidemic” (p. 160)#

Original claim. - MAFF considered and rejected banning ruminant protein from all animal feed, because it “would have deprived the rendering industry of its principal market” (citing Inquiry testimony). - Cross-contamination between feed for pigs and poultry and feed for cattle followed for “the next six or so years, greatly prolonging the BSE epidemic” (p. 160). The chapter gives no source for the size of the effect.

What happened since. - The ban worked, but leaked. - Phillips calls the 1988 ban “the appropriate measure” and quotes an epidemiologist calling it “spectacularly successful”. It cut an escalating infection rate “to a fraction” (Executive Summary §6; paras 202–203). - APHA birth-cohort data show the same thing. UK cases among cattle born in 1987 numbered 37,478. For those born in 1988 the figure was 22,758, for 1989 12,880 and for 1990 5,821. These are minimum figures, because 43,350 cases have no recorded year of birth. - The leakage was large. - Cases in cattle born after the 18 July 1988 ban reached 41,538 by June 2000 (para 338). That is roughly 23% of all UK cases. - Nearly 12,000 were born in the rest of 1988 and over 12,000 in 1989. Later cohorts: over 5,600 born in 1990, 4,500 in 1991, 3,000 in 1992, 2,200 in 1993 and 1,000 in 1994 (paras 207–208). - Phillips concludes that most of the post-1990 cases “resulted from cross-contamination of cattle feed with pig and poultry feed, containing infective MBM” (meat and bone meal) (para 208). - A September 1990 ban on specified offal in pig and poultry feed should have closed the route. It was “unenforceable and widely disregarded” (Executive Summary §6). - Experiments later showed that as little as 1 g of infected brain could infect a cow (para 209). - An all-species ban closed the route. - Mammalian meat and bone meal was banned from all farm-animal feed on 29 March 1996 (para 337), with full enforcement taken to date from 1 August 1996. - UK cases among cattle born in 1995 numbered 1,248. For 1996 the figure was 111, and for 1997 it was 58 (APHA). - The EU banned processed animal protein in the feed of all farmed food animals from 1 January 2001 (Council Decision 2000/766/EC; not re-read). Ireland, for example, went on feeding meat and bone meal to non-ruminants until then (Harney et al., 2003). - Residual cases continue. - Great Britain has recorded 191 cases in cattle born after the reinforced ban: 180 classical, 6 H-type and 5 L-type. The data run to 31 August 2026 (APHA). - Classical cases were confirmed in 2018, 2021, 2024 and on 18 August 2026. The latest was in a cow born in May 2019 in Dumfries and Galloway, detected in routine testing of fallen stock (APHA; Scottish Government, 19 August 2026). - EFSA (2025) says the 2018, 2021 and 2024 cases show that cattle were exposed after the reinforced ban. - A case-control study linked these late cases to home-mixed feed (Ortiz-Pelaez et al., 2012). - A modelling study found a common annual decline of 33.9% across EU member states. It concluded that spontaneous occurrence does not explain most of these cases (Arnold et al., 2017). - On motive. Phillips puts more weight on misunderstanding than on commercial interest: - Officials assumed about 60 infections a month, while the true rate was 10,000 or more a month (para 344). - Nobody considered the minimum infective dose (paras 210–214). - Phillips does note the “shared reluctance” with the feed trade (para 350). - It judged that banning meat and bone meal from all feed in 1994 would have been “disproportionate” without evidence that BSE infected humans (para 346). - I did not re-check the chapter’s specific 1988 source (Inquiry transcript, 29 June 1998). - Later reversal. As risk fell, the EU partly reversed its feed ban: - non-ruminant animal protein was allowed in fish feed from 2013 (Regulation (EU) No 56/2013); - pig protein was allowed in poultry feed, and poultry protein in pig feed, from 2021 (Regulation (EU) 2021/1372). I did not re-read either text.

Verdict: strengthened. - The consequence the chapter asserted without a source is now well quantified. About a quarter of all UK cases came from infections after the 1988 ban, and cases collapsed in the first cohort born under an all-species ban. - The “six or so years” is about right (1988 to 1994–96). - The attribution to renderers’ commercial interest is partly supported. Phillips gives equal or greater weight to ignorance about infective dose and the scale of infection.

Weight. This is strong evidence for digest insight 5 (partial restrictions leak through adjacent uses). Add the counterweight: the partial measure was also dramatically effective. The lesson is to track and close the remaining pathways, not that partial measures fail.

Claim 4. About 1,200 clinical cases in 1987 and the first half of 1988, most of them eaten, could have been removed for about GBP 1.5m (p. 158); about 30,000 infected cattle at least halfway through incubation were eaten between mid-1988 and the end of 1989 (Dealler, 1996; p. 161)#

Original claim. Compensation of “no more than GBP 1 000 per animal” for the roughly 1,200 recorded clinical cases would have totalled “approximately GBP 1.5 million” (p. 158). Between the Southwood recommendation (mid-1988) and the offal ban (end of 1989), “an estimated 30 000 infected cattle … at least halfway through the average incubation period” were consumed (p. 161).

What happened since. - Clinical cases and the delay. - APHA now records 727 confirmed cases before 1988 and 2,184 in 1988. Roughly 1,200 by mid-1988 is therefore plausible, although the counting bases differ. Phillips reports 421 confirmed cases by 22 April 1988 (para 242). - Phillips agrees that the slaughter-and-compensation scheme “could and should have been introduced months earlier” (Executive Summary §7). It was introduced in August 1988. - The chapter’s arithmetic does not add up: 1,200 × £1,000 is £1.2m, not £1.5m. - Total exposure was far larger than the 1988–89 window suggests. - Smith and Bradley (2003) estimate that 1–3 million cattle were probably infected, most of which were slaughtered for food before they showed clinical signs. - Back-calculation using the post-2001 testing data found two to four times more infections than earlier estimates, because clinical cases were substantially under-recorded (Donnelly et al., 2002). - Chen and Wang (2014), drawing on Donnelly et al. (2002), counted slaughtered or unreported cattle in the last year of incubation: 42,809 in 1980–89 and 236,763 in 1990–96. That metric is narrower than Dealler’s “at least halfway through incubation”, and the offal ban cut the infectivity each animal carried. Even so, exposure continued, and by animal numbers grew, after 1989. - The third appendix survey found abnormal prion protein in both the “unexposed” cohorts. One reading of that result is that human exposure began in the late 1970s and continued into the late 1990s (Gill et al., 2020). - Dealler (1996) is an early single estimate. I could not read it, and I found no later study that reproduces or contradicts it directly. It is consistent in order of magnitude with later work. But because the later modelling raised the estimated number of infections, it is more likely an underestimate than an overestimate.

Verdict: partly held up. - The 1,200 figure and the delay are confirmed. - The £1.5m contains an arithmetic slip. - The 30,000 figure is plausible but unverified. - Hindsight undercuts the chapter’s framing. Most infectivity eaten came from cattle with no clinical signs, and exposure ran until 1996. Removing the clinical cases cheaply in 1987–88 was right, and Phillips agrees it was overdue. But it would have averted only a small share of total human exposure.

Weight. Keep this as an example of poor decision quality: a cheap, obviously sensible step was delayed for fiscal and presentational reasons. Do not use it as an estimate of avoidable harm. Digest insight 11 (refusing trivial early costs means larger later ones) stays “suggestive”.

Claim 5. The eventual number of vCJD cases was unknown: perhaps “no more than another 100”, perhaps “up to a million” (Collinge, 1999; p. 164)#

Original claim. Total costs cannot yet be estimated, “not least because we still cannot estimate how many people will eventually succumb” to vCJD (Box 15.3, p. 164).

What happened since. - The UK clinical toll. - The National CJD Research and Surveillance Unit (NCJDRSU) records 178 deaths from definite or probable vCJD: 123 confirmed at post-mortem and 55 probable. None were alive at 2 December 2024. - Deaths peaked at 28 in 2000. There were 5 in 2011, 1 in 2013 and the last in 2016. - The NCJDRSU website now requires a login, so I could not check figures after December 2024. - Worldwide. There were 232 cases worldwide: 178 in the UK, 28 in France and 5 in Spain (Watson et al., 2021). - How projections narrowed. - In 2003, Ghani et al. still gave a 95% confidence interval of 10 to 7,000 future deaths. - In 2010, Garske and Ghani projected a potentially long but uncertain tail. Future cases would come from people with previously little-affected genotypes and from blood transfusion. They found no self-sustaining epidemic. - Genotype. The UK case that died in 2016 was the first definite vCJD case in a person with the MV genotype at codon 129 of the prion protein gene (Mok et al., 2017). All earlier clinical cases were MM. - Secondary transmission. - Among 67 recipients of blood from donors who later developed vCJD, there were 3 clinical vCJD cases and 1 asymptomatic infection. One further asymptomatic infection was found in a patient with a bleeding disorder who had received UK plasma products (UKHSA, 2025). - No transfusion-transmitted case has occurred since all blood was leucodepleted in 1999 (Watson et al., 2021). - Subclinical prevalence. - Appendix surveys estimated about 1 in 4,000 in 2004. - The 2013 survey estimated 493 per million, about 1 in 2,000 (95% CI 282–801 per million). It also found a high share of VV genotypes among the positive samples (Gill et al., 2013). - Appendix-3 then found similar prevalence in people operated on before 1980 and in people born after 1996. That leaves two readings: a background prevalence that may not progress to disease, or exposure lasting from the late 1970s to the late 1990s (Gill et al., 2020). - A survey of 63,007 tonsils found no positives (Clewley et al., as reported by Watson et al., 2021).

Verdict: held up. The chapter predicted nothing. It correctly reported deep uncertainty, and the clinical outcome fell at the low end of its range. The uncertainty it described about subclinical carriers and onward spread persists 25 years later.

Weight. This supports digest insights on latency and irreversible exposure: exposure that cannot be undone produces long-tailed uncertainty that lasts for decades and reaches into adjacent systems such as blood and surgery. Precaution in those adjacent systems, such as leucodepletion, appears to have worked. Do not use the “up to a million” figure rhetorically. It was an early outer bound that did not materialise clinically.

Claim 6. Costs: GBP 4.2bn of forecast UK agriculture-department spending for 1996–2001; a GBP 700m-a-year export trade lost; a 36% real fall in the UK beef market; GBP 1.15bn of lost value added (p. 164)#

Original claim. Box 15.3 (p. 164) gives four figures: - UK agriculture departments’ forecast spending for 1996–2001 was £4.2bn, mostly compensation (House of Commons, 1999); - the export ban of March 1996 cost a trade worth £700m a year; - the value of the UK-produced beef market fell 36% in real terms in the first year; - lost value added came to £1.15bn (DTZ Pieda, 1998).

What happened since. - The forecast is correctly cited. The Public Accounts Committee report BSE: the cost of a crisis (HC 790, 1999) gives £4,210m for 1996–2001. The largest item is £2,240m for the Over Thirty Month Scheme, which kept older cattle out of the food chain. The rest is £1,340m of temporary industry support, £160m for the selective cull and £470m for administration and controls. - Actual public spending (Lord Whitty, HL Deb 27 January 2004, c38W):

Year £m
1996–97 1,548
1997–98 1,017
1998–99 555
1999–2000 412
2000–01 364
2001–02 398
2002–03 479

Verdict: held up. - The figures are accurately drawn from their sources. - The five-year forecast was slightly high, but spending continued to 2005 and beyond. - Export restrictions lasted far longer than the chapter could have known: ten years in the EU and more than 20 in some major markets.

Weight. The cost magnitudes are solid and can be cited with these updates. Two refinements for the lens: - The £1,774m of EC payments equals roughly 36–37% of UK public spending on BSE over 1993–2003 (or 1996–2003). The cost of the failure was therefore partly shared across the EU. - Some of the spending bought back credibility, not safety. MAFF regarded the selective cull and the traceability system as largely irrelevant to eradication, but essential to lifting the ban (PAC, 1999, para 6). Lost credibility had its own price.

Claim 7. Precautionary policy needs regulation separated from sponsorship (the Food Standards Agency), freedom of information, scientific advice separated from policy-making, and plural, open-data research; policy after 1996 had already become “more precautionary” through openness (pp. 165–166)#

Original claim. - Precaution “would have necessitated, firstly and most fundamentally”, separating regulation from sponsorship (p. 165). - It would also have necessitated open discussion of the costs and benefits of a wide range of options. - A Freedom of Information Act “might” change the culture. So “might” institutional separation of advice from policy-making, and pluralistic, open-data research. - The more independent SEAC after 1996 and the Food Standards Agency (FSA) show that policy “has become more precautionary … because it has become more open and accountable” (pp. 165–166).

What happened since. - The reforms were largely enacted. - The FSA began work on 1 April 2000. Its main objective is to protect public health “and otherwise to protect the interests of consumers in relation to food” (Food Standards Act 1999, s. 1). - It may publish its advice to ministers (s. 19). It must take into account risks, “including any uncertainty”, and the costs and benefits of acting or not acting (s. 23). - Section 23 almost exactly writes the chapter’s options-and-costs recommendation into law. - MAFF was replaced by Defra in June 2001 and formally dissolved (SI 2002/794). - The Freedom of Information Act 2000 gave a general right of access from January 2005. - In the EU, the European Food Safety Authority (EFSA) was created under Regulation (EC) 178/2002. That regulation also writes the precautionary principle into food law (Art. 7) and formally separates risk assessment by EFSA from risk management by the Commission. France (1999) and Germany (2002) created separate assessment bodies. I did not re-read these legal texts. - SEAC’s later terms of reference required it to convey “the nature and extent” of scientific uncertainties with its advice. The government later announced it would be dissolved as a non-departmental public body. Its chair said the need for its advice was “much less than it has been” (archived SEAC website, 2011). - How the FSA performed. - Public confidence stayed high. The share of people confident the FSA can be relied on to protect them from food risks was 84% in 2020–21, fell to 78% in 2023 and recovered to 83% from late 2024 (FSA, 25 June 2026). - After the 2013 horsemeat incident, the Elliott Review (2014) found that the 2010 transfer of authenticity testing and compositional labelling from the FSA to Defra had caused confusion. Many stakeholders believed this contributed to the incident. - The National Audit Office (2019) found a regulatory system “showing signs of strain”, with fewer local-authority food-control staff and delayed checks. - The FSA itself combines risk assessment and risk management, yet it kept public trust. Phillips had found the distinction between evaluating and managing risk hard to draw (para 220). The FSA’s first chair described the lesson of BSE as policing the boundary between the two (Krebs, 2005). - EFSA’s independence became its own controversy. The European Court of Auditors (2012) found that none of four EU agencies audited, EFSA included, adequately managed conflicts of interest, although EFSA’s procedures were among the most advanced. Continuing controversy over the independence and transparency of its assessments contributed to the Transparency Regulation (EU) 2019/1381, which has applied since 27 March 2021. I did not re-read it. - Openness cut both ways. - The clearest post-2000 example of an open, costed decision relaxed a precautionary measure. The FSA’s public review of the Over Thirty Month rule found that keeping it would imply about £2bn per death prevented, and recommended a managed move to testing (Krebs, 2005). The rule was replaced in November 2005. - The ban on beef on the bone that followed SEAC’s options-based advice (Beef Bones Regulations 1997, SI 1997/2959), cited by the chapter as good practice, was itself later lifted. I did not trace the revoking instrument. - The authors’ own later view. - Their 2005 book has a chapter titled “BSE and the partial reform of food policy-making”. It compares the FSA, EFSA, France’s AFSSA and Germany’s BVL and BfR. I read the abstract only. - Millstone (2009, Research Policy) asks how far reforms made after the food-safety scares, in eight institutional settings, lived up to the rhetoric used to justify them. I read the abstract only. - Sponsorship and regulation are still combined for animal disease. Defra, the ministry that sponsors farming, remains responsible for animal-disease controls, including BSE surveillance through its agency APHA.

Verdict: partly held up. - The prescriptions were largely adopted, and the FSA has held public trust, which supports the chapter’s direction. - The evidence does not show that openness made policy “more precautionary”. Its visible effect was more explicit, contestable and reversible decisions, including de-escalation. - Separating risk assessment from risk management turned out to be neither necessary for trust (the FSA did not do it) nor sufficient for independence (EFSA’s controversies). - The 2010 fragmentation of the FSA and the horsemeat incident show that institutional gains can be eroded.

Weight. Treat the remedies as good-governance lessons that later practice supports: independence from sponsors, publication of advice, explicit treatment of uncertainty and costs, and accountability for framing. The claim that they produce more precaution should give way to the claim that they produce more proportionate and accountable decisions.

Claim 8. Jurisdictions that put industrial sponsorship and consumer protection in the same institutions (e.g. Ireland) adopted less precautionary approaches than those that split them (e.g. Austria) (p. 163, citing the BASES project)#

Original claim. In Box 15.2 the chapter says “in general” the location of responsibilities framed national responses. Where sponsorship and consumer protection sat in the same institutions, as in Ireland, approaches were less precautionary than where they were split, as in Austria. The source is the EC-funded BASES project, “n.d.” (p. 163).

What happened since. - No systematic test found. I found no systematic comparative study that tests this relationship across countries while controlling for exposure. The BASES website cited is no longer online, and I could not locate its national reports. Edited comparative volumes exist, notably Ansell and Vogel (eds), What’s the Beef? (MIT Press, 2006), but I did not read them. - Exposure confounds the example. - Austria recorded its first native case in 2001 and 5 classical cases in total. Ireland recorded 1,656 (EFSA, 2025, Table 13). - Austria’s precaution was therefore tested against a far smaller hazard. - Irish controls did lag the UK’s (Harney et al., 2003), which fits the chapter’s description: - the ruminant feed ban came in 1990 (UK 1988); - the offal ban came in 1996 (UK 1989); - meat and bone meal was fed to other species until 2001. - Evidence against a simple institutional explanation: - DH, which had no conflict over food, was as eager as MAFF to reassure (Phillips para 1189). - In Germany, where agriculture and health ministries were separate, BSE was found only once testing began in November 2000, and two ministers resigned in January 2001. The York working paper finds that the UK and German risk-management failures followed broadly similar patterns (Beck et al., 2007). - Once EU rules were harmonised after 2001, BSE prevalence fell at a common rate across member states (Arnold et al., 2017). - In a 2002 estimate of infection in cattle born after mid-1996, levels were highest in Greece, Italy and Belgium and comparably low in Great Britain, Germany and France (Donnelly et al., 2002). That ranking does not obviously track institutional design. - Adopted as policy but not tested. Almost every jurisdiction then created separate food-safety or risk-assessment agencies (claim 7), so the hypothesis became policy without being tested.

Verdict: unclear. The claim may be right. But its evidence base is a single project report that is no longer available, the illustrative pair is confounded by exposure, and later evidence neither confirms nor refutes it.

Weight. Downgrade digest insight 2 (promotion plus regulation subordinates protection) from “moderate” to “weak” as a comparative generalisation. Keep it as a well-evidenced within-case mechanism (claim 1). Do not cite Ireland and Austria as evidence.

Claim 9. As of 2000, BSE was rising in France and Ireland and had appeared in Germany and Spain; given differences in national controls and enforcement, some jurisdictions would face animal and public health challenges “for many years” (p. 163)#

Original claim. Box 15.2 (p. 163) says rising cases in France and Ireland and new cases in Germany and Spain mean “some jurisdictions face considerable animal and public health challenges for many years”.

What happened since. - The EU-27 trajectory. Classical BSE cases in today’s 27 EU member states, by year, from EFSA (2025) and its Zenodo data tables:

Year Cases
2000 481
2001 972
2002 1,006
2003 724
2004 518
2005 334
2006 194
2007 92
2010 26
2012 8
2014 2
2015 1
2016 1
2017–2024 0 each year

Verdict: held up. The EU epidemic rose for two more years and then declined over about 15 years to zero classical cases. France bore a significant human toll. Surveillance still runs at a scale of roughly a million tests a year.

Weight. This supports the chapter’s European argument and digest insight 8 (“no evidence of harm” can be produced by not looking). The jump in 2001–02 is a textbook case of an absence of evidence that was really an absence of testing.

Original claim. - MAFF “insisted that the SBO ban was not necessary on scientific grounds”, which made it hard to convince industry the measures mattered. - In 1995, “some 48 %” of abattoirs visited without notice were failing to comply (p. 162, citing HC-331, 1996). - An enforcement representative is quoted calling the controls “a bit of window dressing”.

What happened since. - Phillips confirms the non-compliance and shows it was worse. - When the new Meat Hygiene Service (MHS) took over from local authorities in April 1995, “widespread failures” to stain offal were found. - Every slaughterhouse handling cattle was visited without notice in June 1995. More than half were not meeting the staining rules, and 60 of 435 were not separating offal correctly (paras 452–457). - Once properly enforced, the tonnage of offal reaching renderers more than doubled. That suggests about a third had previously evaded the controls (para 461). - Spinal cord was found in four carcasses that had passed inspection (para 462). - The 1990 ban on offal in animal feed was “unenforceable and widely disregarded” (Executive Summary §6). - Phillips supports the chapter’s causal link, as one cause among several. - The precautionary measures were never presented as protecting against a real risk to humans. That possibility “was not communicated to the public or to those whose job it was to implement and enforce” them (key conclusions). - The human offal ban was presented in terms that “underplayed its importance as a public health measure” (Executive Summary §7). - Phillips attributes the “lack of diligence in implementing Regulations” in part to “the success of continuous efforts” to prevent public concern (para 1186). - It also lists other causes (para 456): local-authority budgets and recruitment problems, especially for vets; weak line management; meat inspectors becoming “almost part of the plant staff”; local reluctance to press marginal slaughterhouses; and a deregulatory “light touch” culture. - Rapid improvement once enforcement was taken seriously. - After the minister personally insisted on full compliance, the share of slaughterhouses failing in some respect fell from 31% in October to 13% in November 1995. By January 1996 only 5% of 344 visits were unsatisfactory (paras 463–464). - Problems persisted in lesser form. An EU inspection in 2004 still found “shortcomings … in various areas in Great Britain” (Regulation (EC) No 657/2006, recital 8).

Verdict: held up. The factual claim is confirmed and the scale was, if anything, larger. The causal link to MAFF’s downplaying of the rules is endorsed by the official inquiry as one of several causes. The quick turnaround once the purpose was stressed and enforcement was centralised supports the chapter’s reading.

Weight. Digest insight 6 (understating why a rule exists erodes compliance) can move from “moderate” to “strong” for this case. Phillips’s list of other causes should travel with it: resourcing, capture of inspectors by the plants, and a deregulatory climate.

Minor factual checks#

Implications for the section’s transferable insights#

The numbering follows the digest.

  1. Categorical safety claims make later precaution self-incriminating. Strong. Phillips’s “sedation” and “leaning into the wind” findings (paras 1179–1185) and the MRM minute (“It was agreed not to raise it”, p. 162) confirm it. Frame it as over-reassurance, not deception.
  2. Promotion plus regulation subordinates protection. Moderate within the case; weak as a comparative law. The evidence is Phillips paras 1173 and 1189, the unavailable BASES source, and the confounded Ireland–Austria pair.
  3. Protective actions read as signals get suppressed. Strong. Examples: notifiability (p. 159), MRM and head meat (p. 162), and the exclusion of the public health laboratory service (p. 163).
  4. Controls designed around convenience leave residual exposure. Strengthened. Dorsal root ganglia in MRM until December 1995, spinal cord in inspected carcasses, cases in cattle born after both feed bans, and EFSA’s reading of the 2018–2024 cases.
  5. Partial restrictions leak through adjacent uses and trade. Strengthened and quantified: about 23% of UK cases, and a collapse in the first cohort born under the all-species ban. Add that the partial measure was also highly effective.
  6. Understating why a rule exists erodes compliance. Strong for this case (Phillips para 1186), alongside resourcing and inspector-capture causes.
  7. Curated advisers and mixed advice presented as “purely scientific”. Moderate to strong. Phillips on Southwood’s unexplained basis and lost caveats, and on SEAC’s muted 1994 reassessment, supports it. Later institutions encoded the remedy: SEAC’s uncertainty remit and FSA s. 23.
  8. “No evidence of harm” produced by not looking. Strengthened. EU testing from 2001 found BSE where it had been assumed absent. Great Britain under-recorded clinical cases two- to fourfold. Appendix surveys found hidden human carriage.
  9. Accountable decision-makers, not scientists, should own the level of protection. Supported by later practice. The Over Thirty Month decision priced protection explicitly (about £2bn per death prevented) and decided it in public.
  10. Strategies that depend on controlling information fail abruptly. Held. The March 1996 collapse, public “betrayal”, and export restrictions lasting ten years or more.
  11. Refusing trivial early costs means paying far larger later ones. Suggestive only. Public spending was £288m in 1986–96 and about £4.8bn in 1996–2003, but these are not a like-for-like counterfactual. Much of the later cost bought back credibility and market access rather than reducing risk.

New lessons from hindsight that the chapter does not draw: - Precaution needs exit criteria. Several BSE measures were later relaxed through open, costed processes: the Over Thirty Month rule, the EU feed ban, and negligible-risk status. Openness served proportionality in both directions. - Latent, irreversible exposure moves the risk into adjacent systems such as blood and surgery, and keeps uncertainty open for decades. Precaution in those adjacent systems, such as leucodepletion in 1999, appears to have worked. - “Eradication” is not a stable end state. Atypical forms and occasional late classical cases, including one in a cow born in 2019, require surveillance to continue indefinitely. - Lost credibility carries its own large, separate cost: measures needed to restore trust and trade rather than to reduce risk.

Method and access notes#

Sources#

BSE Inquiry and Parliament

  1. Phillips, Lord, Bridgeman, J., Ferguson-Smith, M. The BSE Inquiry Report, Vol. 1: Findings and Conclusions. HC 887-I. London: The Stationery Office, October 2000. Copy used: https://jesip.org.uk/uploads/media/incident_reports_and_inquiries/BSE%20Outbreak%20Conclusion.pdf (also at https://www.infectedbloodinquiry.org.uk/sites/default/files/documents/CABO0000141_004%20-%20The%20BSE%20Inquiry%20Report%20Volume%201%20Findings%20and%20Conclusions%20-%2003%20Nov%202000.pdf)
  2. The BSE Inquiry Report, Vol. 10: Economic Impact and International Trade (October 2000), paras 2.66, 4.34, 5.3, 5.6. Wayback copy of the contents page (captured 3 October 2002): http://web.archive.org/web/20021003125536/http://www.bse.org.uk:80/report/volume10/toc.htm
  3. Hansard, HC Deb 7 June 1990, “Agricultural Council” (statement and exchanges, around cols 906–907). https://api.parliament.uk/historic-hansard/commons/1990/jun/07/agricultural-council
  4. House of Commons Committee of Public Accounts. BSE: The Cost of a Crisis. HC 790, Session 1998–99 (1999). Wayback copy (captured 27 October 2016): http://web.archive.org/web/20161027104622/http://www.publications.parliament.uk/pa/cm199899/cmselect/cmpubacc/790/79003.htm
  5. HL Deb 27 January 2004, c38W (Lord Whitty, written answer on the cost of BSE to public funds, 1993–94 to 2002–03). https://www.theyworkforyou.com/wrans/?id=2004-01-27a.38.3
  6. HC Deb 15 March 2007, c488W (Barry Gardiner, written answer: replacing the Over Thirty Month scheme saved about £300m a year). https://www.theyworkforyou.com/wrans/?id=2007-03-15b.126666.h

Official statistics and announcements

  1. Animal and Plant Health Agency. Cattle: TSE surveillance statistics (page last updated 16 September 2026). https://www.gov.uk/government/statistics/cattle-tse-surveillance-statistics. Data files: Current BSE statistics (data to 31 August 2026), https://assets.publishing.service.gov.uk/media/6aaa5992f1f8d2a39605f884/Current_BSE_Statistics.xlsx; Historical BSE statistics (data to 31 May 2025), https://assets.publishing.service.gov.uk/media/68a6dfc22a1dfc29763d51b5/Historical_BSE_Statistics.xlsx
  2. Scottish Government. “BSE” (press release on a classical BSE case in Dumfries and Galloway), 19 August 2026. https://www.gov.scot/news/bse-3/
  3. Defra. “UK international risk status for BSE downgraded in huge boost to farm sector”, 3 June 2025 (as republished by Wired-Gov). https://www.wired-gov.net/wg/news.nsf/articles/uk+international+risk+status+for+bse+downgraded+in+huge+boost+to+farm+sector+03062025142000?open=
  4. APHA Science Blog. “BSE: achieving negligible risk status”, 22 October 2025. https://aphascience.blog.gov.uk/2025/10/22/bse-achieving-negligible-risk-status/
  5. EFSA. The European Union summary report on surveillance for the presence of TSE in 2024. EFSA Journal 2025; 23: e9732 (Tables 13–14 and text). https://doi.org/10.2903/j.efsa.2025.9732. Historical tables 15–16: https://doi.org/10.5281/zenodo.17378970 (published 10 November 2025)
  6. EFSA. The European Union summary report on surveillance for the presence of TSE in 2023. EFSA Journal 2024 (abstract). https://doi.org/10.2903/j.efsa.2024.9097
  7. National CJD Research and Surveillance Unit. Creutzfeldt-Jakob disease in the UK (by calendar year), updated 2 December 2024. Wayback copy (captured 14 December 2024): http://web.archive.org/web/20241214193026/https://www.cjd.ed.ac.uk/sites/default/files/figs.pdf
  8. UK Health Security Agency. Creutzfeldt-Jakob disease (CJD) surveillance update 2024, 26 June 2025 (data to 31 December 2024). https://www.gov.uk/government/publications/creutzfeldt-jakob-disease-cjd-surveillance-update-2024
  9. gov.uk announcements on reopening export markets (titles and summaries read only): “China lifts ban on exports of British beef” (27 June 2018), https://www.gov.uk/government/news/china-lifts-ban-on-exports-of-british-beef; “China opens doors to British beef” (18 October 2019), https://www.gov.uk/government/news/china-opens-doors-to-british-beef; “Japan opens market to British lamb and beef” (10 January 2019), https://www.gov.uk/government/news/japan-opens-market-to-british-lamb-and-beef; “First exports of UK beef to the USA in 20 years underway” (30 September 2020), https://www.gov.uk/government/news/first-exports-of-uk-beef-to-the-usa-in-20-years-underway

Epidemiology and the human toll

  1. Smith, P. G., Bradley, R. Bovine spongiform encephalopathy (BSE) and its epidemiology. British Medical Bulletin 2003; 66: 185–198 (abstract). https://doi.org/10.1093/bmb/66.1.185
  2. Donnelly, C. A., Ferguson, N. M., Ghani, A. C., Anderson, R. M. Implications of BSE infection screening data for the scale of the British BSE epidemic and current European infection levels. Proc R Soc B 2002; 269: 2179–2190 (abstract). https://doi.org/10.1098/rspb.2002.2156
  3. Chen, C.-C., Wang, Y.-H. Estimation of the exposure of the UK population to the BSE agent through dietary intake during the period 1980 to 1996. PLoS ONE 2014; 9: e94020. https://doi.org/10.1371/journal.pone.0094020
  4. Anderson, R. M., et al. Transmission dynamics and epidemiology of BSE in British cattle. Nature 1996; 382: 779–788 (abstract only). https://doi.org/10.1038/382779a0
  5. Arnold, M. E., Simons, R. R. L., Hope, J., Gibbens, N., Adkin, A. L. Is there a decline in BSE cases born after reinforced feed bans? A modelling study in EU member states. Epidemiology and Infection 2017 (abstract). https://doi.org/10.1017/S0950268817001236
  6. Ortiz-Pelaez, A., et al. Case-control study of cases of BSE born after July 31, 1996 (BARB cases) in Great Britain. Veterinary Record 2012 (abstract). https://doi.org/10.1136/vr.100097
  7. Watson, N., et al. The importance of ongoing international surveillance for Creutzfeldt–Jakob disease. Nature Reviews Neurology 2021; 17: 362–379. https://doi.org/10.1038/s41582-021-00488-7
  8. Ritchie, D. L., Peden, A. H., Barria, M. A. Variant CJD: reflections a quarter of a century on. Pathogens 2021; 10: 1413. https://doi.org/10.3390/pathogens10111413
  9. Gill, O. N., et al. Prevalent abnormal prion protein in human appendixes after BSE epizootic: large scale survey. BMJ 2013; 347: f5675. https://doi.org/10.1136/bmj.f5675
  10. Gill, O. N., et al. Prevalence in Britain of abnormal prion protein in human appendices before and after exposure to the cattle BSE epizootic. Acta Neuropathologica 2020 (abstract). https://doi.org/10.1007/s00401-020-02153-7
  11. Mok, T., et al. Variant Creutzfeldt–Jakob disease in a patient with heterozygosity at PRNP codon 129. NEJM 2017; 376: 292–294 (citation only). https://doi.org/10.1056/NEJMc1610003
  12. Ghani, A. C., Ferguson, N. M., Donnelly, C. A., Anderson, R. M. Factors determining the pattern of the vCJD epidemic in the UK. Proc R Soc B 2003 (abstract). https://doi.org/10.1098/rspb.2002.2313
  13. Garske, T., Ghani, A. C. Uncertainty in the tail of the variant Creutzfeldt-Jakob disease epidemic in the UK. PLoS ONE 2010; 5: e15626 (abstract). https://doi.org/10.1371/journal.pone.0015626
  14. Collinge, J. Variant Creutzfeldt-Jakob disease. Lancet 1999; 354: 317–323 (abstract). https://doi.org/10.1016/S0140-6736(99)05128-4
  15. Hewitt, P. E., Llewelyn, C. A., Mackenzie, J., Will, R. G. Three reported cases of vCJD transmission following transfusion of labile blood components. Vox Sanguinis 2006 (citation only). https://doi.org/10.1111/j.1423-0410.2006.00837.x
  16. Harney, M. S., et al. vCJD risk in the Republic of Ireland. BMC Infectious Diseases 2003; 3: 28. https://doi.org/10.1186/1471-2334-3-28

EU law and trade

  1. Commission Regulation (EC) No 657/2006 of 10 April 2006 (OJ L 116, 29.4.2006, p. 9), lifting the UK restrictions; recitals 3–14. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32006R0657
  2. Commission Decision 1999/514/EC of 23 July 1999 (date-based export scheme, from 1 August 1999). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31999D0514
  3. Court of Justice, Case C‑1/00 Commission v France, judgment of 13 December 2001. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:62000CJ0001
  4. Regulation (EC) No 999/2001 (TSE Regulation), not re-read. https://eur-lex.europa.eu/eli/reg/2001/999/oj
  5. Council Decision 2000/766/EC (EU-wide suspension of processed animal protein in the feed of farmed animals from 1 January 2001), not re-read. https://eur-lex.europa.eu/eli/dec/2000/766/oj
  6. Commission Regulation (EU) No 56/2013 (non-ruminant processed animal protein in aquaculture feed), not re-read. https://eur-lex.europa.eu/eli/reg/2013/56/oj
  7. Commission Regulation (EU) 2021/1372 (pig and poultry processed animal protein cross-feeding), not re-read. https://eur-lex.europa.eu/eli/reg/2021/1372/oj
  8. Regulation (EC) No 178/2002 (General Food Law; EFSA; Art. 7 precautionary principle), not re-read. https://eur-lex.europa.eu/eli/reg/2002/178/oj
  9. Regulation (EU) 2019/1381 (Transparency Regulation; applied from 27 March 2021), not re-read. https://eur-lex.europa.eu/eli/reg/2019/1381/oj

Institutions and governance after 2000

  1. Food Standards Act 1999, ss. 1, 19, 22, 23. https://www.legislation.gov.uk/ukpga/1999/28
  2. The Ministry of Agriculture, Fisheries and Food (Dissolution) Order 2002, SI 2002/794. https://www.legislation.gov.uk/uksi/2002/794/contents/made
  3. Freedom of Information Act 2000. https://www.legislation.gov.uk/ukpga/2000/36
  4. Krebs, J. R. The Croonian Lecture 2004. Risk: food, fact and fantasy. Phil Trans R Soc B 2005 (read through a summarising tool). https://doi.org/10.1098/rstb.2005.1665
  5. Elliott, C. Elliott Review into the Integrity and Assurance of Food Supply Networks: Final Report (July 2014; published 4 September 2014), para 7.7. https://assets.publishing.service.gov.uk/media/5a7e20b740f0b62305b80fb8/elliot-review-final-report-july2014.pdf
  6. National Audit Office. Ensuring food safety and standards. HC 2217, 12 June 2019. https://www.nao.org.uk/reports/ensuring-food-safety-and-standards/
  7. Food Standards Agency. FSA Strategy: Annual Update on Progress Indicators (FSA 26/06/04), 25 June 2026, paras 3.5–3.12. https://www.gov.uk/government/publications/fsa-260604-fsa-strategy-annual-update-on-progress-indicators
  8. European Court of Auditors. Management of conflict of interest in selected EU agencies. Special Report No 15/2012. https://www.eca.europa.eu/Lists/ECADocuments/SR12_15/SR12_15_EN.PDF
  9. Spongiform Encephalopathy Advisory Committee website (statement on dissolution; terms of reference). Wayback copy (captured 26 February 2011): http://web.archive.org/web/20110226063114/http://www.seac.gov.uk:80/
  10. The Beef Bones Regulations 1997, SI 1997/2959. https://www.legislation.gov.uk/uksi/1997/2959/made

Later historiography and commentary

  1. Forbes, I. Making a crisis out of a drama: the political analysis of BSE policy-making in the UK. Political Studies 2004 (abstract). https://doi.org/10.1111/j.1467-9248.2004.00483.x
  2. van Zwanenberg, P., Millstone, E. BSE: a paradigm of policy failure. Political Quarterly 2003; 74 (not read). https://doi.org/10.1111/1467-923X.00509
  3. van Zwanenberg, P., Millstone, E. BSE: Risk, Science and Governance. Oxford University Press, 2005 (contents and the abstract of Ch. 9, “BSE and the partial reform of food policy-making”, read through OpenAlex). https://doi.org/10.1093/acprof:oso/9780198525813.001.0001
  4. Millstone, E. Science, risk and governance: radical rhetorics and the realities of reform in food safety governance. Research Policy 2009; 38: 624–636 (abstract, via the Sussex figshare record https://figshare.com/articles/journal_contribution/Science_risk_and_governance_radical_rhetorics_and_the_realities_of_reform_in_food_safety_governance/23364047). https://doi.org/10.1016/j.respol.2009.01.012
  5. Beck, M., Kewell, B., Asenova, D. BSE crisis and food safety regulation: a comparison of the UK and Germany. University of York Management School Working Paper No. 38, 2007 (not peer reviewed). https://eprints.whiterose.ac.uk/3477/
  6. Ansell, C., Vogel, D. (eds). What’s the Beef? The Contested Governance of European Food Safety. MIT Press, 2006 (not read). https://doi.org/10.7551/mitpress/7216.001.0001
  7. Alarcon, P., Wall, B., Barnes, K., Arnold, M., Rajanayagam, B., Guitian, J. Classical BSE in Great Britain: review of its epidemic, risk factors, policy and impact. Food Control 2023; 146: 109490 (not read; access blocked). https://doi.org/10.1016/j.foodcont.2022.109490

EEA 2013 volume

  1. EEA. Late lessons from early warnings: science, precaution, innovation, EEA Report No 1/2013, Annexes: Annex 2, Table A2.1 (p. 702) and pp. 716 ff. (Table A2.15); index. https://www.eea.europa.eu/en/analysis/publications/late-lessons-2/late-lessons-chapters/late-lessons-ii-annex-1/@@download/file
  2. EEA 2013, Ch. 22 (pp. 545–546). https://www.eea.europa.eu/en/analysis/publications/late-lessons-2/late-lessons-chapters/late-lessons-ii-chapter-22/@@download/file
  3. EEA 2013, Ch. 25, “Why did business not react with precaution to early warnings?” (p. 612). https://www.eea.europa.eu/en/analysis/publications/late-lessons-2/late-lessons-chapters/late-lessons-ii-chapter-25/@@download/file
  4. EEA 2013, Ch. 27, “More or less precaution?” (Box 27.1, p. 649; p. 658). https://www.eea.europa.eu/en/analysis/publications/late-lessons-2/late-lessons-chapters/late-lessons-ii-chapter-27/@@download/file