LL1-15 digest: Ch15 ‘Mad cow disease’ 1980s–2000: how reassurances undermined precaution#

Source: EEA, Late lessons from early warnings (2001), report pp. 157–167 (PDF pp. 157–167). Authors: Patrick van Zwanenberg and Erik Millstone. The chapter gives no affiliations. [Background – verify: science-policy researchers at SPRU, University of Sussex.] Other voices: No panels. Box 15.4 summarises the Phillips Inquiry. The authors accept its main criticism and dispute two of its conclusions. (The boxes are unsigned; their wording suggests the authors wrote them. Only Table 15.1 is marked “Source: EEA”.) Full notes: notes/LL1-15.md

Core argument#

UK policy-makers, including Gillian Shephard and John Gummer, whose 1998 Inquiry testimony is cited (fn 11) [Background – verify: both former Ministers of Agriculture], later claimed their handling of BSE had been ultra precautionary and rigorously science-based (p. 157). That is the authors’ paraphrase, not a direct quotation. Drawing mainly on the BSE Inquiry’s record, the authors argue the opposite.

Root cause: one ministry (MAFF) both promoted agriculture and protected consumers, so exports, profits and public spending came before health. In the long run this failed the industry and the Exchequer too (pp. 157, 164–165).
The distinctive twist: reassurance itself became the obstacle. Once beef had been declared safe, any further protective step, however cheap, implied the reassurance had been false. The authors say measures were then refused “not because of their immediate costs” but for presentational reasons (pp. 161–162). Their own evidence shows cost and export concerns were intertwined with presentation.

Key evidence#

Early warnings overridden (1986–88) - Officials saw a possible human risk from 1986 (pp. 157–158). - For the first 20 months, fear of undermining confidence in British beef at home and abroad, and reluctance to increase public spending, dominated MAFF policy-making. Even making the disease notifiable was rejected partly because it might signal that meat or milk was dangerous. The Chief Medical Officer learned of BSE 17 months after MAFF (pp. 159–160). - About 1,200 clinical cases were recorded in 1987 and the first half of 1988, and most were sold as food. They could have been removed for about £1.5m (p. 158). In February 1988 the minister rejected his officials’ advice to slaughter and compensate. The reasons given were unidentified “offsetting savings”, consistency with a rejected sugar-beet compensation claim, and exports (p. 159).

Controls shaped by commercial interests (1988–89) - The 1988 feed ban covered ruminants only; an all-species ban was rejected because it would have cost renderers their main market. Cross-contamination followed for about six years, “greatly prolonging” the epidemic (uncited) (p. 160). - The 1989 offal ban covered tissues chosen because they were easiest to remove and of lowest commercial value, not because only they were infective. It came only after an adviser disclosed he had recommended an offal ban to the pet-food industry, and after the meat-products and pet-food industries told MAFF they would remove offal themselves (p. 160). - An estimated 30,000 infected cattle, at least halfway through incubation, were eaten between mid-1988 and the end of 1989. This is a single estimate (Dealler 1996) (p. 161).

Public certainty, private doubt (1989–90) - In May 1990 the government’s expert committee (SEAC) advised that “no risk” could not be stated categorically. In June 1990 the minister told Parliament there was “clear scientific evidence” that beef was “perfectly safe” (p. 161). - A draft SEAC document on beef safety, intended for the Chief Medical Officer, was circulated for amendments. The MAFF official forwarding it told ministers that its “most potentially inflammatory” drafting had been edited out (p. 161). - Two cheap measures were blocked lest they invite demands for more: a ban on mechanically recovered meat (“It was agreed not to raise it”), and a head-meat control proposed by the environmental health officers’ institute. In the head-meat case, ministers wanted the ban and civil servants blocked it (p. 162).

Weak enforcement and controlled knowledge - MAFF said its own offal ban was not necessary on scientific grounds. The authors link this to weak compliance: in 1995, unannounced visits found some 48% of abattoirs visited failing to comply, and an enforcement representative said the rules had seemed “a bit of window dressing” (p. 162). The causal link rests on that one testimony. - Advisers were “carefully selected”. This rests on one official’s retrospective testimony, and no excluded expert is named. The national disease-surveillance service was excluded. A key scrapie-to-cattle experiment did not start until 1996, and the only random abattoir survey (1999) covered cattle outside the food chain (pp. 162–164).

Europe and costs - UK exports of meat and bone meal to the EU roughly doubled from 1988 to 1989, after the domestic feed ban, and some was fed to cattle. EU-wide legislation lagged six years (p. 163). - A forecast £4.2bn in public spending (1996–2001), mostly compensation to industry. A £700m-a-year export trade lost, and a 36% real fall in the value of the UK beef market in the first year. Future vCJD cases were estimated at anywhere from 100 to a million (p. 164).

Authors’ lessons and recommendations#

Diagnosis: regulations were “too little and too late” and poorly enforced. The reassuring narrative was sustained by controlling information and advice, and by advisers’ acquiescence in presenting mixed scientific and non-scientific advice as “purely scientific” (pp. 164–165).

Remedies (p. 165). Two are framed as what precaution “would have necessitated”: - separate regulation from sponsorship (“firstly and most fundamentally”) - acknowledge how little “sound science” was available, and discuss openly the costs and benefits of a wide range of options, including inaction

Others are hedged (“might”): - a Freedom of Information Act - separating scientific advice and research from policy-making - plural, open-data research

Ministers should own and justify the framing assumptions; advisers should show how they gathered and interpreted the evidence.

The authors cite the Food Standards Agency and the more independent post-1996 SEAC as early signs of progress. SEAC’s options-based advice on beef on the bone is their example (pp. 165–166).

Main mechanisms#

The central mechanism is a self-reinforcing reassurance trap: a convenient hypothesis led to public certainty and claims of total control. That meant refusing incremental measures and avoiding new knowledge, which ended in brittle collapse (pp. 161–164). Around it sit:

a sponsor–regulator conflict of interest, at UK and EU level
protective actions treated as signals of danger (pp. 159, 162–163)
selective use of analogies with known diseases (pp. 157–161), which is the note-taker’s synthesis rather than a mechanism the authors name
expert advice used for legitimation, including self-censorship by advisers (pp. 160, 162, 165)
leakage through partial bans and exports (pp. 160, 163)
latency and irreversible exposure (pp. 161, 164)

Transferable insights (technology-neutral)#

Categorical safety claims make later precaution self-incriminating and collapse a range of graded options into “hold the line” or “do everything” (pp. 161–162). Strong: contemporaneous minutes, though presentation and cost motives are intertwined.
Promotion and regulation in one body subordinates protection (pp. 157, 164–165). Strong for this case; moderate as a generalisation, since comparative data are thin (p. 163).
Protective actions judged as signals get suppressed, even when they cost nothing (pp. 159, 162–163). Strong within the case.
Controls designed around convenience, then presented as complete, leave residual exposure (pp. 160–161). Strong.
Partial restrictions leak through adjacent uses and trade (pp. 160, 163). Strong on the facts; moderate on the size of the effect.
Understating why a rule exists erodes compliance (p. 162). Moderate.
Curated advisers, and value-laden advice presented as “purely scientific”, turn expertise into legitimation (pp. 160–165). Strong on the facts; moderate on interpretation. Adviser selection rests partly on retrospective testimony.
“No evidence of harm” can be produced by not looking (pp. 163–164). Moderate.
Science cannot set the level of protection; accountable decision-makers should own the framing (pp. 159, 165). Asserted.
Strategies that depend on controlling information fail abruptly (p. 164). Moderate.
Refusing trivial early costs can mean paying far larger later ones (pp. 158, 164). Suggestive: the two figures are not a like-for-like counterfactual.

Main caveats#

Contested intent. The narrow charge, that policy-makers claimed a certainty they “knew to be unavailable” (p. 161), is well documented: they were told certainty was impossible. The broader charge, that government “covertly subordinated” health to agricultural sales (p. 164), goes further than the documents show about what ministers believed. It conflicts with the Phillips Inquiry’s findings that government was anxious to protect health and that MAFF did not lean towards producers. The authors find those findings “difficult to see” as consistent with the evidence (p. 165).
Mixed evidence base. Contemporaneous minutes sit alongside retrospective testimony given after the 1996 crisis. The chapter does not distinguish the two, and some items presented as contemporaneous (Meldrum’s June 1988 remark, SEAC’s May 1990 advice) are cited only to 1998 Inquiry transcripts.
Untested counterfactuals. That candour would have allowed more precaution without an earlier market collapse, and that the blocked measures would have cut exposure materially, are assumed, not tested. The chapter does not assess whether the imperfect 1988–89 controls still brought the epidemic down.
Omissions and errors:
no counts of cattle or human cases
no named dissenting scientists (warnings from “the wider scientific community” are mentioned only in general terms, pp. 161, 164)
no assessment of whether post-1996 measures were proportionate
no mention of the Southwood committee’s own reassuring conclusion [Background – verify: its 1989 report judged human health implications “most unlikely”]
Table 15.1 groups the feed ban under 1989, though it was announced in June 1988 (pp. 160, 163)
£1.5m does not match 1,200 cases × ≤£1,000 (p. 158)
What the case supports. It strongly supports procedural precaution: honesty about uncertainty, independence, openness and investment in knowledge. Much of this matches ordinary good-governance lessons. It says little about how strong substantive precaution should be.