BSE: A Paradigm of Policy Failure

1

This is a pre-copy-edited, author-produced PDF of an article that was subsequently published in Political Quarterly. The citation is: van Zwanenberg, P. & Millstone, E. (2003) ‘BSE - A Paradigm of Policy Failure’, Political Quarterly, Vol. 74, No. 1, pp. 27-37

BSE: A Paradigm of Policy Failure PATRICK VAN ZWANENBERG AND ERIK MILLSTONE A political crisis over BSE erupted on 20 March 1996, some three and a half years after the veterinary epidemic had passed its peak. By 2002 the reported incidence of BSE in the UK was lower than at any time since 1987. The incidence of variant CJD has not been rising as rapidly as some had feared. It will soon be lawful to sell British beef throughout the entire EU, including France. The report of the public inquiry into BSE was published in October 2000, and the government published its considered response in September 2001. The Ministry of Agriculture was replaced in summer 2001 by the Department for Environment, Food and Rural Affairs (DEFRA), and the Food Standards Agency has been operating since April 2000. One might be forgiven for thinking that BSE had lost much of its previous scientific and political significance, and that the relevant lessons had all been learned. In this paper we will argue that, while some progress has been made, not all the lessons of the BSE saga have yet been learned. The importance of the BSE (bovine spongiform encephalopathy or ‘mad cow disease’) saga does not lie just in its uniqueness or the severity of its consequences. It is important, in part, because of the many respects in which it was unexceptional. The way in which the politics and science of BSE were entangled within UK policy-making institutions, the way in which those relationships were officially misrepresented, and the problems to which those practices subsequently gave rise, are characteristic of many riskappraisal and decision-making regimes across a wide range of different types of risks and different national and multinational jurisdictions.

The UK, along with virtually all national and multilateral jurisdictions, is struggling with a crisis of science and governance. Many leading actors and commentators acknowledge that traditional ways of conducting, and representing, science-based risk policy-making processes have become scientifically and democratically unsustainable, and that a new approach must be developed. There is, however, far less agreement about the alternatives. The challenge of deploying science in policy decision-making processes in ways that are properly democratic without being anti-scientific, and rigorously scientific without being anti-democratic, has yet to be convincingly met. An analysis of the history of the BSE saga provides an unrivalled opportunity to reflect on the challenge. The episode has been subject to a more detailed public examination than any other science-based risk policy. Lord Phillip’s two-and-a-half-year public inquiry into the UK government’s handling of BSE was exhaustive in its investigation of those aspects of policy selected for analysis and it assembled a stunningly comprehensive documentary archive, all of which is publicly available.1 Diagnosing the shortcomings in BSE policy-making The Phillips Report highlighted 136 separate ‘lessons to be learned’ from the BSE saga. As far as the adequacy of the government’s policy response to BSE between 1986 and 1996 was concerned, shortcomings in the design of measures, bureaucratic delays in their introduction and problems with enforcement were singled out for criticism. Overall, however, Phillips and his colleagues concluded that the government’s policies had been fundamentally sensible and appropriate.2 They argued that, because animal feeding practices had evolved to include the recycling of rendered carcass remains in feed rations, and given the unusual nature of the disease itself, BSE would have been ‘a tragic disaster’ whatever policy response had been adopted.3 That suggests that the government was blown off course by an unmanageable challenge, rather than that the manner in which it sought to respond was flawed. The inquiry Report acknowledged that there had been some serious failures, but argued that most of the important mistakes concerned communication, rather than more fundamental problems with the substantive business of assessing and managing risk. For

2

nearly a decade ministers and senior officials had repeatedly assured the public that British beef was unproblematically safe, though sufficient evidence to support such assertions was never available. According to Phillips and his colleagues, a major consequence of this risk communication strategy was that when the government performed a volte-face in March 1996, acknowledging that BSE had probably been transmitted to humans causing a new variant of CJD, the public concluded that it must have been deliberately misled. That failure has been blamed for triggering an evaporation of trust in both government institutions and scientific expertise. Phillips and his colleagues were right to emphasise that, given intensive farming practices, the long incubation period of BSE, and the inability to diagnose the disease before clinical symptoms appeared, the risks could not have been completely avoided. The policy challenge posed by BSE always has been complex and difficult; but it need not necessarily have been catastrophic. We contend that the UK government’s response to the emergence of BSE failed not just because it did not enable the general public to understand and accept that policy-makers were being sensible, competent and reasonable. It failed, first, because it subordinated consumer protection and public health to an economic and political agenda. Second, it misrepresented the objectives of policy, and the nature of, and grounds for, policy decisions. Third, it sought to maintain those misrepresentations by acquiring, interpreting and representing scientific information and advice in ways that were unscientific and anti-scientific. The policy-making context The policy decisions that produced the BSE saga need to be understood in their political and institutional contexts. The macropolitical context included a postwar consensus that it was necessary, desirable and appropriate for the state, in this context the Ministry of Agriculture, Fisheries and Food (MAFF), actively to support the commercial and industrial objectives of farmers and the food industries, and to give that goal a very high priority within its overall portfolio. A key institutional characteristic of MAFF was also that its remit included responsibility for promoting the agricultural and food industries and for regulating them to protect consumers, public and environmental health. This feature was characteristic of a widespread traditional practice, in both the UK and

3

internationally, of locating within individual institutions responsibility for both promoting and regulating particular industrial sectors. MAFF was, moreover, constrained by the ideological policy assumptions and aspirations of the Thatcher government, which included deregulatory and anti-regulatory aspirations, and a desire to reduce public expenditure. It was within that context that MAFF sought to represent its policy decisions as based on, and only on, sound science. BSE: the initial challenge to policy-makers in the mid-1980s When BSE was first recognised as a novel brain disease of UK cattle in November 1986, the policy challenge was a difficult one. A very similar disease that occurs in sheep, known as scrapie, was not thought to present any risk to humans. No one could be sure, however, that BSE would be similarly innocuous. As one senior MAFF scientist put it accurately, but privately, in 1988 ‘We cannot answer the question is BSE transmissible to humans. That natural experiment is underway in the human population and it remains for epidemiologists to collect data and produce a hypothesis based on it.’4 That was why the Department of Health (DH) established the CJD Surveillance Unit in 1990, on advice from the Southwood Working Party. In the interim, however, many experimental scientific steps could have been taken to diminish many of the key policy-relevant uncertainties, of which only relatively few were taken before March 1996. Eradicating the potential risks from BSE, over and above those that had already and unavoidably been taken prior to the official discovery of the disease, was always going to be difficult and expensive. By the time BSE reached the government’s policy agenda, virtually the entire national herd had potentially been exposed to contaminated animal feeds containing recycled slaughterhouse wastes. During the long incubation period infected animals were impossible to identify, but they were nevertheless a significant source of infectivity when incorporated into either animal or human food chains. The pivotal policy challenge was to decide what steps, if any, should be taken to diminish or to eradicate the unknown and indeterminate risks. A full and rapid BSE eradication programme in the UK in the mid-1980s would have required the slaughter and disposal of almost the entire UK herd, and the ‘decontamination’ of the entire animal feed chain. It would have cost in the order of £20 billion at current prices. It is difficult to

4

imagine any government, then or now, having been keen, or even willing, to devote such a huge sum to an entirely unknown and speculative risk. Very many other policy options were available, however, that could have provided a range of different levels of protection at far lower cost. The main drivers of BSE policy Whenever policy decisions about BSE and public health were taken prior to March 1996, none of the participants could be certain whether BSE posed a risk to human health. There was no direct evidence that BSE would be pathogenic to humans, but that possibility could never be ruled out. The documentary records show, however, that on all but a handful of occasions, the dominant concern of policy-makers was their aspiration to maintain stable levels of demand for British beef in domestic and export markets, and their desire not to increase commercial costs or public expenditures. A concern with public health was never a dominant factor in ministerial decision-making. Whenever market stability or departmental expenditures were threatened, the preferred policy was to gamble that the risks to public health would turn out to be zero or so low as to be acceptably slight. Public legitimations of those decisions always represented them, however, as science-based and consumer-orientated. Several important sets of regulations that served to protect public health were, nevertheless, introduced somewhat belatedly in the late 1980s. Ministers decided to introduce regulations sharply diminishing the amounts of BSE-infectious material entering human food chains; but they did so to protect their own reputations and to try to maintain market stability at the lowest possible cost to both the Exchequer and the meat industry. BSE policy-making was not a classic case of regulatory capture. MAFF did what it did because it was pursuing what it took to be its own interests (which it construed as the public interest), and what it thought was in the best interests of the domestic meat industry—but that was not how the meat industry always saw it. Ironically, it was the threat of unilateral action by the meat products industry in the spring of 1989 that forced ministers eventually to introduce a very important set of measures banning various ‘specified bovine offals’ (or SBOs) from the human food chain.

5

The introduction of regulatory restrictions During the first eighteen months of the BSE saga, from late 1986 to mid-1988, no regulations of any sort were introduced. Ministers and senior officials in MAFF had been primarily concerned about the impact that public knowledge of BSE would have on the export trade in beef and cattle. For nearly a year MAFF adopted a policy of trying to ensure that as few people outside the official veterinary community as possible learned about the existence of the disease. Consequently, no steps of any sort were taken during this period to diminish possible risks to veterinary or public health. Clinically diseased animals were routinely slaughtered and processed for human consumption alongside healthy cattle. The policy of concealment and denial was based not on scientific evidence or advice, but, to the contrary, on a decision to try to avoid consulting or involving scientific advisers. However, the number of reported cases of BSE was rising exponentially, despite the fact that most private veterinary surgeons had intentionally been kept ignorant of the disease. In February 1988, after media reports of the disease began to emerge, senior MAFF officials recommended that the minister should exclude cattle overtly and clinically afflicted with BSE from the food supply. They did this primarily because they thought the government would later be held responsible if it turned out that the disease could transmit to humans. Ministers resisted that advice because they were anxious that any regulatory action would undermine confidence in export markets. Ministers agreed to exclude overtly diseased animals from the food chain only after MAFF officials had insisted on consulting the government’s Chief Medical Officer (CMO), located in the DH. The CMO, in turn, insisted on the creation of an external expert advisory group—the Southwood Working Party (SWP)—not because he doubted that MAFF or the DH possessed requisite expertise, but because he judged that advice from eminent external experts would significantly contribute to changing ministers’ minds. The first step taken by the SWP (none of whose members was an expert on transmissible spongiform encephalopathies or TSEs) was to recommend that animals clinically affected with BSE should be excluded from the human food chain, and this advice ministers reluctantly chose to accept. The SWP did not provide scientific expertise

6

that was unavailable within the Civil Service; it provided external leverage from eminent chaps. Providing scientific legitimation for policy decisions The implicit rationale for establishing the SWP depended on an understanding that it would provide not just scientific advice but also specific policy recommendations. In order to do that, the committee members would have had either to make their own political judgements about the goals of policy and the acceptability of possible risks, or to endorse officials’ and ministers’ judgements on those matters. In practice, the SWP was subject to covert pressure concerning both the issues it addressed and the kind of scientific and policy considerations that it took into account. It was, for example, directly but discreetly told by the MAFF permanent secretary not to make any recommendations that would lead to an increase in public expenditure, and it was subsequently asked to avoid alarming the public by underplaying its concerns about risks of exposure to the BSE agent from pharmaceuticals and from occupational exposures. It was also expected to consider the impact of its deliberations on the economic interests of the meat trade; that was why Southwood and his colleagues were persuaded to drop a recommendation to ban the use of contaminated feed for pigs.5 The SWP’s recommendations were subsequently misrepresented, however (often by the same officials who had put pressure on the working party), as having been based purely on objective scientific considerations. This illustrates a more general pattern of scientists, and their expertise, being used as a political resource; thus policy-makers were able to make important policy decisions, prior to soliciting expert advice, and then obtain a spuriously scientific endorsement for those decisions. That misrepresentation contrived to delete several important considerations from the public representation of policy deliberations over risks. A consequence of the way in which that process was conducted was that it was harder for ministers and other policy actors to appreciate the possible risks and the extent of the uncertainties, and so they failed to appreciate the scope for, or the strength of the case for, exercising precaution. It also gave the misleading impression that decisions about the acceptability of risks were purely scientific, when in practice they were profoundly political.

7

An early draft of the SWP’s recommendations on BSE, for example, had suggested that human consumption of nervous system and lymphatic tissue from all cattle should be banned. That measure would have removed those tissues expected to contain the highest levels of infectivity from animals that were incubating BSE but not yet displaying clinical symptoms. That recommendation was, however, omitted from subsequent drafts; the text of the final report discounted the risks posed by subclinically infected animals, characterising them as negligible. Southwood later explained that he thought such a bovine offals ban would have been ‘a no-goer. They [MAFF] already thought our proposals were pretty revolutionary.’6 As a consequence, many senior policy actors, unaware of the tacit political judgement shaping the SWP’s advice and remarks, thought that there was no scientific justification whatsoever for imposing a ban on bovine offals. Key officials moreover, frequently made that faulty assumption, with serious consequences. Even though a ban on nervous and lymphatic cattle tissues had not been recommended by the SWP, several articulate critics from the scientific and medical profession began publicly challenging the government about the absence of such controls. In the spring of 1989 MAFF learned that the pet-food industry had decided a year earlier, independently of MAFF, to remove the high-risk bovine tissues from commercial petfood products, after having privately obtained advice from a consultant who was also one of MAFF’s expert scientific advisers. Agriculture ministers also learned that a major food industry trade association was considering a similar voluntary ban from its members’ products. The secretary of state, John MacGregor, decided that he could not allow MAFF to be upstaged by the industry, so in June 1989 he announced that an SBO ban would be introduced. It was one of the most important public health measures to diminish the risks from BSE; but the motivation for its introduction had little to do with public health, and a great deal to do with the political credibility of the minister and the desire for market stability. It was a consequence of contingent and incidental factors that the SBO ban, as well as the earlier slaughter policy, was introduced. Credit for the adoption of those policies belongs to the media, responsible officials, leading parts of the meat industry, a few expert advisers and the wider scientific and medical community—not to ministers or the government.

8

Regulatory rigidity The SWP concluded that it was ‘most unlikely that BSE will have any implications for human health’.7 That was, however, a provisional and partial judgement. It was based on the assumption that BSE had derived from scrapie in sheep, and on a further assumption, namely that BSE would behave in exactly the same way as the scrapie agent did, once it had jumped species into cattle. There was virtually no evidence either as to the origin of BSE (there still isn’t), or that the BSE agent, if it derived from a strain of scrapie, would behave just like the scrapie agent. On the contrary, experimental evidence indicated that the pathogenicity and host range of a particular TSE typically change once the agent has ‘passaged’ into a new host, and in ways that are hard to predict. The SWP had based its conclusions on a very optimistic scenario. The working party acknowledged at the time that its assessment was essentially ‘guesswork’, and later that it had been ‘mindful of the disastrous consequences of an alarmist report’.8 The conclusion stood in marked contrast to the comments made by the chair of the committee prior to its first meeting, when he told a journalist that ‘If the agent has crossed from one species to another there is no reason why it should not cross from cattle to man.’9 In 1989, however, the SWP’s optimistic scenario was seized upon by policy-makers and represented then, and on all subsequent occasions until March 1996, as final, conclusive and definitive. Ministers and senior officials repeatedly insisted, therefore, that the presence of BSE in British dairy and beef herds posed no threat to human health. Unfortunately, such claims were misleading. The evidence was never that reassuring in the first place, but it subsequently became progressively less reassuring as new information repeatedly emerged which undermined the hypothesis that BSE would behave just like scrapie. Even after the SBO ban had been introduced, infectivity was still entering the human food chain, because the ban did not cover all nervous and lymphatic tissues or certain other (commercially valuable) tissues that might have been expected to carry infectivity. The government’s scientific advisers made sure, however, that some key policy-makers were aware of those facts. Ministers and officials also insisted that they had been, and were, doing what, and only what, their expert scientific advisers recommended, and that it would have been

9

irresponsible to do anything else. In May 1990, for example, the agriculture minister John Gummer defended BSE policy in the House of Commons by asserting that the government had no alternative but to accept the advice of the experts.10 The impression conveyed was that BSE was purely a technical problem, and the UK government’s response to it was singularly and unproblematically rational. In those circumstances, dissident experts who publicly articulated their interpretations of what was actually an enormously uncertain and incomplete set of evidence were discounted, disparaged or ridiculed. Expert advisers had to be carefully managed by officials in order that the scientists did not undermine the government’s approach. On several occasions officials and ministers first decided what the policy should be and then tried to have their advisers construct advice, and representations of risk, that supported the line that had already been chosen, while subsequently claiming that advisers’ judgements and policy recommendations were based on, and only on, sound science. For example, the Spongiform Encephalopathy Advisory Committee (SEAC) was asked for a statement on the safety of beef in 1990. An early draft from SEAC drew attention to the fact that infectivity was still entering the food chain. MAFF officials simply deleted the ‘inflammatory’ paragraphs.11 The implications of the government’s narrative of adequate knowledge, robust science and zero risk extended well beyond what Phillips and his colleagues called ‘sedation’. The entire policy machinery in effect crystallised and rigidified around that narrative, demonstrating a remarkable lack of responsiveness to the remaining uncertainties or their diminution, and to the emergence of new knowledge and information. A crucial problem was that the commanding heights of the regime became so myopic and rigid that it was unable or unwilling to recognise that risks could be other than negligible. Having articulated a narrative to the effects that the science was robust and the risks were zero or negligible, MAFF found it very difficult to accept and respond to new evidence that implied that its assumptions about the risks (and the policy that flowed from those assumptions) might need to be revised. In May 1989, in a remarkably revealing document, a senior MAFF official from the Food Science Division discussed whether his branch should conduct a risk assessment

10

for bovine-derived human foodstuffs by drawing on analogies with the scrapie infectivity levels that had been found experimentally in various sheep tissues. The official explained that the assessment would probably estimate significant levels of infectivity in tissues such as heart, liver, kidney and muscle. He concluded that there would be no way of reducing such calculated exposure except by recommending further restrictions to ministers; he therefore advised that no such risk assessment should be conducted.12 The most important aspects of BSE policy were based not on ‘sound science’ but rather on decisions to avoid involving science, or to involve only ‘managed science’. In 1990, proposals from the European Commission to tighten the SBO controls by removing obvious nervous and lymphatic tissue from cuts of beef were resisted, but not because of the cost implications. Although ministers were willing seriously to consider tightening the regulations, MAFF officials insisted that no such action should be taken because ‘[a] statutory provision would . . . increase public concern as it would be assumed that such removal had not been the normal practice and could lead to pressure for the removal of all such material from meat, which would be impossible to implement as lymphatic material is throughout the carcass and is exposed wherever meat is cut’.13 Having previously failed to eradicate or substantially diminish possible risks, senior MAFF officials realised that introducing new regulations threatened to make explicit the fact that those risks were not, and had never been, negligible or certain. Many other, relatively low-cost steps could have been taken to learn about and diminish the risks to human consumers. Most of those (such as a ban on using so-called ‘mechanically recovered meat’) were avoided, partly to avoid damaging the competitiveness of the meat trade, but also to sustain the illusion of zero risk and complete control. From the point of view of almost all of our continental neighbours, one of the more problematic aspects of the BSE saga concerns decisions taken by senior officials in 1988 to try to mitigate the relatively minor economic costs to the renderers and the animal feed industry of a domestic ban on feeding ruminant protein to cattle by turning a blind eye to increased exports of contaminated feedstuffs. Those officials also persuaded scientific expert advisers not to mention the possible risks in public or in print, and deliberately chose not to warn importers that the material was contaminated. The strategy of concealment was reversed in 1989, but by then it had already ensured that

11

BSE would eventually become a global problem, and that the phrase ‘perfidious Albion’ would again be used in mainland Europe. A general pattern? A fundamental characteristic of the British approach to policy-making on BSE, therefore, was that while policy was represented to the general public as if it were based on, and only on, science evidence and the advice of the outstanding experts, in practice many of the important policy decisions were made before the scientists were recruited or consulted, and attempts were routinely made to obtain their endorsement of those prior decisions. The objectives of policy were concealed and/or misrepresented, risks were seriously and consistently understated, and the degree to which the available evidence and expert advisers supported those policies was misrepresented. Scientists were not completely manipulated, nor were they putty in the hands of the officials. Indeed, it was because the science was not, and could not be, entirely subordinated to political control that MAFF’s narrative of reassurance was eventually torpedoed in March 1996. Officials and ministers were, however, actively trying to ensure that scientific deliberations reached the conclusions they preferred, and they repeatedly tried to hide behind the advice of their external scientific experts, and to disparage any scientists who were critical of their reassuring narrative. MAFF officials responsible for BSE policy never showed a willingness to acknowledge either the extent of their ignorance or their tactics for coping with that ignorance. That pattern was by no means unique to BSE policy, or even to food safety policy. Selectively understating scientific (and other forms of) uncertainty is too often a conditioned reflex in governmental circles, partly because politicians do not want to reveal how poorly informed they often are, but also because they want to close off what might otherwise be seen as plausible and legitimate grounds for contesting policy. Policy implications If policy-making is to become democratically and scientifically legitimate, policy-makers will have to abandon the pretence that their decisions are based on, least of all only on, sound science. It will also be necessary to disclose, rather than conceal, scientific

12

uncertainties and differences of opinion among those with relevant expertise. When the uncertainties have been acknowledged, the political dimensions can no longer be concealed, and scientists should then not be expected to decide policies. Those imperatives are often interpreted to mean that scientists must be insulated from political influences, and that is how large parts of the scientific community interpret the lessons of the BSE saga. That is, however, an oversimplified interpretation because scientific understanding and expert advice never were, and never could have been, developed independently of their social, political and economic contexts. If scientists are to comment on what is known about the likely consequences of following, or failing to follow, a range of different possible courses of action, they need to know what the objectives of policy are, what courses of action are under consideration, and which could be implemented. The solution is, therefore, not to pretend that experts deliberate in a vacuum or that scientists are indifferent to any and all competing social interests or values, but rather that those considerations should be dealt with in an open and accountable fashion. Furthermore, the policy objectives which frame the scientific agenda are at least as important as the scientific evidence and its subsequent interpretation. If the ostensible and actual objective of policy is to protect public health, then it would be far more straightforward for scientific evidence and judgements to be developed in a sound, prudent and responsible fashion than for contrary considerations covertly to intrude. The BSE saga and crisis have provoked a wave of structural reforms in several jurisdictions. In the UK it led to the break-up, and subsequently the abolition, of MAFF. In April 2000 the Food Standards Agency (FSA) was established to take over MAFF’s former responsibilities for food safety and consumer protection policies. Separating industrial sponsorship from responsibility for regulatory policy-making was essential but not sufficient. A process of reform in the procedures for making and legitimating policy in the FSA is currently under way. The UK government’s response to the Phillips Inquiry stipulated that departments should not ask advisory committees to deal with issues beyond their competence, nor ask scientific committees to make risk management decisions. The FSA recently and boldly recommended that all its expert advisory committees should conduct their business in

13

open sessions. It stipulated that unorthodox and contrary scientific views should be considered, and that advisory committees should always provide a clear audit trail showing how and why they reached their decisions, where differences of opinions had arisen, and which assumptions and uncertainties were inherent in their conclusions.14 If those guidelines are followed in practice, then British policy-makers and their expert advisers are likely to specify and observe a far clearer separation of the scientific and political aspects of risk policy-making than hitherto. How that separation is to be achieved, delineated and policed is yet not entirely clear, nor is it at all clear how scientific and non-scientific considerations will ultimately be coupled together. Most of the shortcomings in BSE policy-making before April 2000, when the FSA was established, would be far less likely to recur were the FSA’s guidelines to be fully implemented. To the extent that some implementation deficits may occur, a risk remains that some of the key failures of BSE policy-making could be repeated. Institutional and procedural reforms in Europe The BSE crisis, and other food safety scares, have also had a considerable effect on the European Commission. In the late 1990s responsibility for providing scientific advice about food safety was moved away from the Directorate General for Industry and the Internal Market (DG-III) to what was then DG-XXIV, and is now called DG-SANCO or the Directorate General for Health and Consumer Protection. At that stage, however, regulatory authority remained with DG-III. In the spring of 2000, and in the wake of the crisis over dioxin contamination of animal feeds, the Commission endeavoured institutionally to separate regulation from sponsorship. Responsibility for consumer protection and food safety passed to DG-SANCO. In January 2000 the Commission published a White Paper on food safety. It proposed establishing a new integrated statutory framework covering the EU’s entire food chain ‘from the farm to the fork’ and the creation of a European Food Authority (EFA) to provide authoritative, independent, science-based advice to the Commission, and specifically to DG-SANCO. That body has subsequently been redefined as the European Food Safety Authority (EFSA) and was scheduled to begin work in the autumn of 2002.

14

However, the EFSA has, at the time of writing, not yet been established, and its modus operandi remains hard to predict. The wording of the Commission’s texts is ambiguous as between restricting the EFSA to providing strictly scientific advice on policy options and expecting it to provide specific prescriptive policy recommendations, or even decisions. That ambiguity was evident in the remark of a senior DG-SANCO official who said that SANCO wants the EFSA to be ‘independent but not out of control’. It is not obvious that those two characteristics can readily be reconciled, nor that the creation of the EFSA will overcome the limitations of the status quo ante. In recent years analogous developments have taken place in several other EU member states. The Irish government created the Food Safety Authority of Ireland, the French government established the Agence Française de la Sécurité Sanitaire des Aliments (AFSSA) and the German government has created the Bundesamt für Verbraucherschutz und Lebensmittel Sicherheit. At the global level, there is some evidence that a process of reform is under way at the Codex Alimentarius Commission, and the joint expert committees of the WHO and FAO. In those latter institutions, however, reform is proceeding at a glacial pace. Summary and conclusions The report of the Phillips Inquiry suggested that policy-making processes were fundamentally sound but failed to command the trust they genuinely deserved; it underdiagnosed the flaws in BSE policy-making. The FSA’s guidelines on the use of scientific advice suggest that the agency’s interpretation of the BSE saga is closer to ours than to Phillips’. If the FSA’s latest proposals were fully implemented it would be far more difficult than hitherto for policy-makers to hide behind their scientific advisers. Acting to promote the commercial interests of the food industry while pretending to protect public health would become scientifically and politically harder. Implementing the guidelines should help ensure that the roles of scientific advisers and policy-makers are more clearly and appropriately differentiated, and that each group is properly, but separately, accountable for its judgements and decisions. UK public policy on BSE (from 1986 to 1996, at least) provides a paradigm—and all too typical—case of risk policy-making that had little in the way of either scientific or

15

democratic legitimacy. Institutions are now being created and reformed, and they are struggling to articulate and implement a new paradigm. Several signs of progress have emerged, but it is too soon to tell what their impact will be. MAFF is no more, but neither the FSA nor DEFRA has yet fully clarified the relationship between the scientific aspects of appraisal and the political aspects of decision-making. The same is true for DGSANCO and the new agencies in Ireland, France and Germany. More fundamental reforms may therefore be anticipated before the scientific and democratic legitimacy of policy-making is enhanced. Notes 1 The supporting evidence is available at www.bse.org.uk (accessed 26 September 2002). 2 L. Phillips, J. Bridgeman and M. Ferguson-Smith, The BSE Inquiry, London, Stationery Office, 2000, vol. 1, Executive Summary, p. xviii and para. 1153. 3 Ibid., para. 108. 4 R. Bradley, BSE Research Projects, minute dated 19 July 1988. BSE Inquiry Year Book Reference Number 88\07.19\2.1-2.2, available at www.bse.org.uk (accessed 26 September 2002). 5 E. Millstone and P. van Zwanenberg, ‘Politics of Expert Advice: Lessons from the Early History of the BSE Saga’, Science and Public Policy, 2001, pp. 99–112. 6 F. Pearce, ‘Ministers Hostile to Advice on BSE’, New Scientist, 30 March 1996, p. 4. 7 Report of the Working Party on Bovine Spongiform Encephalopathy, London, Ministry of Agriculture, Fisheries and Food, 1989, p. 21. 8 R. Southwood, A. Epstein, W. Martin and L. Walton, Witness Statement No. 483A, 1999, para. 4, available at www.bse.org.uk (accessed 26 September 2002). 9 J. Erlichman, ‘Butchers Selling Diseased Meat’, Guardian, 29 June 1988, p. 4. 10 House of Commons Debates, 6th ser., vol. 173, col. 82. 11 Phillips et al., BSE Inquiry, vol. 11, para. 4.118. 12 Ibid., vol. 6, para. 3.202. 13 ibid., para. 4.450 14 UK Food Standards Agency, Report on the Review of Scientific Committees, 15 April 2002; see http://www.food.gov.uk/news/newsarchive/58746.

16

17