ArticlePDF Available

Politics of Expert Advice: Lessons from the Early History of the BSE Saga

  • Science Policy Research Unit University of Sussex
  • Centro de Investigaciones para la Transformación


This paper analyses the dynamics of the interactions between scientific and non-scientific considerations in providing scientific advice to policy, focusing on the first scientific committee to advise on BSE (bovine spongiform encephalopathy) policy-making in the UK and the political and social roles it was expected to play, and in practice played, in policy-making. The paper argues that the Committee was both deliberately and inadvertently utilised to provide spurious scientific legitimation for policy decisions which government officials believed ministers, other government departments, the meat industry and the general public might not otherwise accept. It demonstrates how those social roles rendered the spectrum of policy choices available on BSE opaque, allowed officials to undermine the democratic accountability of ministers, and contributed to making a very serious problem considerably worse. Some practical lessons are outlined for the organisation of scientific expertise in political affairs.
Science and Public Policy, volume 28, number 2, April 2001, pages 99–112, Beech Tree Publishing, 10 Watford Close, Guildford, Surrey GU1 2EP, England
Scientific advice
Politics of expert advice: lessons from the early
history of the BSE saga
Erik Millstone and Patrick van Zwanenberg
This paper analyses the dynamics of the interac-
tions between scientific and non-scientific con-
siderations in providing scientific advice to
policy, focusing on the first scientific committee
to advise on BSE (bovine spongiform encephal-
opathy) policy-making in the UK and the
political and social roles it was expected to play,
and in practice played, in policy-making. The
paper argues that the Committee was both
deliberately and inadvertently utilised to provide
spurious scientific legitimation for policy deci-
sions which government officials believed min-
isters, other government departments, the meat
industry and the general public might not
otherwise accept. It demonstrates how those
social roles rendered the spectrum of policy
choices available on BSE opaque, allowed
officials to undermine the democratic account-
ability of ministers, and contributed to making a
very serious problem considerably worse. Some
practical lessons are outlined for the organisa-
tion of scientific expertise in political affairs.
Dr Erik Millstone is at SPRU (Science and Technology Policy
Research), Mantell Building, University of Sussex, Brighton BN1
9RF, UK; Tel: +44 1273 877380; Fax: +44 1273 685865; E-mail: Dr Patrick van Zwanenberg is also
at SPRU; E-mail:
These authors have adopted the practice of alternating the
names on all collaborative publications and therefore no signifi-
cance should be attributed to the sequence of names.
IN MARCH 1996 the British Government an-
nounced that a new variant of Creutzfeldt–Jakob
Disease (CJD) was probably linked to exposure to
the cattle disease bovine spongiform encephalopathy
(BSE), thus unleashing the most damaging science-
based political crisis that has ever occurred in Britain.
What was so astonishing about events in the spring of
1996, and so damaging to the UK Government’s cred-
ibility, was that ministers and senior officials had in-
sisted, on innumerable previous occasions, that the
presence of BSE in British dairy and beef herds posed
no threat whatsoever to human health.
Ministers frequently argued that British beef was
“perfectly safe” (MAFF, 1990a; Gummer, 1990a),
that regulations went far beyond what was needed to
protect the public (Maclean, 1990; Meldrum, 1995)
and that those conclusions were fully supported by
scientific evidence and expertise. Indeed, central to
the Government’s depiction of BSE policy-making
was a strict reliance on advice from a series of inde-
pendent expert advisory committees. Ministers and
officials repeatedly insisted that BSE policy was “in-
formed and governed” by that advice (Howe, 1998,
paragraph 14), often implying that expert committees
played not only an influential, but a determinate, role
in decision-making (Gummer, 1990b).
The contrast between that reassuring narrative and
the statements of 20 March 1996 was startling.
Members of the Government’s Spongiform Encephal-
opathy Advisory Committee told the press that the
future number of variant CJD cases was “totally
unpredictable” and a “major cause for concern”
(MacIntyre et al, 1996). Furthermore, it is now clear
that the scientific advice received by ministers and se-
nior officials in the UK had indicated that the avail-
able evidence did not support the conclusion that BSE
Science and Public Policy April 2001 99
0302-3427/01/020099-14 US$08.00 Beech Tree Publishing 2001
was risk-free nor that British beef was without prob-
lems. Such disjunctions have thrust issues concerning
the role of scientific advice in policy-making to an un-
precedented level of scrutiny, both in the UK and
throughout the rest of Europe.
In this paper, we describe and analyse the early his-
tory of BSE policy-making in the UK to explore the
political and social roles that expert advisory commit-
tees can and do play in regulatory policy-making. Our
aim is to illuminate the dynamics of the interactions
between scientific and non-scientific considerations
in providing scientific advice to policy and to explore
the implications of those dynamics for robust regula-
tory policy-making.
We shall analyse empirically selected aspects of
the operation of the first expert advisory Committee
on BSE, known as the Southwood Working Party.
That Committee was established in the Spring of 1988
and ceased to operate following the publication of its
report and recommendations in February 1989.
Using the material publicly available as a result of
the Phillips Inquiry into BSE we shall identify some
of the multiple political and social roles that the Com-
mittee was expected to play, and in practice played, in
BSE policy-making. We shall argue that the expert
Committee was both deliberately and inadvertently
utilised to provide a spurious scientific legitimation
for policy decisions which Government officials be-
lieved ministers, other government departments, the
meat industry and the general public might not other-
wise accept.
We identify some of the structural and procedural
features of expert committees that help to account for
their playing those social roles and show how those
social roles rendered the spectrum of policy choices
available on BSE opaque, allowed officials to under-
mine the democratic accountability of ministers, and
contributed to making a very serious problem even
worse. We conclude by outlining some practical
lessons for the organisation of scientific expertise in
political affairs.
Relationships between science and policy
One of the key characteristics of the way in which
many policy-making bodies represent, and sometimes
organise, the institutional role of scientists in policy-
making is the assumption or assertion that scientific
considerations are not just a necessary basis for deci-
sion-making, but that they are by themselves suffi-
cient. Until October 2000, for example, British
ministers defended their decision not to compensate
people who had succumbed to variant CJD by claim-
ing that the Government could not be held responsible
for the BSE saga because it had always followed
scientific advice (Channel 4, 2000).
In effect, ministers were trying to respond to criti-
cisms of regulatory decision-making by representing
policy as involving purely scientific judgements.
Ministers and officials all too frequently have made
the converse assertion by claiming that there was no
scientific justification for policies that they did not
adopt or did not wish to adopt.
For over 30 years, however, most policy analysts
have recognised that there can never be a purely scien-
tific justification for any policy and that assertions to
the contrary seriously misrepresent real decision-
making processes (Gilpin, 1962; Weinberg, 1972). A
central criticism of that model of policy-making has
been that its plausibility invariably requires politics to
masquerade as science, with scientists either making
covert political judgements or having their judge-
ments politically misrepresented as sufficient and
Policy-relevant science is very rarely complete and
undisputed. Risk assessments are almost always
based on equivocal fragments of evidence, and their
conclusions are chronically uncertain. Decisions
about when to act, or when not to act, and about what
types of actions are appropriate, thus involve beliefs
about the severity of the putative risks which impos-
ing regulatory controls might enable us to avoid.
Furthermore, even if we could reliably know the
likelihood and seriousness of all possible types of
harm, those facts would never be sufficient to deter-
mine policy. Policy decisions require making
100 Science and Public Policy April 2001
Politics of expert advice
Erik Millstone is a senior lecturer at SPRU (Science and
Technology Policy Research), Sussex University. He has
been researching into the causes and consequences of
technological change in the food industry since 1974. Re-
cently his work has focused on the role of scientific evidence
and expertise in public policy-making in relation to topics
such as lead pollution, BSE and GM foods. His recent publi-
cations include articles in Zeitschrift für das gesamte
Lebenmittelrecht on EU food policy, Nature Medicine on
trust among science, scientists and institutions, Swiss Politi-
cal Science Review on food safety, and New Genetics and
Society on traumas of biotechnology, and Nature on GM
Patrick van Zwanenberg works at SPRU (Science and Tech-
nology Policy Research), Sussex University. His research
focuses on the role of science and expertise in public and
environmental health policy-making. He has been research-
ing into BSE policy for the last two years, but is increasingly
working on ways to operationalise the precautionary
principle and more generally on issues of science and
governance. His recent publications include articles on
expert risk assessment in Science, Technology & Human
Values, and BSE policy in Late Lessons from Early
Warnings: The Precautionary Principle in Policy
Policy-relevant science is very rarely
complete and undisputed: risk
assessments are almost always based
on equivocal fragments of evidence,
and their conclusions are chronically
complex judgements about the kinds of (uncertain).
risks, and levels of possible risk, which are accept-
able, in exchange for various presumed or anticipated
Even amongst that fraction of public officials
which recognises that science, by itself, is insufficient
to determine policy, it is widely assumed that the pro-
duction of scientific advice is, or should be, a politi-
cally and socially neutral process. Those protagonists
typically argue that it remains vital to insist on a strict
separation between ‘risk assessment’ (which is
deemed to be purely scientific) and the downstream
functions of ‘risk evaluation’ and ‘risk management’
(which may depend on non-scientific considerations).
as in Figure 1.
Evidence produced by scholars who have analysed
risk assessments that were supposedly purely scien-
tific indicates, however, that it is not possible to con-
struct a risk assessment from available scientific data
without embedding it with some prior, socially de-
rived framing assumptions (Jasanoff and Wynne,
1998). These concern, for example, assumptions
about the overall goals and parameters of policy, the
types of risk to appraise and the types to discount, the
types of disciplinary approach to bring to bear on a
problem, the types of model that are appropriate, and
the types of uncertainty to engage with, respond to,
and highlight (Robbins and Johnston, 1976; Wynne,
1992; Levidow et al, 1997; van Zwanenberg and Mill-
stone, 2000).
Drawing on these insights, a representation of the
structural features of a richer ‘contemporary’ model
of the relationships between scientific advice and pol-
icy-making than is commonly supposed by regulatory
policy-makers, is provided in Figure 2.
One of the fundamental premises of this model is
that all science-based risk assessments are inevitably
framed by some set of prior socially-based consider-
ations. In this model, scientific considerations are,
therefore, located within (or even perhaps sand-
wiched between) two sets of non-scientific consider-
ations. The upstream framing assumptions are
concerned, for example, with issues such as deciding
which changes are to count as harmful impacts and
which to discount, the choice of baseline against
which to evaluate such impacts, and the potential
routes of exposure to assess and those to disregard.
By contrast, downstream evaluations might be
concerned, for example, with the ways in which the
possible risks should be set against the extent and dis-
tribution of the social and/or commercial benefits of
accepting or diminishing those risks. They might also
Science and Public Policy April 2001 101
Politics of expert advice
Figure 1. Modern model of risk policy-making
Figure 2. Contemporary model of risk policy-making
include the civil liberties implications of imposing re-
strictions on the actions of individuals or corpora-
tions, or the relative suitability of different policy
instruments. While the graphical representation of
this model in Figure 2 has not been elaborated with ar-
rows representing feedback linkages, we do assume
that there are likely to be links between upstream
framing assumptions and downstream evaluative
One assumption underlying our approach is that, if
this model is a correct representation, existing institu-
tional structures systematically misrepresent the ways
in which decisions are actually reached. In particular,
upstream framing assumptions have typically not
been acknowledged; they are either left implicit, or
concealed within closed scientific discussions, and
surreptitiously embedded within expert judgements
and advice that have been misrepresented as if they
were purely scientific.
Our analysis draws on the ‘contemporary’ model to
describe and analyse how, in practice, scientific and
non-scientific considerations were coupled together
in BSE policy-making and the role that the South-
wood Working Party played in those processes. We
shall assume that these categorisations are not just
useful for post hoc analyses but also serve to indicate
the ways in which decision-making institutions and
processes should be structured.
Challenge of BSE
BSE was first recognised as a novel cattle disease
in November 1986 by scientists at the Central Vet-
erinary Laboratory (CVL) of the UK Ministry of
Agriculture, Fisheries and Food (MAFF). CVL pathol-
ogists discovered lesions in the brains of the afflicted
animals that were similar to a sheep disease called
scrapie that is endemic in the UK sheep population.
Scrapie belongs to a family of very poorly understood,
untreatable and invariably fatal brain diseases known
as transmissible spongiform encephalopathies (TSEs).
TSEs can affect both animals and humans but a TSE
had never previously been recorded in cattle.
TSEs are transmissible, both within a species and,
in some cases, between species. MAFF scientists sus-
pected that BSE had derived from sheep infected with
scrapie and was being transmitted through contami-
nated feed. Meat and bone meal derived from the
rendered remains of sheep, cattle and other animals
was routinely incorporated into animal feedstuffs.
Contaminated feed was quickly suspected and later
confirmed as the principal vector of the disease but
whether BSE had in fact derived from scrapie, or from
another source, was, and remains, unknown.1
The emergence of BSE posed, and continues to
pose, an enormously difficult challenge for public
policy because we have been profoundly ignorant
about so many aspects of the disease. Most crucially,
in the late 1980s, we did not know whether BSE was
pathogenic to humans. There was no evidence,
despite many years of epidemiological research, that
eating sheep meat from scrapie-infected animals
could cause Creutzfeldt–Jakob Disease (CJD) — the
most common human TSE.
Unfortunately, however, each TSE was thought to
possess a distinct host range and policy-makers could
not be sure that the agent that caused BSE had in fact
derived from scrapie. Even if the scrapie pathogen had
jumped species into cattle, the available evidence in-
dicated that one could not be sure that the disease
would subsequently have the same transmission
characteristics as scrapie once present in another host
Even if policy-makers were to have assumed that
BSE was pathogenic to humans, uncertainties about
how to manage that risk remained profound. The
pathogenic agent, believed by many to be an abnormal
and virtually indestructible type of protein known as a
prion, gives rise to a disease with a very long incu-
bation period which can be transmitted well before
clinical symptoms appear. There was, however, no
ante-mortem diagnostic test that could detect the path-
ogen in animals before clinical symptoms appeared
and so asymptomatic cattle could not be identified nor
differentiated from cattle which were uninfected.
It was known, from studies on scrapie, that the
pathogen is found in its most concentrated form in
brain and other nervous tissues but it is not necessarily
confined to those tissues. When cattle are incubating
the pathogen we do not know whether or not some tis-
sues are pathogen free. It may be that the pathogen is
confined exclusively to some tissues, or it may be that
all and any tissue will contain some of the pathogenic
agent, though at varying concentrations. Even if tis-
sues were assumed to carry the pathogenic agent, no-
body knew what the levels of infectivity in the various
tissues would be, or how this could vary over the pe-
riod of incubation and neither did anyone know
whether there might be a threshold of human exposure
below which the risk would be negligible.
In 1987, as cases of BSE rapidly escalated through-
out the UK, and for many years thereafter, it was not
possible reliably to predict the effects of eating British
cattle and cattle products. Scientific knowledge was
fragmentary and inconclusive and, in the short term,
there were few experimental means of substantially
reducing those uncertainties. Policy-makers had no
choice but to take urgent decisions about a novel dis-
ease the implications of which were unknown.
Initial policy response
Policy responsibility for BSE rested primarily with
MAFF, a department that was expected simulta-
neously to promote the economic interests of farmers
and the food industry whilst also protecting public
health from food-borne hazards. During the first 15
months of the BSE epidemic, the predominant con-
cern for ministers and senior officials in MAFF,
despite recognising the potential hazards from BSE
102 Science and Public Policy April 2001
Politics of expert advice
(Phillips et al, 2000, volume 3, paragraph 1.37) was
that any knowledge about the emerging disease in the
UK herd might provoke consumers and overseas trad-
ers to stop purchasing, consuming and importing Brit-
ish cattle, beef and beef products.
To avoid publicity about the new disease, MAFF
policy-makers therefore took no regulatory measures
whatsoever,3and attempted to keep all knowledge
about the new disease within the Ministry and a small
part of the official veterinary community. Not even
the Department of Health was informed about BSE.
Establishing an expert Committee on BSE
By early 1988, however, over a year after the first
cases had been diagnosed, reported cases of BSE were
increasing rapidly and knowledge of the epidemic
was beginning to diffuse into the public domain. Dif-
ficult questions were being raised both by the media
and in Parliament about the risks of consuming dis-
eased cattle, which at the time were entering the hu-
man food chain. MAFF officials became concerned
that unless ministers introduced a policy to slaughter
and destroy clinically affected animals they would be
held responsible if it later transpired that BSE was
transmissible to humans (Phillips et al, 2000, volume
3, paragraph 5.41).
Senior MAFF officials therefore recommended in
February 1988 that the Agricultural Minister, John
MacGregor, authorise the introduction of a slaughter
and compensation policy. MacGregor refused to ac-
cept that advice, primarily because such action would
set a precedent that was inconsistent with the Govern-
ment’s policy of insisting that industry pay the costs
of eradicating animal and plant diseases, and also be-
cause of a lack of scientific evidence of a risk (Phillips
et al, 2000, volume 3, paragraphs 5.58, 5.186 and
Officials had warned MacGregor that the only
credible option available to him if he rejected official
advice was to obtain the support of the Department of
Health’s (DoH) Chief Medical Officer (CMO).
MAFF officials therefore informed the CMO, Donald
Acheson, about BSE, some 17 months after MAFF
was first alerted.
Acheson has claimed that he was told by MAFF of-
ficials that a formal risk appraisal was a prerequisite
for the introduction of a slaughter and compensation
policy in order to justify the costs to the Treasury, al-
though that version of events was denied by MAFF of-
ficials (Phillips et al, 2000, volume 3, paragraph
5.209). The Phillips Inquiry has concluded that
MAFF officials’ testimony was accurate and that
Acheson must have been mistaken (Phillips et al,
2000, volume 3, paragraph 5.213). Nevertheless the
contemporaneous minutes of a meeting between
Acheson and MAFF officials do note that: “Mr
Cruickshank [a senior MAFF official] thought it nec-
essary to assess the risk in humans in order to justify
the cost of any control measures taken by MAFF”.
Whether MAFF officials did in fact argue that such
appraisal needed to be conducted in advance of any
control measures is disputed, but it is clear that minis-
ters had already rejected official advice in part be-
cause they wanted more positive evidence of a risk to
humans before the expense of Government compen-
sation could be recommended (Phillips et al, 2000,
volume 3, paragraph 5.193). The immediate, and
explicitly political, problem was thus one of how to
respond to the almost total absence of scientific
evidence given that ministers wanted an evidential
rationale before they would concede that action
needed to be taken.
Our interpretation of events is that, since evidential
support could not be provided, Acheson and other of-
ficials chose to seek ‘eminential’ support for the pol-
icy that ministers had thus far rejected (compare with
Isaacs and Fitzgerald, 1999). Acheson himself might
have provided that support but as he later recalled: “I
regard it as highly unlikely that my personal advice,
unsupported by experts, would have been accepted by
the Treasury” (Acheson, 1999, paragraphs 27–28).
Acheson and other DoH and MAFF officials then
appointed an expert Committee of eminent scientists
to provide urgent advice specifically on bovine-
derived pharmaceutical products and the disposal of
affected animals. Acheson told colleagues that he
thought it highly probable that the advice would be
that carcasses of affected animals should not go for
human consumption (Phillips et al, 2000, volume 3,
paragraph 5.216).
Just to be sure, Acheson wrote to the chair of the
new Committee suggesting that “it may be desirable
for your group to have an early meeting to decide
whether any urgent action is prudent” (BSE Inquiry,
1999a, paragraph 42). When the Committee did in-
deed quickly recommend taking clinically affected
carcasses out of the human food chain, Acheson wrote
to the Health Minister arguing that “… such action is
indeed essential” (Phillips et al, 2000, volume 3, para-
graph 5.109)
We suggest that the Committee was established
primarily to provide officials in MAFF and the De-
partment of Health with a political resource with
which to push for the introduction of regulations
which they believed agricultural ministers and the
Treasury would not otherwise accept. The fact that
none of the Committee members was an expert on
Science and Public Policy April 2001 103
Politics of expert advice
We suggest that the committee was
established primarily to provide
MAFF and DoH officials with a
political resource with which to push
for regulations which they believed
agricultural ministers and the
Treasury would not otherwise accept
TSEs was thus relatively unimportant to that objec-
tive. The one required quality was eminence, which
the members possessed in abundance.
The Chair of the Committee was Richard
Southwood, a Professor of Zoology at Oxford who
had previously chaired both the Royal Commission
on Environmental Pollution and the National Radio-
logical Protection Board; his name was suggested by
Acheson who knew Southwood personally (BSE In-
quiry, 1999a, paragraph 18). The other members
were: Anthony Epstein, emeritus Professor of Pathol-
ogy at Bristol University and joint discoverer of the
Epstein Barr virus; Professor Bill Martin, a veterinar-
ian, and former director of the Moredun Research In-
stitute in Edinburgh; and Sir John Walton, former
Professor of Neurology at Newcastle upon Tyne.
MAFF and DoH officials had suggested that Martin
and Walton join the Committee, whilst Epstein was
appointed at Southwood’s suggestion. All four mem-
bers had served previously on numerous government
committees (Southwood et al, 1999a, paragraph 1).
The ‘Southwood Working Party’, in common with
most UK advisory committees, had no staff, no bud-
get; the members were unpaid and they were expected
to carry on with their existing professional duties. It
was a joint committee of MAFF and the DoH and both
departments contributed to the committee’s secretar-
iat. The Working Party was particularly dependent on
its secretariat and other officials from MAFF and the
DoH. Indeed officials drew up a list of questions that
they thought the Committee should address in provid-
ing advice to ministers, drafted the answers to those
questions, supplied most of the data and evidence and
drafted much of the Committee’s final report.
Asymmetries of knowledge and resources are typi-
cal features of the relationship between government
departments and small ad hoc advisory committees
but the structural characteristics of UK committees
mean that such asymmetries are especially
Expected role of Southwood Working Party
The Working Party was formally appointed to “advise
on the implications of Bovine Spongiform
Encephalopathy and matters relating thereto”
(MAFF/DoH, 1989). That is particularly ambiguous
as between giving scientific or policy advice, or both,
but of course the implicit rationale for establishing the
Committee depended on an understanding that the
members would provide not just scientific advice but
also explicit policy advice with specific policy recom-
mendations; a role that the members were content to
The danger for MAFF was that, once appointed, an
expert committee with a remit to provide policy advice
might erode the Ministry’s autonomy over control of
BSE policy. As one MAFF official later acknowl-
edged, speaking generally about expert committees
(BSE Inquiry Transcript, 29 June 1988, page 79):
“… you have to turn to external bodies to try to
give some credibility to public pronouncements,
[but] you are very dependent therefore on what
the Committees then find. … Really the key to it
is setting up the Committee, who is on it, and the
nature of their investigations.”
When negotiations were underway to set up the
Working Party, the preference of senior MAFF offi-
cials had been for Southwood to provide advice alone,
on the basis of help from civil servants, and without
any terms of reference (BSE Inquiry, 1999a, para-
graph 26). However, both Acheson and Southwood
insisted that external experts join the Committee and
that the Committee should be formally appointed
(BSE Inquiry, 1999a, paragraph 48).
Over the lifetime of the Working Party, MAFF of-
ficials consistently attempted, discreetly, to exert as
much control as possible over the sensitive political
functions of risk evaluation and risk management.
Even before the Working Party began meeting,
Southwood has recalled that MAFF’s Permanent Sec-
retary “… expressed the hope to me that any recom-
mendations we would make ‘would not lead to an
increase in public expenditure’” (Southwood, 1998,
paragraph 25). In practice the Working Party some-
times sought to retain autonomy for policy decisions.
We describe, in the next section, some of the tensions
between the Government and the Working Party.
Although the Committee was willing to take an
overt role advising on risk evaluation and manage-
ment, it appeared reluctant, initially, to contribute to,
or to endorse, the Ministry of Agriculture’s risk com-
munication strategy. That strategy was concerned pri-
marily with trying not to frighten the British public or
overseas markets and governments, but Southwood
did not always support it. For example, at about the
time of the Working Party’s first meeting Southwood
was explicit about the absence of evidence regarding
BSE. He told journalists that: “[i]f the agent has
crossed from one species to another there is no reason
why it should not cross from cattle to man”
(Erlichman, 1988)
Soon after the Working Party began meeting, how-
ever, it started, in effect, to acquiesce with the Gov-
ernment’s risk communication priorities. For
example, the first draft of the Committee’s report,
written a few weeks after the comment to journalists,
described the risks of transmission to humans as “re-
mote”, but that phrase had been provided by civil ser-
vants (BSE Inquiry Transcript, 21 July 1999, pages
27–28). The Committee members later recalled that
they were “… mindful of the disastrous consequences
of an alarmist report...” (Southwood et al, 1999b,
paragraph 4) and that the purpose of their report was,
in part, to persuade the public not to react
By “disastrous” the Committee members ex-
plained that they meant, in part, the possible economic
consequences for farmers and those involved in the
livestock industry (BSE Inquiry Transcript, 21 July
104 Science and Public Policy April 2001
Politics of expert advice
1999, page 7), by which they presumably meant the
implications that might arise if members of the public
stopped purchasing British beef. We describe in sub-
sequent sections some of the ways in which that up-
stream framing assumption shaped the assessment
and representation of risks.
Role actually played by Working Party
In this section we describe three areas of the
Southwood Working Party’s deliberations that illus-
trate some of the tensions concerning both the multi-
ple tasks of risk assessment, risk management and
risk communication and the relationships between
officials and the Working Party.
Controls on the animal feed chain
Immediately prior to the establishment of the
Southwood Working Party, MAFF banned the use of
potentially contaminated ruminant protein (that is,
slaughterhouse waste from cattle and sheep) in the
feed of other ruminants. Non-ruminants, such as pigs
and poultry, could still be fed with the contaminated
protein even though nobody knew whether or not they
might also be susceptible to BSE. MAFF officials had
in fact considered, and then rejected, a ban on feeding
ruminant protein meal to all animals because that
would have deprived the rendering industry of its
principal market (the bulk of animal protein was fed to
pigs and poultry)4(BSE Inquiry Transcript, 29 June
1998, page 35). Senior veterinary officials were nev-
ertheless aware that their decision was a gamble. In
June 1988, the Chief Veterinary Officer, Keith
Meldrum, admitted privately to a colleague that:
“[t]he most we could say is that any ruminant
protein fed to [pigs] might contain the agent of
BSE or scrapie. Whether or not infection would
be established in the pig and whether it might
replicate is unknown.” (BSE Inquiry Transcript,
16 June 1998, page 99).
The Southwood Working Party was concerned that
feeding contaminated material to pigs and other non-
ruminants might be hazardous and the draft con-
clusions to the report in effect contained a recom-
mendation that the use of ruminant meat and bone
meal should be banned from the feed of all herbivores
(Phillips et al, 2000, volume 4, paragraph 9.21;
Cruickshank, 1998, paragraph 7.2). When ministers
and senior MAFF officials first saw the draft conclu-
sions it “caused great anxiety” (Cruickshank, 1998,
paragraph 7.1). As the junior Agriculture Minister
Donald Thompson commented:
“I cannot say how strongly I regard this matter.
Of course we must take all due care but the envi-
ronmental, economic and competition conse-
quences would be dire if Prof S was to go
forward” (Phillips et al, 2000, volume 4,
paragraph 9.25).
Officials decided that the Chief Veterinary Officer,
Keith Meldrum, should attend the final Working
Party meeting and explain MAFF’s perspective.
Southwood refused to allow Meldrum to participate
in the Working Party’s last meeting but prior to that
meeting the attention of two members of the Commit-
tee was drawn to the practical problems their proposal
might provoke (Phillips et al, 2000, volume 4, para-
graph 9.28). In January 1989 one of those members,
Bill Martin, wrote to Sir Richard Southwood arguing
that “I think we have to be restrained in the view we ex-
press in the report, on this subject” (Phillips et al, 2000,
volume 4, paragraph 9.26).
Southwood and his colleagues responded by drop-
ping their recommendation.5Crucially, however, the
final report contained almost no discussion about the
possible risks to pigs. It would have been difficult
presentationally to claim, on the one hand, that the
Working Party was concerned about the risks to pigs,
but on the other hand, to in effect claim that there was
no need for regulatory restrictions on the contents of
pig feed, at least not without making explicit the fact
that the Committee was not making purely scientific
Following the publication of the Southwood re-
port, the absence of a recommendation to extend the
ruminant feed ban to pigs provoked enormous criti-
cism from, amongst others, prion disease specialists
and medical scientists, the Welsh Chief Medical Offi-
cer, and the National Farmers Union (Ferriman, 1989;
Phillips et al, 2000, volume 5, paras 3.161 and 3.168).
The animal feed industry too was unhappy with the
practice of including at least some ruminant tissues in
animal feedstuffs, and at least one of the members of
the Southwood Working Party was unhappy to learn
that infected ruminant meal was still being fed to Brit-
ish pigs (Phillips et al, 2000, volume 4, paragraph
9.34). Even the Agriculture Minister asked officials if
it might not be appropriate to extend the feed ban to
cover pigs as a precautionary measure (Phillips et al,
2000, volume 5, paragraphs 3.118).
MAFF officials, isolated in their determination to
maintain what they perceived to be the interests of the
rendering industry, responded to those criticisms by
misrepresenting the absence from the Committee’s
Science and Public Policy April 2001 105
Politics of expert advice
The Southwood Working Party’s
conclusions about extending the feed
ban were subordinated to the prior
policy judgements of MAFF officials:
officials then represented the
Committee’s conclusions as purely
scientific in nature
report of any discussion of pigs as ‘scientific’ proof
that the Committee was confident that pigs posed no
problem. A MAFF press release, dated 9 January
1990, noted for example that:
“There is no scientific justification to extend the
ruminant feed ban to pigs and poultry. The
Southwood Report acknowledged the impor-
tance of the feed ban for ruminants, but did not
recommend that it be extended to pigs and poul-
try.” [emphasis added] (MAFF, 1990b)
Officials made a similar private response to their
ministers and to the animal feed industry (Phillips et
al, 2000, volume 5, paragraphs 3.102 and 3.125).
To summarise, the Southwood Working Party’s
conclusions about extending the feed ban were subor-
dinated to the prior policy judgements of MAFF offi-
cials. Officials then represented the Committee’s
conclusions as purely scientific in nature, drawing on
the spurious scientific veneer of authority provided by
the Working Party to defend officials’ policy deci-
sions in the face of criticism from the public, the
animal feed industry and even their own ministers;
a tactic that the members of the Working Party did
not overtly challenge.
An unfortunate consequence of the decision
not to ban infected ruminant protein in pig and poul-
try feed was that for the next six or so years cross-
contamination occurred between feed destined for
cattle and feed destined for other animals, greatly
prolonging the BSE epidemic (Phillips et al, 2000,
volume 1, paragraphs 376 and 377). Furthermore, we
still do not know whether or not BSE is pathogenic to
pigs and other farm animals (Hill et al, 2000).
Controls on the human food chain
The Southwood Working Party had been quick to
insist that clinically affected animals be removed
from the human food chain but it did not make any
recommendations as regards restrictions on the
consumption of asymptomatic cattle (that is, cattle
infected with BSE but not yet exhibiting clinical
symptoms). Asymptomatic cattle were far greater
in number than clinically affected animals and were
potentially just as hazardous, but as there was no di-
agnostic test with which to identify them, any re-
strictions, such as the removal of potentially highly
infectious central nervous system and lymphatic
tissues, would have to be imposed on the entire cat-
tle herd.
The members of Southwood Working Party have
recalled that they were indeed very worried about al-
lowing consumption of at least some types of bovine
offal (BSE Inquiry Transcript, 11 March 1998, page
142) and in fact the first draft of their report had
suggested that “[b]rain and lymphoid tissue could be
omitted from meat pies without any important conse-
quences for the industry and certainly without any cu-
linary loss” (Phillips et al, 2000, volume 4, paragraph
10.53). The Committee subsequently decided,
however, to drop any such recommendation.
The rational for that decision was not recorded in
contemporaneous minutes, but in written evidence to
the Phillips Inquiry the members of the Working Party
have claimed that they did not consider the costs and
practicalities of banning bovine brain tissue to be pro-
portionate to their beliefs about the possible risks
(Southwood Working Party, 1999, paragraph 28).
Elsewhere, however, Southwood has suggested that
their decision was influenced not by their own inde-
pendently formed judgements but by their assumption
about what might and might not be politically
In 1996, Southwood explained: “We felt [a ban on
bovine brain] was a no-goer. They [MAFF] already
thought our proposals were pretty revolutionary”
(Pearce, 1996). In other words, direct pressure from
officials was not required, in this case, to influence the
Committee’s eventual recommendations — mere
anxiety seems to have been sufficient.
Once again the Working Party did not make ex-
plicit the political judgements that it had made when
deciding not to recommend an offal ban but instead
chose to provide the impression, in public, that the
risks were negligible, as if the decision not to recom-
mend restrictions was purely scientific. They stated,
for example, that the risks from asymptomatic cattle
would not even justify labelling of products contain-
ing central nervous system tissue (MAFF/DoH, 1989,
paragraph 5.3.5).
As with the animal feed issue, the absence of an
offal ban attracted much criticism following the
publication of the Working Party’s report. Ministers
recognised that some infected animals might be enter-
ing the food chain before they developed clinical
symptoms and one of MAFF’s junior ministers asked
his officials to look into the prospect of a ban on con-
sumption of older cows (Phillips et al, 2000, volume
6, paragraph 3.149). MAFF’s own veterinary scien-
tists clearly recognised that there was a rationale for
further regulatory restrictions. MAFF’s head of veter-
inary pathology told the Chief Veterinary Officer that:
“It is quite clear that in an infected animal in the
late pre-clinical phase (ie that which concerns
the Parliamentary Secretary) that lympho-
reticular tissues [and] certain other tissues
including some endocrines and gut plus CNS
[central nervous system] are highly likely to be
infected to a high titre of significance to humans
if the agent is transmissible to them.” (Phillips et
al, 2000, volume 6, paragraph 3.136)
That advice did not, however, appear to reach minis-
ters who, instead, were told by senior MAFF officials
that: “… there would not appear to be any reason to
take any further action at this stage, particularly since
Southwood did not recommend it. In fact he
[Southwood] concluded that even the labelling of
products would not be justified in the present state of
106 Science and Public Policy April 2001
Politics of expert advice
knowledge” (BSE Inquiry, 1999d, paragraph 83).
Thus once again officials represented (although per-
haps not deliberately) the Working Party’s risk man-
agement decisions as purely scientific as a means of
justifying their policy commitments in response to ex-
ternal and internal pressures.
Later in the spring of 1989, agriculture ministers
learnt that the major pet food manufacturers had ex-
cluded certain cattle offal from their products and that
a major food industry trade association was consider-
ing a voluntary offal ban (Phillips et al, 2000, volume
6, paragraph 3.203; BSE Inquiry, 1999d, paragraph
101). The Minister, not wishing to be upstaged by the
industry, and for other reasons too, announced a simi-
lar ban for the human food chain,6although it was not
introduced for another nine months. With hindsight,
the offal ban was a critical element of public health
protection against BSE. It was, however, introduced
for political rather than public health reasons.
The delay of over a year, from the time restrictions
on offal consumption first appeared as a suggestion in
the draft of the Working Party’s report, to the actual
introduction of the ban, meant that the offal from an
estimated 25,000 infected cattle, that were at least
halfway through the average incubation period for
BSE, had entered the human food chain (Dealler,
1996). Furthermore, the Working Party’s portrayal of
the risks from asymptomatic animals inadvertently
undermined the rational for subsequently introducing
the regulation, because few actors believed that the of-
fal ban possessed any scientific legitimacy. The Agri-
cultural Minister, John MacGregor, recalled, for
example, that: “some people were concerned about
sub-clinical animals … others were not, because Sir
Richard Southwood had made clear that was not a real
problem” (Phillips et al, 2000, volume 6, paragraph
Instead the offal ban was widely believed to have
been introduced to reassure an irrational and ill-
informed public (Phillips et al, 2000, volume 6, para-
graphs 3.313-3.317). Not surprisingly, the controls
were poorly enforced (Phillips et al, 2000, volume 1,
paragraph 1186).
Risk communication and non-food borne exposure
This section focuses on examples of the tensions be-
tween the provision of scientific advice and the objec-
tives of risk communication. As noted previously, the
members of the Working Party were anxious that their
report should not be ‘alarmist’. One of the ways in
which that upstream framing assumption shaped risk
assessments was in relation to discussions of uncer-
tainty. In general, the Working Party has recalled that:
“[w]e were concerned to identify any real and
material risks to human health about which
something could be done. Whilst our Report
drew attention to the uncertainties that existed,
we did not and do not consider that the highlight-
ing of numerous ‘potential’ risks was then or is
now desirable especially when such potential
risks are based on speculation rather than scien-
tific knowledge.” [emphases added] (South-
wood Working Party, 1999, paragraph 9)
More specifically, for example, the Committee was
very reassuring, in print, about the risks from asymp-
tomatic animals. Another scientist wrote to
Southwood criticising his Committee for under-
emphasising those risks (Phillips et al, 2000, volume
4, para 10.65). Southwood replied by arguing that:
“… it was extremely difficult to steer the proper
course between causing excessive alarm and undue
complacency” (Phillips et al, 2000, volume 4, para
10.66) and that given the likelihood that BSE would
behave like scrapie his Committee had decided not to
press the point about sub-clinically infected animals.
Although the Working Party’s report sometimes
attached caveats to its more reassuring statements, the
fact that the evidence was so fragile and indirect was
not always made explicit. Yet privately, the Commit-
tee acknowledged that their scientific assessment was
essentially “guesswork” (Phillips et al, 2000, volume
4, paragraph 10.33). Several ministers and officials
have claimed that they were largely unaware of the
fragility of Southwood’s conclusions (Phillips et al,
2000, volume 4, paragraph 11.6).
Another more specific manifestation of the inten-
tion not be alarmist was in relation to non-food
sources of exposure to the BSE pathogen. Southwood
has recalled that: “[w]e really thought the medical
problem was severe” (Phillips et al, 2000, volume 4,
paragraph 10.83) because of the use of bovine materi-
als in the manufacture of pharmaceutical products
such as vaccines. The Committee was also concerned
about occupational routes of exposure (Phillips et al,
2000, volume 4, paragraph 6.11).
Early drafts of the Working Party’s report had
drawn attention to those concerns but the members
were persuaded by the Department of Health secretar-
iat to modify the relevant passage of the Committee’s
report on the grounds that public confidence in the
vaccination programme might have been put in jeop-
ardy (and because the authorities had already pri-
vately been alerted to the possible risks). Indeed, the
Working Party adopted verbatim the DoH officials
suggested wording for the relevant section of the
Science and Public Policy April 2001 107
Politics of expert advice
Working Party members were
persuaded by the DoH secretariat to
modify the passage about use of
bovine materials in pharmaceutical
products on the grounds that public
confidence in the vaccination
programme might have been put in
report (Phillips et al, 2000, volume 7, paragraph 5.21).
The Working Party’s final report stated that the risks
of transmission via medicinal products and occupa-
tional routes of exposure appeared “remote”
(MAFF/DoH, 1989, paragraph 5.3.5) but that term
was absent in earlier drafts of the report (BSE Inquiry,
1999e, paragraph 117).
As far as occupational risks were concerned,
Southwood had told officials in November 1988 that
current Health and Safety Executive regulations were
inadequate but he had been told by the Chief Veteri-
nary Officer, Keith Meldrum that “… new precau-
tions would escalate the issue unnecessarily when we
were saying there was no hazard to man from BSE …”
(Phillips et al, 2000, volume 4, paragraph 6.11).
Southwood agreed that a general warning from the
Health and Safety Executive about handling animals
would perhaps suffice. When asked by the Phillips In-
quiry why the Working Party had described occupa-
tional risks as remote, Southwood replied that “…
causing alarm — which may be totally unnecessary
— is not something a responsible body does” (BSE
Inquiry Transcript, 21 July 1999, page 96). Another
member of the Committee, Bill Martin, explained
“One could see slaughtermen downing tools and
going on strike because of the possible risks of
handling cattle. There were all sorts of possibili-
ties there. I could not see veterinarians doing
that, but they would take precautions by wearing
long plastic sleeves and so on in any calving they
were doing. But slaughtermen and abattoir
workers were a different category. They could
down tools and refuse to handle cattle.” (BSE
Inquiry Transcript, 21 July 1999, page 97)
Although the Working Party had deliberately misled
the public in describing their beliefs about non-food
borne risks from BSE, the members of the Committee
defended that decision by claiming that they had pri-
vately alerted the Medicines Control Agency and the
Health and Safety Executive to their real beliefs about
the potential risks of transmission and that the term re-
mote was accurate in the sense that precautions had al-
ready been, or were about to be, taken ( Phillips et al,
2000, volume 4, paragraph 10.87). That ultimately
turned out to be naïve because those and other agen-
cies were persuaded by the Working Party’s published
advice that the risks were already remote (Phillips et
al, 2000, volume 4, paragraphs 10.83–10.109). Ex-
isting stocks of vaccines, for example, continued to be
used until 1993 and it took the Health and Safety Ex-
ecutive three years to issue advice to high risk trades
(Phillips et al, 2000, volume 1, page xxvii).
Given the long-standing practice of recycling slaugh-
terhouse waste in animal feed, it is hard to see how all
risks from BSE could have been avoided once the dis-
ease spread amongst British cattle. Human exposure
to the BSE pathogen undoubtedly occurred well be-
fore the British animal disease surveillance system
first discovered BSE. There were, however, very
many ways in which the UK Government could have
responded to the emergence of the disease that would
have provided a far greater degree of protection to
public health. Several of the reasons why this did not
happen concern the ways in which scientific advice
was, and to some extent still is, organised in the UK.
A recent official review of British expert commit-
tees involved with food safety acknowledged that
“[h]istorically, it has not always been clear where the
responsibilities of the advisory committees end and
those of the Government begin” (DTI, 2000, page 4).
In this paper, we have illustrated some of the conse-
quences of that lack of clarity and the complexities of
the interactions between science and policy.
First, we have argued that the initial purpose of set-
ting up the advisory committee was to provide politi-
cal as opposed to scientific support to officials who
wanted to persuade ministers to sanction the removal
of clinically infected animals from the food chain.
This provides a striking example of scientists, and
their expertise, being used as a political resource, and
partly acquiescing in that use. We have shown, more-
over, that the advisory committee was only able to
perform that function because its deliberations, much
of the evidence available to it, and its interactions with
civil servants and ministers, were not available for
public scrutiny.
Secondly, as implied by the ‘contemporary’ model
of science in policy, as given in Figure 2, we have ex-
plained how in practice the Southwood Working
Party was sandwiched very firmly between a set of up-
stream framing assumptions concerning the objec-
tives of public policy (that is, not to scare the public
and to spend as little public money as possible) and
downstream judgements about the acceptability of
uncertain risks in exchange for the economic and
political benefits of sustaining (at least temporarily)
the UK meat and dairy trade.
Yet we have shown that there were significant dif-
ferences between the role that the Southwood
Working Party ostensibly played, namely a purely
scientific one, and the political roles that it was in-
tended to play and which, in practice, it eventually
played. We have shown how technical analyses and
representations of risks were subordinated, directly
and indirectly, to the prior decisions and commit-
ments of risk managers and how expert scientific ad-
visors were willing to acquiesce to an arrangement
under which the scientists provided political advice
that was misrepresented to the general public, and
other actors, including ministers, as if it were purely
We have also shown that the members of the Com-
mittee and the relevant government departments
failed explicitly to acknowledge the policy contexts
and their complexities, within which appraisal was
108 Science and Public Policy April 2001
Politics of expert advice
conducted. We have provided examples of how politi-
cal framing commitments shaped the production of
scientific advice even though the subsequent reports
of the committees were presented in ways which con-
cealed those contexts and which suggested that the
risk assessments were constructed in isolation and
abstraction from their context.
Those arrangements suited officials and ministers
because it allowed them to argue that they were doing
what, and only what, their scientific advisors recom-
mended, and it allowed officials to use the ostensible
scientific authority of the Committee to persuade the
public, the Treasury, other government departments,
ministers and the beef industry to accept their policy
preferences. It also flattered the scientists by repre-
senting them as authoritative and influential.
We have argued, however, that those arrangements
not only allowed political decisions to be taken under
the guise of science but also that the spectrum of pol-
icy choices available on BSE was rendered opaque
and that in practice officials undermined the demo-
cratic accountability of ministers. Not only did those
arrangements give rise to political and democratic
problems, they also contributed to making a very seri-
ous problem considerably worse.
We have shown how the multiple, but opaque, po-
litical and social roles played by expert committees
entailed, albeit sometimes inadvertently, that vitally
important regulations concerned with pharma-
ceuticals, occupational health and the removal of
cattle offal from the human food chain to protect
public health were widely viewed within the trade as
unnecessary. Furthermore, those arrangements have
done much to shatter public trust and confidence in
‘scientific’ advice and public policy-making.
We have therefore provided an analysis of the role
and conduct of the Southwood Working Party and in-
dicated the utility of the ‘contemporary’ model as a
general account of the relationships of science and
policy. Finally, although none of our findings is in-
consistent with the evidence available to the Phillips
Inquiry into BSE, the Inquiry did not draw attention to
most of our empirical and analytical conclusions.
The BSE saga demonstrates that in UK policy-making
there is too little clarity about how the scientific
aspects of risk assessment and the political aspects of
risk management can, and should be, both differenti-
ated from each other and coupled together. On the one
hand, it is crucial to ensure that analyses and repre-
sentations of risks are not subordinated to the prior
decisions of risk managers, and that ministers do not
hide behind their officials and advisors, pretending
that they are merely following the advice of scientific
experts and not responsible for any of the key judge-
ments or decisions. On the other hand, a complete
separation of science and policy would not be desir-
able or feasible: scientific decision-making cannot
operate in a policy vacuum and policy decision-
making cannot operate in a scientific vacuum.
Explicit political framing assumptions
Upstream framing assumptions are sometimes made
explicit in committees’ terms of reference but tradi-
tionally their remit is often vague and ambiguous and
few of the framing assumptions are traditionally cap-
tured in the terms of reference. If the relationships
between science and policy-making are to be
transparent, rather than opaque, prior framing as-
sumptions should be explicit rather than implicit. This
also implies an arrangement under which policy
contexts are revealed rather than concealed.
Those responsible for risk assessments should
therefore be provided with an explicit set of policy-
framing assumptions about their remit and responsi-
bilities, for example, about which kinds of risk they
should address and by reference to what criteria.7
Those are political judgements that, once having
been made, can be addressed scientifically by risk
assessors, but they are not in themselves scientific
judgements about risk.
Decisions about the choice of such framing
assumptions should rest primarily with those who are
democratically accountable, namely ministers. Other-
wise we must expect that risk assessments will remain
entangled with policy decisions in ways that will com-
promise the independence of scientific advisors, and
allow political decision-making to masquerade as if it
were based solely on science. Scientific advisors
should, however, be at liberty to introduce additional
framing assumptions, so long as those are acknowl-
edged and explicitly justified.
Scientific advisors should, moreover, be responsi-
ble for demonstrating that their advice has in practice
been framed within parameters that can be justified, if
not by the prior framework set out by ministers, at
least by a subsequent democratically accountable pol-
icy framework. If expert scientific advisory commit-
tees were to explain how they had reached their
conclusions, given the available evidence and uncer-
tainties, and the framing assumptions provided, that
would oblige them to reveal at least some of the non-
scientific considerations that had influenced their
thinking and analyses, and the ways in which they had
been deployed.
Science and Public Policy April 2001 109
Politics of expert advice
Those responsible for risk assessments
should be provided with an explicit set
of policy-framing assumptions about
their remit and responsibilities, for
example, about which kinds of risk it
should address and by reference to
what criteria
Scientific advice: pluralistic and conditional
The task of having to choose how to respond to uncer-
tain risks will never be straightforward but hopefully
it can be well informed. One of the conditions for be-
ing well informed would be to oblige expert advisors
to draw to policy-makers’ attention as many of the
uncertainties and gaps in knowledge as possible. If
scientific advisors are responsible for providing polit-
ical as well as scientific advice, as was the case with
Southwood, then there will inevitably be incentives to
limit a discussion of the uncertainties.
As we have shown in this paper, under such cir-
cumstances committee members may be aware of
many types of uncertainty when producing assess-
ments of risks and when making recommendations,
but often fail to make some of those key uncertainties
explicit. Consequently, ministers and senior policy-
makers may not be aware of the fragility of scientific
knowledge and thus the spectrum of possible risks and
subsequent policy options.
Advisory committees should therefore provide
predominantly scientific advice. That, however,
would involve committees confining their comments
to the following types of consideration:
·the existence of possible risks;
·what is known about their significance and
·the limits of our knowledge about those risks and
the extent of the attendant uncertainties;
·the adequacy with which those risks have been in-
vestigated; and
·the further investigations that could be undertaken
and what they might show.
Such scientific considerations can never, by them-
selves, entail specific judgements about what levels of
risk, and which levels of uncertainty, are deemed to be
socially acceptable. Scientific advice to policy-
makers cannot, therefore, be definitive, monolithic or
prescriptive, rather it should be pluralistic and
conditional8(Stirling, 1999; EEA, forthcoming).
Expert scientific advisors should be expected to
discuss a range of possible courses of action and they
should set out the possible and probable consequences of
following, or failing to follow, those different courses of
action, but their expert knowledge, by itself, cannot jus-
tify the prescriptive advocacy of any single policy.
Nevertheless, scientists cannot conduct their risk assess-
ments in a policy vacuum; they need to be aware of the
policy options available to ministers, and the kinds of
approach which ministers envisage adopting.
Advisors should be informed of the range of possi-
ble courses of action that ministers might take and
should be asked to perform their assessments within
that framework. Scientific advisors should, however,
be at liberty to introduce alternative policy options.
The process could be iterative.
Advisors’ discussions of uncertainty should be par-
ticularly explicit. For each policy option, scientific
advice should involve the presentation of a range of
possible risk scenarios, each of which could be linked
to a scientific judgement about the relative plausibil-
ity of each scenario, and the consequences of a possi-
ble error of scientific judgement in believing each
scenario to be true. Advisors could also indicate
which types of evidence, if they were to emerge,
might lead them to re-evaluate their judgements.
If scientific advice was conditional and pluralistic
in the senses outlined above, ministers would be pre-
sented with a discussion of the possible risks and ben-
efits of taking, or failing to take, a wide range of
possible policy options. Ministers would therefore be
as well informed as possible about risk appraisal op-
tions but the responsibility for deciding what consti-
tutes a proportionate response, for example to BSE, in
a context in which there is an absence of complete or
unequivocal scientific knowledge, would rest with
democratically elected representatives rather than
with experts or officials. This would not only help
Ministers exercise their political judgements, it would
also make policy-making processes more transparent
and democratically legitimate and it would enhance
the robustness of regulatory policy-making by cre-
ating a clear point of engagement for different public
and professional perspectives on any given regulatory
Institutional independence and external scrutiny
Procedural demarcation of science and policy may be
insufficient to prevent risk assessment decisions from
being determined by prior risk-management priori-
ties, or decisions that transcend the limits of science
from being misrepresented as if they were scientific.
As the UK Government and the European Commis-
sion have realised, both an institutional separation of
risk assessors from risk managers and a move to open
up risk-assessment processes to external scrutiny are
This has, for example, taken the form in the UK of
the newly created Food Standards Agency (FSA), at
arm’s length from the Department of Health. In the
Commission there is an analogous plan to create a Eu-
ropean Food Authority at arm’s length from what had
been DG-XXIV, and what has now became known as
DG-SANCO, or the Directorate General for Health
and Consumer Protection.
In the UK, the intention was that the FSA would
provide scientific and administrative expertise, and
that its remit would be to advise ministers, but that
policy-making would be the responsibility of minis-
ters. In practice, however, the FSA is taking resp-
onsibility for making risk-management decisions.
Furthermore, it is often not explaining the non-
scientific considerations by reference to which those
judgements are being made.
The FSA is committed to a far greater degree of
transparency than has been traditional in British regu-
latory agencies but that innovation only applies to this
agency and it remains at the FSA’s discretion. If BSE
110 Science and Public Policy April 2001
Politics of expert advice
policy-making in the UK had been governed by a
proper freedom of information regime, rather than by
official secrecy, MAFF’s policies would have been
scientifically and politically unsustainable.
Secrecy allowed ministers and senior officials to
maintain the public pretence that policies were fo-
cused on consumer protection and based on secure
scientific foundations, while privately acknowledg-
ing that regulations were being decided entirely on
political rather than scientific grounds. At present,
however, the UK Government’s Freedom of Informa-
tion Act, that received royal assent on 30 November
2000, will not be sufficient to prevent the repetition of
such deceit (Stationery Office, 2000; Campaign for
Freedom of Information, 2000)
It is not uncommon for ostensibly scientific screens to
be invoked to conceal a Byzantine complexity of dif-
fering and cross-cutting agendas or for multiple
agents to claim scientific legitimacy for their differing
preferences, decisions and actions. The kinds of frag-
mentation and failures of communication within what
might otherwise be thought of as well-integrated and
cohesive policy networks might well be duplicated in
other regimes of public governance. It seems prima
facie unlikely that the pattern we have highlighted
was unique to this area of policy. This is an issue to
which we intend to return at a later date.
1. Volume Two of the Phillips Report (Phillips et al, 2000) provides
a readable account of existing scientific knowledge about BSE,
and other TSEs, in the late 1980s.
2. Experimental evidence indicated that it was then not possible
to predict what the properties, including the pathogenicity, of a
given strain of scrapie would be once it had ‘passaged’ to an-
other species (Kimberlin et al, 1987). MAFF scientists were
clearly aware of that evidence. For example, in July 1988, the
Assistant Chief Veterinary Officer told the UK rendering trade
association that: “… the BSE agent could not be equated with
scrapie agent and, even if they were the same, they would not
necessarily behave in the same way after jumping the species
barrier …” (BSE Inquiry, 1999c, paragraph 224).
3. Even the option of making the disease notifiable, an essential
precursor to learning about the epidemic, was rejected be-
cause, as one official put it, such action “… might imply to the
general public we know something they don’t like the meat or
milk is a source of danger for humans” (Phillips et al, 2000, vol-
ume 3, paragraph 2.130).
4. The implications of a wider ban on feeding animal protein to an-
imals went even further than the immediate loss of markets to
the rendering industry. If, for example, ruminant protein was
banned from pig feed it would be an obvious step to ask why it
was still considered safe to feed ruminant protein to humans (at
that time even cattle with clinical symptoms of BSE were still
going into the human food chain).
5. As Southwood later acknowledged “[r]ecognising the practical
problem of extensive quantities of [meat and bone meal], we
did not recommend extending the ban to pigs and poultry”
(Southwood et al, 1999a, paragraph 124).
6. As a DoH minute of a meeting held on 6 June between the
Agriculutre Minister, MAFF officials and the deputy Chief Medi-
cal Officer noted: “[f]rom the outset it was clear Mr MacGregor
believed there was an overriding political case for action even if
the science had not changed. He was under increasing
pressure from the public and the industry to ban bovine offal
from the human food chain. There was talk of the food industry
taking its own measures and he could not allow MAFF to be up-
staged by the industry” (BSE Inquiry, 1999d, para 135).
7. The influence of such framing assumptions is particularly ex-
plicit in the agricultural biotechnology arena. For example, US
risk assessments are framed by the assumption that the only
risks from the introduction of GM (genetically modified) crops
that are deemed relevant are those which might have an ad-
verse impact on the financial viability of US farming, while in
most EU member states the risk assessment is framed more
widely to include adverse effects on other flora and fauna
(House of Lords, 1998). In Austria, the assessment of risk is
framed even more widely because the Austrians require evi-
dence not merely that a new GM technology is no worse than
currently prevailing intensive farming, they want it to be at least
as environmentally benign as organic farming, and moreover
to provide definite environmental benefits when compared to
currently available alternatives (Torgesen and Seifert, 2000).
8. Our thanks to Andy Stirling for this specific wording.
Donald Acheson (1999), Witness Statement no 251a, in Phillips et
al (2000), Disc 11.
BSE Inquiry (1999a), “The Southwood Working Party”, Revised
Factual Account no 1, available at, last
accessed 7 November 2000.
BSE Inquiry (1999b), “Slaughter and compensation”, Revised Fac-
tual Account no 6, available at, last
accessed 7 November 2000.
BSE Inquiry (1999c), “Ruminant feed ban”, Revised Factual Ac-
count no 7, available at , last accessed 7 November 2000.
BSE Inquiry (1999d), “The Introduction of the SBO ban”, Revised
Factual Account no 8, available at, last
accessed 7 November 2000.
BSE Inquiry (1999e), “Government and the Southwood Working
Party”, Revised Factual Account no 9, available at, last accessed 7 November 2000.
BSE Inquiry Transcripts (2000), in Phillips et al (2000), Disc 12.
Campaign for Freedom of Information (2000), BSE and Secrecy:
Implications for the Freedom of Information Bill, 10 November
(Campaign for Freedom of Information, London).
Channel 4 (2000), Beef Encounters, broadcast on 8 October 2000.
Alistair Cruickshank (1998), Witness Statement no 75 to BSE In-
quiry, in Phillips et al (2000), Disc 11.
Steve Dealler (1996), “Can the spread of BSE and CJD be pre-
dicted?”, in S C Ratzan (editor), The Mad Cow Crisis (UCL
Press, London).
DTI, Department of Trade and Industry (2000), Review of Risk Pro-
cedures used by the Government’s Advisory Committees Deal-
ing with Food Safety, Report of the Group led by Sir Robert May,
Chief Scientific Advisor, July (Department of Tradeand Industry,
EEA (forthcoming), Late Lessons from Early Warnings: The
Precautionary Principle in Policy 1896–2000 (European Envi-
ronment Agency, Copenhagen, February 2001).
James Erlichman (1988), “Butchers selling diseased meat”, Guard-
ian, 29 June.
European Commission (2000), White Paper on Food Safety in the
European Union, COM(99)719, 12 January.
Annabel Ferriman (1989), “Danger in food chain: pigs may carry
‘mad cow’ disease”, Observer, 17 September, page 3.
R Gilpin (1962), American Scientists and Nuclear Weapons Policy
(Princeton University Press, Princeton).
John Gummer (1990a), Hansard, 8 June, column 906 (The Statio-
nery Office, London).
John Gummer (1990b), Hansard, 21 May 1990, 173(110), column
82 (The Stationery Office, London).
Andrew F Hill, Susan Joiner, Jackie Linehan, Melanie Desbruslais,
Peter L Lantos, and John Collinge (2000), “Species-barrier-
independent prion replication in apparently resistant species”,
Proceedings of the National Academy Of Sciences, 97(18),
pages 10248–10253.
House of Lords, Select Committee in the European Communities
(1998), Regulation of Genetic Modification in Agriculture,Re
port, 15 December, HL Paper 11-I (The Stationery Office,
Earl Howe (1998), Witness Statement no 301, in Phillips et al
(2000), Disc 11.
Science and Public Policy April 2001 111
Politics of expert advice
David Isaacs and Dominic Fitzgerald (1999), “Seven alternatives to
evidence based medicine”, British Medical Journal, 319, 18–25
December, page 1618.
Sheila Jasanoff and Brian Wynne (1998), “Science and decision-
making”, in S Rayner and E L Malone (editors), Human Choices
and Climate Change: volume 1 — the societal framework
(Battelle Press, Ohio).
R H Kimberlin, S Cole andCAWalker (1987), “Temporary and per-
manent modifications to a single strain of mouse scrapie on
transmission to rats and hamsters”, Journal of General Virology,
68, pages 1875–1881.
Les Levidow, Susan Carr, David Wield and Rene von Schomberg
(1997), “European biotechnology regulation: framing the risk
assessment of a herbicide-tolerant crop”, Science, Technology
and Human Values, 22, pages 472–505.
Donald MacIntyre, Liz Hunt and Charles Arthur (1996), “Ministers
admit possible link between mad cow disease and human
deaths”, Independent, 21 March, page 1.
David Maclean (1990), Comments in This Week, transmitted on
ITV, 17 May.
Keith Meldrum (1995), Comments in “BSE the hidden epidemic”,
World In Action, transmitted on ITV, 13 November.
MAFF, Ministry of Agriculture, Fisheries and Food (1990a), British
Beef Is Safe, Gummer, News Release 185/90, 15 May (Ministry
of Agriculture Fisheries and Food, London).
MAFF, Ministry of Agriculture, Fisheries and Food (1990b), Gov-
ernment Action on BSE, News Release FF 1/90, 9 January
(Ministry of Agriculture Fisheries and Food, London).
MAFF/DoH, Ministry of Agriculture Fisheries and Food/Department
of Health (1989), Report of the Working Party on Bovine
Spongiform Encephalopathy (Ministry of Agriculture Fisheries
and Food, London).
Fred Pearce (1996), “Ministers hostile to advice on BSE”, New Sci-
entist, 30 March, page 4.
Lord Phillips of Worth Matravers, June Bridgeman and Malcolm
Ferguson-Smith (2000), The BSE Inquiry: Report: evidence and
supporting papers of the Inquiry into the emergence and identifi-
cation of Bovine Spongiform Encephalopathy (BSE) and variant
Creutzfeldt–Jakob Disease (vCJD) and the action taken in re-
sponse to it up to 20 March 1996 (The Stationery Office,
David Robbins and Ron Johnston (1976), “The role of cognitive and
occupational differentiation in scientific controversies”, Social
Studies of Science, 6, pages 349–368.
Richard Southwood (1998), Witness Statement no 1 to BSE Inquiry,
in Phillips et al (2000), Disc 11.
Richard Southwood, Anthony Epstein, William Martin and Lord
Walton (1999a), Witness Statement no 483, in Phillips et al
(2000), Disc 11.
Richard Southwood, Anthony Epstein, William Martin and Lord
Walton (1999b), Witness Statement no 483A, in Phillips et al
(2000), Disc 11.
Southwood Working Party (1999), Witness Statement no 1D, in
Phillips et al (2000), Disc 11.
Stationery Office (2000), available at http://www.legislation.hmso., last accessed 30 April
2001 .
Andy Stirling (1999), On Science and Precaution in the Manage-
ment of Technological Risk, volume I (Institute for Prospective
Technological Studies, Seville)
Helge Torgesen and Franz Seifert (2000), “Austria: precautionary
blockage of agricultural biotechnology”, Journal of Risk Re-
search, 3(3), pages 209–218.
Patrick van Zwanenberg and Erik Millstone (2000), “Beyond skepti-
cal relativism: evaluating the social constructions of expert risk
assessments”, Science, Technology and Human Values, 25(3),
pages 259–282.
Alvin Weinberg (1972), “Science and trans-science”, Minerva, 10,
pages 209–222.
Brian Wynne (1992), “Uncertainty and environmental learning:
reconceiving science and policy in the preventive paradigm”,
Global Environmental Change, 2, pages 111–127.
112 Science and Public Policy April 2001
Politics of expert advice
... Background. Academic research on technical advice to policy commonly focuses on social and related policy areas such as health, education and crime (Oliver et al. 2014) and disciplinary advice from science disciplines (Jasanoff 1994;Millstone and van Zwanenberg 2001). Little or no prior research in the social sciences have explored engineering expertise in policy domains where such advice is critical (e.g. ...
... There is a perspective based around 'evidencebase/informed policy making' (Greenhalgh and Russell 2009;Head 2010;Nutley et al. 2007); there is a perspective around 'science advice' (Gluckman 2014;Jasanoff 2013;Pielke 2007;Schenkel 2010) and there is a perspective around what can be called 'expertise in policy' (e.g. Millstone and van Zwanenberg 2001;Page 2010). These literatures can be mapped in different ways revealing certain kinds of gaps which the study here seeks to fill. ...
... 'Science advice' and 'expertise in policy' also often tend to focus on health-related topics (including epidemics, xenotransplantation (Weale 2001) or antimicrobial resistance (Spruijt et al. 2014)) but also are concerned with other physical science-oriented areas such as agricultural biosciences (e.g. GM crops, and diseases in cattle: Millstone and van Zwanenberg 2001;Oreszczyn 2005;Schwach et al. 2007). A similar Scopus search as per above for the phrase "science advice" returns 141 articles stretching back to 1963 with a similar topical orientation in the titles of the top 10 most-cited papers. ...
Most, if not all empirical research on evidence-based policy has three features: firstly, it typically focuses on the application of science and scientific expertise on policy; secondly, it is executed by ‘outsider’ researchers who are not part of the public administration or policy-making process but observers of it (for example, Stevens, 2010); and thirdly, the major topical focus is in social policy areas such as health, education and crime (Oliver et al, 2014). This study advances the perspectives on evidence-based policy making by exploring the role of engineering expertise in policy making. We first make the case that, although related, science and engineering represent different epistemic communities in relation to policy practice. This difference, we argue, can give rise to particular styles of interaction that can make the governance of engineering expertise in policy making different to that for science or scientists. We then report on the findings of a study of the relationship between a new engineering team in a UK ministry with a technical portfolio and the policy colleagues they worked with across a range of programme areas. Through 18 interviews with policy officials, we identify a range of interactions that imply a need to consider styles of management and approaches to internal deployment of experts within policy organisations, as well as the implications for policy making and engineering expertise, given the way policy and engineering practices overlap. Key messages Engineering advice has never been properly identified and studied in the academic social science literature to date. Engineering advice is an important and potent source of evidence in policy making in topical areas like energy policy. In contrast to science advice, engineering advice as a practice significantly overlaps with policy practice meaning important conflict or complementarity is possible, dependent on how the advice is deployed. </ul
... Instead, they concluded that "secrecy allowed ministers and senior officials to maintain the public pretence that [. . . ] policies were based on secure scientific foundations, while privately acknowledging that regulations were being decided entirely on political rather than scientific grounds" [148]. ...
Full-text available
Epidemic disease spreading is conventionally often modelled and analyzed by means of rate and diffusion equations, following the paradigms of well-controlled chemical reactions and diffusive dynamics in a test tube. Yet, serious worries that this suggestive and appealing similarity might be a false friend were already voiced by the pioneers of mathematical epidemiology. A century later, we can draw on cross-fertilizations from network and game theory and the emerging field of eco-evolutionary dynamics to substantiate them. Epidemiological spreading is thereby revealed as a fundamentally heterogeneous and erratic process that shares certain properties with more unwieldy phenomena, such as earthquakes, hurricanes, traffic jams, and stock crashes. They are all characterised by high tail risks that materialize very rarely but fatally. That they arise from bursts of unlikely chains of localized random "superspreader" events, by which micro-scale fluctuations and uncertainties may get heftily magnified, makes their accurate prediction and control intrinsically and notoriously hard. That epidemic disease spreading is moreover closely intertwined with equally heterogeneous genetic drift and information feedback adds new challenges -- and chances.
... This was noted in a study by Millstone and van Zwanenberg of the working group advising the UK government during the BSE affair. They describe how the ostensibly independent scientists ended up acquiescing with officials to give messages that were deemed acceptable to ministers and the meat industry, contributing "to making a very serious problem considerably worse" [17] . ...
The COVID-19 pandemic has shone a light on the complex relationship between science and policy. Policymakers have had to make decisions at speed in conditions of uncertainty, implementing policies that have had profound consequences for people's lives. Yet this process has sometimes been characterised by fragmentation, opacity and a disconnect between evidence and policy. In the United Kingdom, concerns about the secrecy that initially surrounded this process led to the creation of Independent SAGE, an unofficial group of scientists from different disciplines that came together to ask policy-relevant questions, review the evolving evidence, and make evidence-based recommendations. The group took a public health approach with a population perspective, worked in a holistic transdisciplinary way, and were committed to public engagement. In this paper, we review the lessons learned during its first year. These include the importance of learning from local expertise, the value of learning from other countries, the role of civil society as a critical friend to government, finding appropriate relationships between science and policy, and recognising the necessity of viewing issues through an equity lens.
... The rapidly developed models were the basis of the decision to undertake widespread preemptive culling of many animals. Post-footand-mouth, modelers got a key role in SAGE; indeed, Neil Ferguson headed the modelling in both foot-and-mouth epidemic and early stages of COVID-19 in the United Kingdom (Millstone 2001;Ford 2020). This continuity between the two big crises could lead us to think that using mathematical modelling in foot-and-mouth was an undisputed success. ...
Full-text available
We argue for a foundational epistemic claim and a hypothesis about the production and uses of mathematical epidemiological models, exploring the consequences for our political and socio-economic lives. First, in order to make the best use of scientific models, we need to understand why models are not truly representational of our world, but are already pitched towards various uses. Second, we need to understand the implicit power relations in numbers and models in public policy, and, thus, the implications for good governance if numbers and models are used as the exclusive drivers of decision making.
... One of those clouds, however, was bovine spongiform encephalopathy (BSE) or 'mad cow disease'. At the same time that Bodmer's committee was writing its report, farmers in the UK were reporting cattle collapsing from a disease that left them staggering and slobbering across the farmyard and dying in considerable discomfort (see Jasanoff, 1997;Millstone and van Zwanenberg, 2001;Reeves, 2002;Frewer et al., 2002, for examples of scholarly approaches to BSE and the crisis it caused in the UK). The Central Veterinary Laboratory first recognised BSE as a novel cattle disease in 1986. ...
Full-text available
Modern science communication has emerged in the twentieth century as a field of study, a body of practice and a profession—and it is a practice with deep historical roots. We have seen the birth of interactive science centres, the first university actions in teaching and conducting research, and a sharp growth in employment of science communicators. This collection charts the emergence of modern science communication across the world. This is the first volume to map investment around the globe in science centres, university courses and research, publications and conferences as well as tell the national stories of science communication. How did it all begin? How has development varied from one country to another? What motivated governments, institutions and people to see science communication as an answer to questions of the social place of science? Communicating Science describes the pathways followed by 39 different countries. All continents and many cultures are represented. For some countries, this is the first time that their science communication story has been told. Edited by: Toss Gascoigne, Bernard Schiele, Joan Leach, Michelle Riedlinger, Bruce V. Lewenstein, Luisa Massarani, Peter Broks Download the book FREE at:
... One of those clouds, however, was bovine spongiform encephalopathy (BSE) or 'mad cow disease'. At the same time that Bodmer's committee was writing its report, farmers in the UK were reporting cattle collapsing from a disease that left them staggering and slobbering across the farmyard and dying in considerable discomfort (see Jasanoff, 1997;Millstone and van Zwanenberg, 2001;Reeves, 2002;Frewer et al., 2002, for examples of scholarly approaches to BSE and the crisis it caused in the UK). The Central Veterinary Laboratory first recognised BSE as a novel cattle disease in 1986. ...
The 1950s British Prime Minister Harold Macmillan is reported to have replied to a journalist’s question as to what was likely to blow his government off course with the words ‘events, dear boy, events’. The development of science communication in the UK from the mid-1980s onwards is one of the best-documented stories in this field, punctuated by a series of reports from both the scientific community and the government itself—and by a number of ‘Macmillian events’ that blew science’s relationship with the wider world hither and thither. Here we offer a series of episodes that changed how we think about the relationship between science and society. We describe these largely chronologically and imply that they heralded new eras of science communication. This does not mean, however, that previous approaches simply disappeared: many old ideas were buried momentarily or continued as an undercurrent, less visible but ready to resurface as and when conditions allowed and required them to do so.
... A 50-year history of land management and land use policy decisions is peppered with the neglect of landowner and farmer perspectives and situated knowledge. This has been demonstrated to ruin relationships and have disastrous and long lasting consequences (for examples see Lowe et al (1997) on agriculture and water pollution [5] ; Brian Wynnes seminal work on the Chernoble fall out and sheep farming "May sheep Safely graze" [6] ; and critiques of the British expertise model and the BSE crisis [7,8] ). Farmer and Landowner motivations for not planting have been identified in research [4,9,10] , however perceptions and preferences in regards reasons for proactive tree planting is less well evidenced (with a few key exceptions in international contexts [11][12][13] ). ...
Full-text available
Confor essay competition entry #thefutureisforestry in response to the question 'how can we motivate farmers to plant more trees' Judges said: “Jenny set up the context for the discussion quickly, then questioned the question - and delivered a very coherent and persuasive argument as to how we can move forward, focused on relationship-building, communication and challenging perceptions. She identified very specific logistical challenges to planting at a local level, plus problems of knowledge and funding.” The essay critiqes the perceptions of both forestry and farming, suggesting that communication is needed to ensure options that are suitable for farmed landscapes landscape, both socially, physically and environmentlally. Essay published on Confor site 8th April 2020 and in the Confor April
Many health-related disasters in the recent past were man-made, or their fundamental mistakes in their prevention or mitigation. In many challenges, mankind was initially unaware, followed by a stage where one side had the interest to deny, and the other side fought against harm. When finally legal and administrative measures are taken, then this happens within the framework of the usual political theater. As increasingly challenges cannot be seen with the naked eye, in their intellectual processing science plays a key role. But also science is not as neutral as it would like to be. Instead of blindly trusting the authorities, it is always better to maintain common sense, look behind the curtains, be aware of conflicts of interest behind all facades, and know that there is no absolute truth.
Full-text available
We argue for a foundational epistemic claim and a hypothesis about the production and uses of mathematical epidemiological models, exploring the consequences for our political and socio-economic lives. First, in order to make the best use of scientific models, we need to understand why models are not truly representational of our world, but are already pitched towards various uses. Second, we need to understand the implicit power relations in numbers and models in public policy, and, thus, the implications for good governance if numbers and models are used as the exclusive drivers of decision making.
Technical Report
Full-text available
Editorial team: Poul Harremoës (Chairman) David Gee (EEA editor) Malcolm MacGarvin (Executive editor) Andy Stirling (Editor) Jane Keys (Editor) Brian Wynne (Editor) Sofia Guedes Vaz (EEA editor) Project managers: David Gee and Sofia Guedes Vaz European Environment Agency Environmental issue report No 22
Full-text available
Constructivist analyses of risk regulation are typically agnostic about what should count as robust or reliable knowledge. Indeed, constructivists usually portray competing accounts of risk as if they were always equally contingent or engaged with different and incommensurable issues and problem definitions. This article argues that assumptions about the equal reliability of competing accounts of risk deserve to be, and sometimes can be, examined empirically. A constructivist approach grounded in epistemological realism is outlined and applied empirically to a particular comparative U.S./U.K. case study of pesticide regulation. The article argues that while the scope for interpretative flexibility when addressing risk issues is clearly extensive, it is not unconstrained. By scrutinizing the structure and coherence of particular risk assessments and policy decisions by reference to both empirical evidence and commonly held robust standards of interpretation, the article argues that the U.K. evaluation was not only less precautionary than its U.S. equivalent, but it was also less well constructed and therefore less reliable. Several social and institutional characteristics of U.S. and U.K. policy making are highlighted that appear variously to facilitate or inhibit the production of reliable knowledge and the making of prudent policy decisions.
Full-text available
Austria has interpreted the precautionary principle and Directive 90/220 in a more stringent way than other EU member states. It continues to ban the import of Bt maize despite the Commission's recurrent warnings. The Austrian standard of GMO risk assessment emphasizes a broad deénition of adverse effects beyond a purely technical account of risk, including effects of agricultural practices. Boundaries between plant, seed, food, and feed assessments tend to blur. It asks implicitly for the demonstration of safety and uses organic farming as a normative reference point. The understanding of precaution goes beyond the Danish approach in extensively interpreting the scope of Directive 90/220. This policy originated from the Environment Agency (UBA) and developed out of the division of labour among government agencies. It is in line with the inherent pater- nalism of Austrian governance as well as with Austrian public sensitivities concerning organic agriculture and food. When public opinion turned hostile to agricultural biotech- nology, the Austrian standard became entrenched and led to Austria's initially lone stance among EU member states. KE Y W O R D S: agricultural biotechnology; GM crops; precaution; policy standard; policy streams; policy entrepreneur; Austria
Full-text available
As products of the "new biotechnology," genetically modified organisms have provoked a wide-ranging risk debate on potential harm, especially from herbicide-tolerant crops. In response to this legitimacy problem, the European Community adopted precautionary legislation, which left open the definition of environmental harm. When the U.K. proposed Europe-wide market approval of a herbicide-tolerant oilseed rape (canola), the proposal encountered dissent from some countries and environmentalisi groups. Further debate on normative judgments became necessary to implement the precaution- ary legislation. In dispute were several regulatory boundaries-of administrative re- sponsibility, causality, acceptability, and evidence. The boundary disputes expressed divergent framings of biotechnological risk, each with its implicit model of the socionatu- ral order In this way, the disputes can illuminate the sorts of risk framings that have already become embedded and standardized in other regulatory sectors.
The author considers the implications for current assumptions about scientific knowledge and environmental policy raised by the preventive approach and the associated Precautionary Principle. He offers a critical examination of approaches to characterizing different kinds of uncertainty in policy knowledge, especially in relation to decision making upstream from environmental effects. Via the key dimension of unrecognized indeterminacy in scientific knowledge, the author argues that shifting the normative principles applied to policy use of science is not merely an external shift in relation to the same body of 'natural' knowledge, but also involves the possible reshaping of the 'natural' knowledge itself.
The interspecies transmission of scrapie is frequently associated with exceptionally long incubation periods at first passage in the new host compared to later passages (the species barrier effect). The basis of this was investigated using the 139A strain of scrapie which had been cloned by three serial passages in mice at limiting infectious doses. Cloned scrapie was passaged through hamsters (twice) or rats (thrice) and then reisolated in mice. Large species barrier effects were encountered on mouse-to-hamster and hamster-to-mouse passage resulting in the isolation of a mutant strain, 139-H/M, with properties very different from 139A. In contrast, the strain reisolated from rats was indistinguishable from 139A. However, a large species barrier was encountered at the mouse-to-rat passage but not at the rat-to-mouse passage. It is suggested that the transmission of scrapie between species may be associated with no change in properties or a permanent change in the scrapie genome due to the selection of mutants. A third possibility, the donor species effect, is a temporary change occurring only at first passage in the new host species which is largely or entirely caused by the introduction of material from the previous host. We speculate that the donor species effect could be explained if some host protein forms a functional part of the infectious agent.
Nature is the international weekly journal of science: a magazine style journal that publishes full-length research papers in all disciplines of science, as well as News and Views, reviews, news, features, commentaries, web focuses and more, covering all branches of science and how science impacts upon all aspects of society and life.