Home

About Green Audit

Green Audit Services

Publications

Papers

Epidemiology

Science & Policy Making

Political Theory

Azande Science

Links

   Publications

 I Don't Know Much About Science

Price £5.00 + £2.00 / €8 + €4.00Green Audit
Castle Cottage, Sea View Place, Aberystwyth, SY23 1DZ
Tel. +44-1970-630215
Email: admin@greenaudit.org
 

This report considers the processes which lead to government decision-making in areas which involve technical or scientific knowledge. It starts by analysing responses to a questionnaire, sent to all Members of Parliament in 1997. This asked MPs to put down their scientific or technical qualifications or other expertise in the area of mathematics and science. In addition, all MPs were classified in terms of their background and qualifications, using publicly available data. Results indicated that about 30 percent of all MPs had no formal qualifications in basic science or mathematics (i.e. ‘O’levels). In addition, over 80 percent of MPs were from backgrounds and professions where no scientific knowledge was necessary or assumed.

The question is raised that such a state of affairs might result in decisions being made where harmful processes became permitted as a result of lack of knowledge or ability to sift through scientific reports. Some examples are given of cases where this may have occurred, for example BSE, GM food and Chernobyl radiation. The example of the Commons Backbench Environment Committee is explored and it is found that there are presently eleven members, none of whom has a degree in a hard science subject and two of whom do not have ‘O’ level mathematics.

In a general discussion, the problems of making decisions on technically or scientifically complex matters are illustrated using recent developments in the area of philosophy of science, collectively known as ‘science studies’. In particular, it is argued, by reference to a number of contemporary issues, including health risks from nuclear power, BSE and genetically modified foods, that the present advisory committee system of scientific advice to government is fundamentally flawed. This is because no expert committee can, according to the discoveries made by ‘science studies’, be truly independent of bias, a problem exacerbated by the large sums of scientific funding available from multi-national corporations. Thus ‘the best available scientific advice’ given out as a platform for government decisions is presently only the consensus of a committee which usually holds a quite specific, and not always balanced, position on the topic in question.

In order to overcome this problem, a novel structure of oppositional committees is advocated. This system is based on the adversarial model of parliamentary democracy and the British legal system and would involve the MPs’ consideration of reports or presentations prepared by two different scientific groups separately commissioned to research the arguments in support of or opposing a particular project or suggestion. This type of approach, which is similar to the STOA unit approach to scientific advice given to members of the European Parliament, makes use of both the strengths of MPs, that of exercising judgement in areas of human cues and communications, and those of scientists, those of weighing and making judgements about questions involving complex scientific information. In addition the report proposes the funding of ‘citizen scientists’ to explore the public health aspects of new technological developments

1. Introduction: Technological Decision-Making Requires Technical Information

‘The government’s position is based on the best scientific advice’. This statement is uttered frequently by politicians seeking to bolster unpopular decisions. But how reassuring is it? The decision to allow beef contaminated with BSE to be fed to vulnerable people, resulting in the deaths of more than 50 UK citizens to date, was based on the best scientific advice. So is the decision to continue to operate the Sellafield reprocessing facility, although the scientific advice which suggests that the levels of radioisotopic contamination and human exposure causes no harm was developed before DNA had been discovered and is looking increasingly insecure.

Developments in technology have now reached the level where the introduction of new processes, substances, or practices may have unforeseen but catastrophic consequences for life on earth. In addition, as the Chernobyl accident has shown, events in one country can cause death or disease, or economically disastrous consequences in a different country many hundreds of miles distant. Who is regulating technological development? We live in an age when industrialists have access to more novel processes, materials and substances than ever before and so decisions about the regulation and direction of production within our economy are inevitably highly technical. Who, apart from a few scientists closely associated with the development of these processes or substances, even begins to understand the possible consequences of their use? And these very people are least likely to be objective when decisions are to be made over holding up the use of the products of their research, especially where large sums of money are involved.

What do these considerations imply for the future of a democracy within which most citizens are apparently incapable of judging scientific facts for themselves? Must we, and the politicians we elect, rely on the expertise and independence of scientists? And to what extent can this independence be achieved when so much university research is now funded by corporations?

We need to be sure that our elected representatives can make a judgement about the quality of scientific advice, but are they educated and equipped to do this? How much scientific education is it reasonable to expect of our MPs? Can anybody do the job effectively without even O-level maths? Our findings show that a proportion of MPs may be doing just that.

2. How Much do MPs Know about Science?

2.1. MPs’ Scientific Qualifications

In order to examine the area of scientific policy-making, we first conducted a survey of MPs. We sent a questionnaire to all MPs, outlining our research project and asking them what scientific, mathematical or statistical qualifications they possessed, or whether they had any experience or had worked in any capacity which they felt might help them understand scientific matters. Had they, for example, conducted any research in a scientific or medical field? We asked them if they had obtained O-level Maths or studied any O-level science subjects at school. About 150 replied with completed questionnaires. Some refused to respond. A few were angry, and stated that they did not need to have any science or maths qualifications to make decisions about scientific matters as they either had advisors or they could consult people who did. We next went to the main publication containing CV information on MPs, Dod’s Parliamentary Companion, and used this information to classify them in terms of sex, age and their general area of expertise or life experience. For this we used the 11 categories shown in Table 1.

Table 1. Categories for classifying MPs’ area of knowledge and expertise

In analysing the questionnaire returns, the educational levels of MPs were classed by stages from no O-levels in science through to PhD level in science using the scheme shown in Table 2. These classification procedures were mainly applied in a negative sense, to establish the proportion of MPs in the sample with little or no experience or qualifications in the area of science, medicine and mathematics.

Table 2. Scientific knowledge level of MPs: categories in questionnaire sample by qualifications or work experience

^aHard sciences are physics, chemistry, mathematics, and medicine; soft sciences are computing and economics.

The results of our analysis of MPs’ knowledge and expertise, based on the category labels given in Table 1, are presented in Table 3. In Table 4 we show the scientific qualifications of the MPs in the sample who returned the questionnaire. In Table 5 we return to the total database and attempt to indicate the university degree status of all MPs by science, arts or no degree.

Table 3. Classification of MPs’ areas of expertise

Does this matter? For many of the political decisions that have to be considered, perhaps not. But in a world that is increasingly affected by scientific and technological knowledge, those who are unable to understand basic scientific or mathematical concepts are at a great disadvantage. In particular, they are prisoners to advice which is given by scientifically literate civil servants, expert committees, and lobbyists whose interests may be tied to transnational corporations. And such advice is increasingly biased towards permitting questionable or hazardous procedures and processes where the result is likely to increase profit or employment.

A good example is afforded by the recent decision to withdraw the contract for the operation of the Atomic Weapons Establishment (AWE), Aldermaston, from Hunting Brae Ltd and give it to a joint team from British Nuclear Fuels, the present operators of the Sellafield reprocessing plant in Cumbria and Lockheed Martin, a US firm with experience of operating nuclear plant. Both the old firm, Hunting-Brae, and the new firms have poor safety records, but the decision to bring in BNFL/ Lockheed was made at the time of international scandal following BNFL’s admission of routinely falsifying safety data on MOX fuel elements which had been sent to reactors in Japan and Germany. At the same time, evidence became available showing that the US government were refusing to give Lockheed Martin any more contracts because of their consistently poor safety record and management attitude to nuclear safety. For this reason, a BBC TV ‘Panorama’ reporter interviewed Defence Minister Baroness Symons in a programme televised on 27 March and asked why the government was using companies with such poor safety records to administer a nuclear plant where nuclear weapons were fabricated. Had she read the reports criticising the safety records of the companies?

In her response, Baroness Symons said:

I wouldn’t necessarily understand the individual bits. I have the humility to say that I am not a nuclear scientist, but there are those who are and who understand the reports in full. I have to rely on those with real expertise.

The implicit assumption made is that these ‘nuclear scientists’ who understood the individual bits would be unbiased either as a consequence of their connections with the nuclear industry or following their education or experience in the field of nuclear physics. It is unlikely, for example, that Baroness Symons consulted with experts who may also have ‘understood the individual bits’ but whose provenance or background was with Greenpeace, Friends of the Earth, or the Nuclear Awareness Group in Reading, the town most likely to be affected by the operation of the plant.

Table 4. Science qualifications of the sample of 93 questionnaire respondents

2.2. MPs’ employment background

There is also the question of the culture of the MPs who make decisions in the area of science. Two professions dominate our legislature: lawyers and academics or teachers. Perhaps for a law-making body that is keen to tell us how to behave this is not surprising. But are these people best equipped to make decisions about practical matters? At very minimum, their respective cultures of certainty or proof differ greatly from those of science. A close third to these two dogmatic professions comes the media and PR sector. Between them these three career backgrounds account for 40 per cent of our elected representatives in the House of Commons. Other, more practical backgrounds are consequently under-represented.

This is a distinct pattern from the population at large, where only 1 per cent are employed in ‘legal activities’ and another slightly over 1 per cent work in marketing, consultancy, and the media. So these two sectors, which represent only 2 per cent of the working population of Great Britain (Digest of Economic Statistics 1996: Table 6.2; UK figures unavailable), contribute 21 per cent of members of the House of Commons. In the UK population as a whole, 8.4 per cent of employees work in the educational sector, so, although this sector is over-represented amongst MPs, it is not as massively over-represented as the communications sectors. Under-represented sectors include particularly production and construction, which account for nearly 23 per cent of the UK workforce, while amongst MPs only 11 per cent have a background in industry as whole, whether white- or blue-collar.

Table 5. University education of all MPs: results by type of degree

Note: Hard sciences are physics, chemistry, mathematics, and medicine; soft sciences are computing and economics.

Table 5 shows that only 8.5 per cent of MPs have a degree in a ‘hard’ science subject; while another 11.2 per cent have a ‘soft science’ degree, this includes some social science subjects which are so soft as to have barely any technical content. Arts degrees are vastly over-represented: nearly half of MPs have one, with a further quarter of MPs having no degree at all. This leads us to question further how scientific advice on policy issues is generated and is converted by scientifically illiterate MPs into statute. In order to better illustrate the problem, we will next focus on the government Backbench Environment Committee and ask how it might consider scientific advice given to it in areas where the consequences of policy might result in hazard to human health.

2.3. The Commons Backbench Environment Committee

The Environment Committee is routinely required to consider information of a mathematical and scientific nature and to make decisions which could affect the health and well-being of very large numbers of people for a very long period of time. The environment is where we all live and where our children will have to live in the future. Technological developments regularly occur where possible subtle toxic or environmental effects of by-products released from a production process (dioxins, CFCs, radioactive discharges) or in some cases, subtle long-term toxic or mutational (cancer) effects resulting from exposure to the new agent itself (BSE, GM foods), may occur. The Environment Committee has to advise on draft legislation which ensures that progress in efficient use of resources is not blocked by irrational fears. We need to accept that history teaches that many of these fears are later discovered to be all to well grounded. Who do the committee listen to and how can they decide? We will first look at the make-up of the present Environment Committee (Table 6).

Table 6. The Commons Backbench Environment Committee

Source: Constituency offices; authors’ questionnaire; The Vacher Dod Guide to the New House of Commons (1997).

The information on qualifications was not easy to obtain. Several members of the committee objected to being questioned and one refused to answer. As far as we could ascertain on the basis of repeated phone calls, of the eleven members of the committee, not a single one has a degree in either physics, chemistry or biology. There are two with ‘A’ level maths and one with ‘A’level physics, none with post-O-level qualifications in chemistry or biological sciences. Examination of the MPs’ areas of expertise and qualifications enables us to see that those who were selected onto this committee could have been more scientifically qualified. We would not wish to disparage the contributions made by the members of the Environment Committee. But it is their job to scrutinise legislation in this area and presumably to take responsibility when errors occur. It is surely unfair to expect such responsibility to be taken in highly technical areas without adequate background knowledge.

What is the problem with this lack of scientific education? Surely we do not want our democratic representatives to be academic-oriented scientists or technocrats. But on the other hand, it is clear that without sufficient grounding in the basics of mathematics and/or science our politicians simply cannot judge the quality of the scientific information they are given. This makes them vulnerable to one-sided lobbying. It makes them accept information on the basis of the credibility of the person communicating it. In this world, Professor Plum must be more right than Dr Foster and no one listens to Mary Green. What should be done to ensure the best outcome? We will return to this, but at this stage we suggest that, as a minimum, the ‘other side’ of any issue should be presented at an early stage in the process. Whatever their scientific education we can assume that politicians are adept at reading human cues, since this is a fundamental requirement for success within a political structure. So the politician’s role might be to judge the validity and integrity of information presented. This would presuppose only a basic level of scientific knowledge on the part of the MP. But it raises new questions. Who is to provide the scientific information, and what is that information worth?

3. What is Good Science?

Science, as practised over the last 300 years, was a philosophical development which looked primarily to experiment and to the empirical data to provide evidence about truth. Its divergence from previous philosophical methods was fundamentally this: that beliefs began with objective consideration of the results of experiment rather than with statements about what seemed likely or about religious expectations or Holy Writ. In its basic form the method is based upon the principle of scientific induction. This states that there exists a form of inference by which laws can be inferred from particular facts, unequivocally.

The classical exposition of the inductive method (originally due to William of Occam) is as what are now called Mill’s Canons, the two most important of which are:

· The Canon of Agreement, which states that whatever there is in common between the antecedent conditions of a phenomenon can be supposed to be the cause or related to the cause of the phenomenon.

· The Canon of Difference, which states that the differences in the conditions under which an effect occurs and those under which it does not must be the cause or related to the cause of that effect.

In addition, the method relies upon the Principle of Accumulation, which states that scientific knowledge grows additively by the discovery of independent laws, and the Principle of Instance Confirmation, that the degree of belief in the truth of a law is proportional to the number of favourable instances of the law.

In addition to the inductive method outlined above, the scientific method includes the range of analytical methods subsumed within Popper’s Doctrine of Falsifiability. This regards science as moving forward through the experimental falsification of existing belief structures. Finally to the methods of inductive reasoning we must add considerations of Plausibility of Mechanism.

These are the methods of science (Mill, 1879; Popper, 1962; Harré, 1985; Papineau, 1996). Those who seek to apply these methods to the examination of a number of contemporary questions might be understandably confused. A good example, which we consider further below, is the question of increases in childhood leukemia associated with nuclear sites. According to all of the routines of science outlined above it should be now universally conceded that low-level radiation exposure to man-made radioactive substances released from nuclear plants like Sellafield, Dounreay and Cap de la Hague cause increases the risk of child leukemia. Application of Mill’s Canons all point to this. The Principle of Accumulation points to this. The Principle of Instance Confirmation is applicable. The continued assertion by the nuclear industry and by government that exposures are too low on the basis of studies of Hiroshima survivors have been (Popper) falsified by many researchers and through many studies. Yet the statutes which permit the releases and fresh leukemias to be induced continue in place. Why? The answer relates to science’s attitude towards evidence, significance, and truth.

4.1. What is evidence?

The dominant position taken by lawyers in our legislature becomes of specific relevance when we consider the question of ‘evidence’, because what lawyers and scientists think of as evidence is entirely different (this point was first made explicitly by Michael Mansfield QC). In a legal context, stating that there is no evidence about, say, a crime is to make a strong statement. For example, the sentence ‘There is no evidence that Mr Blair was present at the scene of the crime’ is to be taken as analogous to the statement ‘Mr Blair was not at the scene of the crime’. It is therefore a statement of his absence, not a mere acknowledgement of ignorance.

In science the position is entirely different. When a government scientist reports to Michael Meacher that ‘there is no evidence that genetically modified crops are damaging to health’, he is simply stating that none of the research studies conducted have found ‘significant’ answers to this question (see the following paragraph), or even that no studies have been carried out at all. If Meacher were to interpret this in the sense of legal evidence he might well take the statement as reassurance as to the negative consequences of genetic modification, rather than the statement of total ignorance it really represents. This important point indicates the crucial importance of the arts, and specifically legal, rather than scientific background of the majority of our decision-makers.

The nature of scientific evidence presented to government committees and thence to ministers is limited and biased in another way. Importantly, most of the scientific research carried out in universities today has been part funded by industry. As a result of the Thatcherite push towards ‘market-driven research’ only research programmes that can eventually yield a profit are likely to be funded. So what is the university scientist to do, if s/he undertakes a research project and finds results which would undermine the product of the very company that funded the research. Rather than stating specific cases (although the example of GMOs outlined in a later section is instructive), perhaps it would be more useful to illustrate this by means of a fictitious case-study.

Nuclear Futures Ltd. is interested in obtaining evidence that shows that there is a threshold of exposure to low levels of ionizing radiation below which there is no harmful effect. This ‘threshold’ hypothesis may follow from the discovery that damaged cells can repair themselves. Why, then, is it believed that low levels of radiation exposure may cause cancer? Camford University Radiation Biology department has a long history of doing funded work for Nuclear Futures Ltd (NFL). After some discussion, Dr Whizz is awarded a two-year contract to examine the hypothesis that there is a threshold for exposure dose below which there is no discernible effect. Dr Whizz and his team decide to irradiate cell cultures with alpha particles from Plutonium. In earlier experiments, he had found that there was no effect, below a certain quite high threshold. But this time, using new techniques which have become available, he follows the cells after exposure through several replications. His results suggest a threshold but it is one in which the effects are much larger at lower doses in the descendants of the irradiated cells. The results are not conclusive, however. What is he to do? Clearly, NFL are unlikely to fund further work which will result in their own demise. And this would result in no more contracts for his department. Maybe the work is quietly shelved. Maybe a preliminary paper is prepared for the International Journal of Radiation Biology but when it is sent, maybe it is returned by a referee who wants further confirmatory results before passing such a contentious and potentially explosive report. Maybe the report describes the results in the body text but the conclusions and abstract are at odds with the detailed result. This latter is becoming a favourite method for providing acceptable reports without actually telling lies, as we shall see below. Dr Whizz is a member of a government advisory committee. He is asked if the installations belonging to NFL may be responsible for increases in childhood cancer nearby. What does he reply? Does he exercise scientific caution? After all, his results were only preliminary. They required more work to confirm. No one has paid for this work.

In case this should seem far-fetched (and we can describe very similar cases involving low-level radiation exposure) we will turn to a recent study involving research into the possible harmful effects of mobile phone radiation. In a recent study, published in the journal Epidemiology, the results were reported of a study of 195,775 employees of the company Motorola, who develop and manufacture mobile phone equipment (Morgan et al., 2000). The study was part-funded by Motorola and carried out by a non-university organisation Exponent Health Group and examined all causes of mortality, with brain cancers, lymphomas and leukemias as major a priori outcomes of interest. The study seemed to report no excess risk from any cause of death among the workers. We may be suspicious that uch a study was funded with the express purpose of reassuring the public and those involved in legislation over mobile phone radiation exposure that there were no harmful effects. What are we to make of this? The raw data were not tabulated, nor made available for independent examination, so we have to assume that the processed results were accurately reduced from the data. Furthermore, no one has access to such data except the company. The abstract of the results states:

However, close inspection of the paper reveals a quite different picture. The study compares mortality risk in the highly educated, upper social class, electronics workers with members of the general public in four States of the USA., Arizona, Florida, Texas and Illinois. Comparison of the death rates reveals that although the Motorola employees enjoy lower death rates (owing to their higher socio-economic status), their death rates from all causes were significantly lower than their death rates from leukemia and lymphatic system cancers and most other cancers (but not brain cancers). The effect was particularly clear in the case of the lymphatic system cancer Hodgkin’s Disease, a result which was not mentioned in the Abstract. If the overall mortality risk from all causes is used as an internal control for the ‘healthy worker effect’ there was a higher risk of dying of most cancer types. And how do we know what the result would have been if Arizona and Texas were not used as external controls, but instead, Ohio and Tennessee were used? In fact, it was possible to use data reported in the paper to argue that there was a strong excess risk of most types of cancer in the workers, since standardised risk by period of employment consistently showed a 50 to 100 percent elevated risk in those who had worked more than 5 years with the radiation relative to those who had worked from 0-5 years. Nothing was made of this result.

There is a great deal of scope in epidemiology for tailoring the results to fit the hypothesis. Yet the results of this mobile phone paper, and others like it, reporting studies which were funded by industry and showing results which counter any suggestion that these industries may be causing harm, are commonplace in the reference section of review committees which give advice to government. Do the committee members look through the original papers? Would they know what to look for?

Aside from the biased research findings government experts do receive, there is also a limitation on what is available to them brought about by the system of peer review, a necessary hurdle for publication in an academic journal. Any researcher on the fringes of the scientific establishment, or whose findings fall outside the accepted scientific paradigm, might find him or herself asking ‘Who is my peer?’ A peer is defined as ‘a person of the same age, status, or ability’, which is unlikely to be the case when a young researcher submits a paper for review. But even more serious is the fact that the reviewers seem to assume that what should be the ‘same’ between reviewer and reviewed is their view of scientific truth. Many reviewers seem to see their role as the reinforcement of the view of the world they themselves hold. So alternative theories are unlikely to be published. When all accepted authorities ruled that the sun revolved around the earth, Galileo was unlikely to find his paper on ‘Let’s Twist Again: The Need for a Real Copernican Revolution’ accepted by Acta Catholica Astronomica. And the same situation persists today. The so-called peers who are allowed to judge what should be published may exercise their private theoretical prejudices at the expense of new ideas, and in the security provided by the system’s secrecy.

According to concerns expressed by the editors of some of Britain’s medical journals, the system of peer review is not effective even in weeding out fraudulent papers whose conclusions are based on bogus research. In the modern competitive academic environment, where researchers are desperate for advancement, evidence of fabrication and falsification of results is growing. Richard Horton, editor of The Lancet, is concerned that ‘the public was increasingly aware that fraud existed and there was a risk that the scientific community, in not acting, would be thought to be sweeping it under the carpet’ (Boseley, 1999: 8). The evidence accumulating about the quality of academic review suggests that the system of peer review is about the exercise of power and prestige within academia, not about a selfless search for scientific truth.

So when government ministers make claims about their decisions being made on ‘the best available scientific evidence’ we should be cautious. This will probably mean that it results from a consensus view of the state of scientific knowledge as agreed by the senior academics who operate the peer review system. It may even mean that their decisions are based on nothing more than arm-waving by their advisers, no genuine evidence being available to consider. Or, worse still, it may mean that decisions are being made on the basis of information provided only by the corporations whose interests are served by a decision in favour of the process or product under examination.

As a result of a recent controversy over the finding of excess breast cancer mortality near the Hinkley Point nuclear power station and accusation of bias towards the nuclear industry by university researchers, journalists from The Big Issue magazine asked the operator of the power station, BNFL, how much university research was being currently paid for by them. They replied that they were currently funding 29 research projects in English universities. The Big Issue revealed that this list excluded two research projects that they themselves had independently disovered but which were not on the list of 29 (The Big Issue South West, 24 April 2000).

Arpad Pusztai, the man at the Centre of the media frenzy over GM foods in late 1998, describes the threat he considers is posed to academic freedom by political and industry pressure in a recent article in The Ecologist. In response to his undertaking and reporting of objective scientific findings he was sacked and slandered, his reputation and scientific career destroyed. His employers abandoned and betrayed him once the political problems posed by his research became clear. But beyond this, he was also maligned by three committes which should be independent but which in this case appear to have been acting under political pressure: the House of Commons Science and Technology Committee, the Royal Society Committee on Toxicology, and the Advisory Committee on Novel Foods and Processes (of whom more below). Such a savage response to unpalatable scientific results not only succeeds in undermining the researcher himself, but also gives a clear signal to other scientists, who are quick to follow the cues about which results are acceptable and which not.

We may leave the drawing of the conclusions of this experience for the future of independent science to Pusztai himself:

It seems that, in the eyes of many senior scientists today, the future of science lies with industry. When scientists who apparently have no obvious financial connection with the biotech industry defend GM crops so blindly, and attack even the mildest critics, slandering their work and abilities in the process, we must ask ourselves what motivates them. And one possible motivation is that, with the rapid disappearance of the State patronage of science, many of these people are genuinely worred about the future funding of scientific research itself.

4.2. What is significant?

The word ‘significant’ is one that within the scientific community has a specific, technical meaning, but can also be interpreted generally by those without a scientific background. When a research finding is said to be ‘significant’ this means that it may be considered to be meaningful, in the sense of not being a chance finding. Since statistics is a methodology based on probability, it accepts a certain level of error as inevitable, meaning that some scientific findings that have passed the ‘significance’ test are still bound to be wrong.

The level of ‘significance’, which, of course, is directly related to the level of error, is chosen by the researcher, and should be set higher if the findings have more potentially dangerous implications. The level of significance generally adopted in scientific research is 5 per cent. This means that researchers are accepting a 5 per cent level of error, or that they will be wrong 1 in 20 times.

The procedure of testing whether results are ‘significant’ is known in the jargon as ‘hypothesis testing’. The scientist tests the ‘null’ hypothesis, which is the proposition that there is nothing unusual going on, or that the ‘distribution’ of results found does not differ from what would be expected by chance. This is illustrated in Figure 1 below, where the curve on the right represents the expected results and that on the left the results that were actually observed. The overlap labelled a represents the probability of making the error identified above, i.e. claiming that there is a finding when in fact that results came from the expected distribution, but a rather unusual part of it.

Figure 1. The Risk of a Type I and Type II error

Statistics defines two types of error that can be made when undertaking research. The first, known as a Type I error, is the one of most concern to scientists. It involves making a claim to have a research finding when in fact the results were generated by chance, and considerable egg on the face. An example might be a medical trial which indicated that a certain drug was effective in slowing the progress of AIDS. Later follow-up research might fail to find a similar result, suggesting that the original findings fell into the 5 per cent error area. For professional and credibility reasons, this is the kind of error most feared by a researcher: the error of claiming a significant result when in fact the finding resulted from chance.

But there is another type of error which is equally important, particularly in terms of potentially harmful consequences of industrial products or processes. This is the Type II error, defined as the failure to find a significant result when the hypothesis is in fact correct. The probability of this second kind of error is represented by the area b on the figure. It represents the risk of carrying out a trial and, for reasons which may relate to technical issues such as the size of the sample, failing to find a statistically significant result. It may not necessarily mean that the hypothesis is wrong, only that significance was not found this time. However, it may allow conclusions to be drawn, either to justify use of a technology or just because of extreme caution, that processes are not causing any ill effects when in fact they are.

4.3. What is scientific truth?

Although scientists may believe that science moves forward through the formal philosophical framework outlined earlier, reality seems rather more down-to-earth. In the last twenty years, sociologists have begun to direct their critical gaze at scientists and their real world. In the fields of sociology and social anthropology, fundamental questions about objectivity led, after the Second World War, to the examination of objectivity and the application of reflexive methods. We cannot escape from our culture, claimed the philosophers. What we appear to find when we look at other societies and cultures is largely a reflection of our own subjective view. And this interpretation is so embedded in the way we ourselves think about or understand the world that what we find is only our own interpretation, based on our own culture, of what we would be doing or thinking if we were the person being studied. Thus, what we find is essentially what we put there ourselves through our interpretative assumptions.

The early search for objectivity led to the belief that science was the most objective description of the physical world, particularly if the descriptions were mathematical. This was because it was believed that there were somehow ‘scientific facts’ wrested from Nature and elevated to the level of ‘physical laws’, like Newton’s Laws of Motion. However, recent close examination of scientists at work and study and of how their theories and discoveries come to eventually be accepted in their own and the wider community came to show that science is not as objective as it believes. ‘Science studies’, as this sociology has come to be known, finds that science is not free from the bias and inaccuracy which permeates all other areas of knowledge, and for the same reasons. Scientists are human beings like non-scientists. And scientific facts are not the unassailable result of forcing Nature to reveal her Truths, but are assembled from the interplay of many different items, actors, machines and procedures, all of which may be faulty, biased, inaccurate or uncertain.

The situation is outlined by the philosopher Bruno Latour, in two books, Science in Action (1987) and Pandora’s Hope (1999). Latour’s conclusions are very relevant to our enquiry. He finds that scientific truths are not unassailable, nor final, nor always without components derived from muddier sources than Nature herself. He also finds that what is accepted at any period of history is a scientific world-view that consists of a system of ‘black boxes’. These are accepted encapsulations of earlier theory that are then used as machines to understand and interpret new discoveries. Most significantly, he finds that as time passes and more knowledge is included in these ‘black boxes’ it becomes increasingly difficult for any scientists to open up or attack the complex system of connections that maintains the ‘black boxes’ or is within them. The current problem is that those who are building the present scientific consensus are those who are funded to do the research by those who have need of the results of this same research to make money. It is therefore quite reasonable to assume that this process leads to the construction of ‘black boxes’ which contain false reasoning, false connections and even false experimental results.

If Latour, and the ‘science study philosophers’ are correct, then what emerges is a need to question the very advice given to government by its expert committees. This is because it is not ‘the best scientific advice’. It is merely the contents of the particular black box which the particular scientific committee made up out of those particular experts they chose to believe. Are these experts unbiased? According to Latour, this is not a question about whether they are slipped buff envelopes containing fivers. It is an inevitable consequence of the fact, as we have pointed out, that they work in an area dominated by funding which is all tending towards increasing profit and research. In the following section we will consider some specific examples of how such advice was later shown to be in error, and resulted in people’s deaths.

5. When the Scientific Advice System Fails in Practice

5.1. Cancer in the offspring of nuclear workers

In the UK, the statutory advisory body in the area of radiation risk is the National Radiological Protection Board. This body has acted in such a questionable manner that the chairman of the Independent Advisory Committee on Sellafield and child leukemia Sir Douglas Black advised in 1984 that the government should set up a second ‘independent’ body to advise on radiation risk, the Committee on Medical Aspects of Radiation in the Environment, COMARE. As we shall see, this device too was to fail. NRPB were criticised for bias by the Royal Commission on the Environment (Flowers), by their own workers (Goss), and by NGOs. Undeterred, they continue to provide a rosy picture of the nuclear industry. A recent attempt to reinforce the perception that man-made radioactivity is harmful is found in their study of cancer in the offspring of nuclear workers. This report, was published independently and in a non-peer reviewed form but was also published in the British Medical Journal (Draper et al. 1997).

Close examination of the results revealed that the children of male nuclear workers were about twice as likely to suffer from leukemia than control children who lived nearby but who did not have fathers who were nuclear workers. For mothers who were nuclear workers the children had five times the chance of developing leukemia, although these results were based on a number of cases too small for statistics to consider it significant. The minimum response on the part of a responsible government to such a shocking finding might be to ban at least women, and possibly men, who might have children in the future from working in the industry. However, the conclusion of the reports was that radiation was not the cause of the increase in leukemia. This was because one of the findings was that the largest external doses were not associated with the highest risk. Implicit in this conclusion was a ‘black box’ belief that effect must be related to dose in a simple, straight-line manner. This itself is not justified since many results have shown that this is not so. But these results and these reports were not mentioned. Facts were selected to provide the conclusion that NRPB sought, and although the paper was criticised in the BMJ for this (Busby and Scott Cato, 1997), politicians continue to be reassured.

5.2. Childhood cancer and electrical power lines

This story illustrates the way in which research conclusions may be steered and spun to show the opposite of the result or something quite different, and it also shows how ‘great men’ are defined and then utilised to marginalise uncomfortable results.

Since 1992, in response to the increase in childhood cancer, the United Kingdom Co-ordinating Committee on Cancer Research (UKCCCR) has organised the United Kingdom Childhood Cancer Study (UKCCS), Chairman, Sir Richard Doll. Doll, an epidemiologist, became famous through his research in connection with the link between cigarette smoking and cancer. The UKCCS study collected information on 3,838 children with cancer and 7,629 control children without cancer throughout the UK. The operation, costing £11 million plus, has been funded jointly by ‘the UK’s leading cancer charities, government and industry’. ‘Government’ here includes the National Radiological Protection Board while ‘Industry’, we learn on looking closer, includes Westlakes (BNFL’s research wing), Scottish Nuclear, the National Grid Company and members of the electricity supply industry.

The study investigates five possible causes of childhood cancer, of which the first two were ionizing radiation and power-frequency electromagnetic fields. The first results to emerge in November 1999 related to the latter cause. The report in The Lancet, was entitled ‘Exposure to power-frequency magnetic fields and the risk of childhood cancer’ (UKCCCR, 1999a). A euphoric UKCCCR press release proclaimed ‘Major study finds no link between overhead power cables and childhood cancer’ (UKCCCR, 1999b). A news release from the Electricity Association echoed this exact headline and went on ‘So strong is this finding that Sir Richard Doll, the eminent scientist [etc.] . . . believes that there is now no justification for further epidemiological studies on EMF and childhood cancer in Britain’ (Electricity Association, 1999).

These conclusions spun to the media were wholly false. The Lancet article was solely about children exposed to magnetic fields in the home. The only result which threw any light upon the effect of overhead high-voltage power cables was an admission, tucked away in a table, that 31 cases and 17 controls lived near such power lines, suggesting that there was an approximate doubling of risk in those near the power lines. Many senior researchers, including those who had been involved in the study, were furious, pointing out that the press release was inaccurate. One of them allegedly admitted that someone had altered it from the agreed text.

The reason for the panic and spin-doctoring lay in another research report in the International Journal of Radiation Biology, published in the same week. In ‘Increased exposure to pollutant aerosols under high-voltage power lines’ (Henshaw et al., 1999) Professor Denis Henshaw and co-workers from Bristol University reported that they had measured significantly increased concentrations of radioactive dust in the vicinity of high-voltage power lines. Henshaw’s team placed their simple tastrak polycarbonate plastic alpha-particle detectors within 250m of the 275 and 400 kilovolt overhead power lines and found that the radiation dose to people was significantly increased. The high electric field was attracting the charged particles of dust which contained both radioactive substances and other pollutant molecules. Here, at last, was a plausible mechanism for the well-attested and widely confirmed association between power lines and childhood cancer. The discovery has obvious legal and financial implications for the electricity industry.

The UKCCCR line was swallowed whole. Power lines were reported safe. Fergus Walsh, BBC Health Correspondent, dutifully relayed this falsehood on the national TV news, deferentially interviewing Sir Richard Doll. The item had no balancing view and when we contacted Walsh after the programme and explained that the study had not even considered high-voltage power-line electric fields he did not seem to understand that there was a difference between electric and magnetic fields. When eventually he realised the nature of the complaint, his defence fell back on the eminence of the good Sir Richard.

This is a straightforward illustration of the pitfalls which exist in a search for truth in the area of scientific advice. Sir Richard Doll has been on many government advisory committees in the UK and the USA including, at present, an advisory committee on the risk of cancer following exposure to high voltage electric and magnetic fields. His interpretation of the results of the study above adds support to the widely held belief (Walker, 1999) that his advice may be biased or plain wrong. Nevertheless, his eminence and the many honours and awards he has been granted by grateful governments and powerful industrial groups are clearly able to influence the media into reporting what they may be unable to resolve themselves and which they are often told is false by people that they believe to be less credible.

The specific criticism concerning the drawing of irrelevant conclusions is also a common theme. In the age of information overload decision-makers have no time to read full reports and rely on reading only conclusions and executive summaries. These sections may be all that is read in political circles, while the meat of the report is scrutinised by other scientists. The two sections may even be written with different audiences in mind, so that they become disjointed, with the conclusions acting as a kind of comforting summary, and bearing little relation to the actual scientific findings which few journalists or politicians have time to read.

6 . The Scientific Advisory Committee: How Does it Work?

To explore how and why the system of science advice is failing we examined the work of three advisory committees in detail. The three committees have been evaluated in terms of their openness and accountability by Democratic Audit (Weir and Beetham, 1999: Table 8.5). The Novel Foods and Processes Committee (ACNFP) received a relatively favourable score, choosing to publish annual reports and registers its members interests, as well as carrying out public consultations and publishing its advice to government. However, the other two committees we consider—the Committee on the Medical Aspects of Radiation in the Environment (COMARE) and the BSE Committee (SEAC)—both score very poorly. COMARE’s only concession to accountability is to publish a register of members’ interests, while SEAC makes no attempt at all to open its work to public scrutiny.

6.1. SEAC: Mad cows and Englishmen

The story of BSE starts before the establishment of SEAC in 1990. It is a long and depressing narrative¾ currently under investigation by the Phillips Enquiry¾ demonstrating a frightening ignorance on the part of MPs and their ‘experts’, combined with an even more terrifying complacency. The result has already been 53 confirmed deaths, with an unknown number of UK citizens still incubating the human form of the disease. While the eventual Phillips Report will no doubt be damning, a more personal and readable version of this saga is offered in paperback form by Stephen Dealler, who, along with Richard Lacey and Helen Grant, is the hero of this sad tale. These three acted as ‘good’ scientists should¾ informing themselves, drawing conclusions based on the best available facts, and then alerting government to the risks. The response was ridicule, humiliation, and the destruction of careers. Richard Lacey was insulted by Lord Soames in the House of Commons and insultingly told to ‘keep taking the pills’. More seriously, his research unit was destroyed around him until he no longer had any staff to manage and was out of a job.

One quotation from Dealler’s book serves to illustrate the attitude towards science of the Backbench Agriculture Committee under its Chairman Gerry Wiggin, who represented an agricultural constituency in Somerset. The meeting took place at the House of Commons in June 1990 when Dealler, Lacey and Grant presented evidence. The following is an account of the reception given to Lacey’s evidence:

the questions asked seemed to be of minor significance¾ mostly about whether he had enough knowledge of the subject or whether he should speak about things that were really somebody else’s province. . . The chairman, Gerry Wiggin, was determined to hear what risks Lacey felt were possible in the short and longer term. Lacey told him directly that this sort of disease could become a major cause of death in Britain, and said we should take action to prevent such a calamity. Wiggin made it clear that he felt that scare stories got nobody anywhere. As I had feared, the committee did not know enough to ask the right questions of Lacey anyway. They did not appear to know that there was no method of testing anything for the presence of infection without actually inoculating it into the brain of a cow. They did not seem to realise that animals inherently lacked any way of forming immunity to the agents of TSEs, nor that the disease could not be destroyed by cooking.

In view of the information provided in the previous section about MPs’ level of scientific education it is not surprising that they could not ask useful questions. But what is shocking about this account is the arrogance and complacency shown by representatives in an area of critical importance to public health.

We can take no comfort from the fact that this was a few years ago and the hope that the new intake of MPs are better motivated. Recent evidence suggests that little has changed. In response to our scientific questionnaire, which focused on the BSE issue, one current Liberal Democrat MP representing a rural constituency replied as follows:

Presumably, the ‘constituents’ this MP is representing are only those with an interest in farming, not those who have to eat its deadly product.

6.2. COMARE: Something rotten in the state of Didcot

Comare (the Committee on the Medical Effects of Radiation in the Environment) was set up in the wake of the Black Report into the Sellafield leukemia cluster. Sir Douglas Black, who chaired the ‘Independent Advisory Committee’ was clearly unconvinced by arguments made by NRPB that the largest source of radioactivity in Europe could not be somehow associated with the discovery of a cluster of childhood leukemia when radiation was the only proven cause of the disease. Environmentalists and anti-nuclear activists had long argued that the government’s advisory committee on radiological protection¾ the National Radiological Protection Board¾ was not independent of the nuclear industry. In 1977, one of NRPB’s own senior researchers wrote in a letter to New Scientist, ‘The Royal Commission on Environmental Protection criticises the NRPB for bias towards underestimating radiation risk and of not being seen to be independent of the UKAEA’ (quoted in Busby 1995, p. 23).

So the purpose of the establishment of COMARE was for the government to have advice which enabled it to protect the public health from the dangers of radioisotopes that was independent of both the nuclear industry and the NRPB. The problem is that COMARE is based at the NRPB offices at Chilton, Oxfordshire. If you phone the COMARE telephone number it is answered with ‘Hello, NRPB’. What is more the three-person secretariat that organises COMARE’s agendas are all on the NRPB payroll. This point has been raised with various leading members of NRPB and the COMARE Chairman as being worrying but they cannot see a problem. When asked why it was necessary for COMARE to be based at NRPB one replied that it enables members to use the NRPB library. But none of the COMARE members themselves are based at Chilton, with the marginal exception of Eric Wright, who works at the MRC Radiation and Genome Stability Unit which is on an adjoining site.

COMARE’s three Secretaries: Roy Hamlet and John Cooper (Scientific) plus a currently empty post as Medical Secretary (formerly C. Sharp) are all NRPB employees who spend a proportion of their time conducting research for NRPB and the rest preparing paperwork for the members of COMARE, whose raison d’être is to be independent of NRPB. The Secretary we spoke to was unwilling to say how much time was spent with which hat on. We were recently amused to have copied to us a letter from the Department of Health in London advising an enquirer into radiation and health that the DoH took this area very seriously and were advised by the ‘independent committee COMARE’ on the effects of radiation. The letter was signed by Dr Roy Hamlet, who signed himself, Radiation Advisor, Department of Health. He does not have far to walk to get advice from the ‘independent committee’ since he is, himself, their secretary.

It seems apparent that in an area of such crucial importance as the medical effects of radiation in the environment it is wrong that there is no full-time member of staff even administering the Committee, never mind carrying out the independent research that all our lives depend on. Its meeting schedule of only four meetings a year also seems entirely inadequate for members to even discuss cursorily the vast number of papers published in this field around the world.

It is hard to form an idea of the independence or otherwise of the eighteen current committee members. They are appointed by the Chief Medical Officer; empty posts are not advertised which immediately raises questions about an old boys’ network. Luckily, because COMARE is a committee which we deal with on a fairly regular basis, we have been able to observe their behaviour fairly closely. This was recently helped when we were leaked a copy of the minutes of their 55th meeting, held on 18 March 1999, at which they discussed the report, written by Green Audit that there was increased cancer near those parts of the Irish Sea coast of Wales where radioisotopes from Sellafield had become concentrated. At this meeting, the authors of this report were not present, but their work was discussed in their absence following a presentation by the director of the Wales Cancer Intelligence Unit arguing that there was no effect and the data used by the Green Audit study was inaccurate. We are not concerned here with the truth or otherwise of this allegation, but direct attention to an exchange which took place in the course of the meeting.

7.13 Professor MacMillan asked whether it was possible to be sure that there was no coastline effect on the incidence of leukemia. Professor Clayton also thought it would be premature to say that the coastline effect did not exist.

The question here is, whether the alarming possibility of increasing Dr Busby’s credibility is more important than investigating what he claims to have found, even though this may save hundreds of lives, if he is right?

6.3. GMO Committee: natural or unnatural rape?

In the case of GM foods the plot thickens: on this issue the government seems to ignore even its own scientific advice. During the latest media GM feeding frenzy it emerged that a report attractively titled Investigation of feral oilseed rape populations: genetically modified organisms research report no 12 had been with the government for two years without being published. Was this something to do with the fact that its authors, the Scottish Crops Institute, had found that GM rape is much hardier than had been thought and could indeed cross-breed with non-GM varieties, undermining an important commercial crop? And how was this related to the fact that permission for farm-scale planting was given by ACRE (the Advisory Committee on Releases into the Environment) in the same month? This was, of course, unsurprising, since the Committee has never turned down an application for approval. It is painfully ironic that so much of the argument revolves around this crop ‘rape’, when many would suggest that it precisely what GM foods are doing to Nature herself.

How objective is science in the area of genetic modification? The lack of objectivity of Ray Baker, head of the Biotechnology and Biological Science Research Council, is evidence from his enthusiastic comment:

From 1985 to 1997, there have been 25,000 field trials of different GM crops of 60 different plants in 45 countries. . . We need to build even more confidence in this technology and it is vital to increase the size of those experiments that are going on.’

Given that this is the head of the research body overseeing science research in this field in the UK, we wouldn’t expect him to have already decided that the role of scientists was to ‘build even more confidence’, particularly when so little research on the technology has been carried out this far. Scientists can only possibly be in an initial exploratory phase of research as far as genetic modification is concerned. Academics who are being advised thus by their research council are the people from whom the members of the government’s advisory committee will inevitably be drawn.

According to the Independent on Sunday a majority of the members of the Advisory Committee on Novel Foods and Processes have links, either personally or institutionally, with the food industry. In an attempt to broaden the interest-base of the Committee it was agreed to appoint a ‘consumer representative’ last year. ‘Maff rejected a sceptic on GM foods from the Consumers’ Association in favour of the wife of a board member of the chemical company Boots’. The Committee’s independence is devastated by the information that it actually commissions no research of its own and relies mainly on information provided by biotech companies. The Committee’s most recent Annual Report (ACNFP, 1997) contains no references to peer-reviewed papers: there are a mere five references, 3 to papers by Zeneca (or whatever), the other 2 being publications from the Committee itself.

The quality of work carried out by the biotechnology companies themselves may well be highly dubious. Monsanto’s own research into the safety of its genetically engineered maize was criticised by ACRE for being based on ‘poor interpretation’ and far below required standards at their January meeting. The minutes record that ‘the molecular data submitted by the applicant did not support the conclusions regarding genomic organisation of the transgenes.’ Curiously, however, the research was re-presented and the application has already been approved.

6.4. Southwood, BSE and NRPB

In conclusion, a final example is now given which ties together low-level radiation and BSE, both hazards which are feared by ordinary people against the advice of scientists and scientific committees who have maintained that the public are acting emotionally, without logic, and irrationally in their fears.

The 1987 government advisory committee on BSE, the Spongiform Encephalopathy Advisory Committee, was chaired by Sir Richard Southwood. Despite receiving evidence to the contrary from eminent research scientists, the Southwood Committee reported to government that BSE was a disease of cows and could not cross the species barrier into humans. We now know that 53 people have died of what is now called human-variant CJD, caused by eating BSE-infected meat, and that Southwood was wrong. What is less well known is that at the time, and until 1995, Southwood was also Chairman of NRPB, advising government on the health effects of radiation. In 1986, when the Chernobyl reactor explosion occurred, large quantities of radioactive substances fell on parts of the UK, particularly Wales and Scotland. NRPB’s advice was that the levels of exposure were too low to have any measurable consequences, and in any case, the radiation would disperse in a month or so. They were soon proved wrong on the second count. Sheep from some upland areas of Wales are still banned from sale for consumption owing to their high radioactivity content nearly 15 years after the event.

However, they have now been proved wrong on the first count also. Infant leukemia increases have now been reported for those children who were in the womb at the time of the contamination in Greece, the USA, Germany and Scotland and Wales (see review in Busby and Scott Cato, 2000). This latter study was able to use the numbers of cases observed, together with the exposure doses and prediction models of the NRPB to show that their error in risk factor for this disease is upward of 100-fold. Such a figure is in the right order to ‘explain’ the Sellafield leukemia cluster and much else besides. In this instance we have incorrect scientific advice being believed because of the credibility of the advisor, but remarkably here the same advisor was responsible for two sets of advice from two different committees which both resulted in deaths in those affected by the advice because of bad decisions by those who they democratically elected to represent them.

Overall, there appears to be a common tone between the government advisory committees studied. The atmosphere would be more appropriate to a gentlemen’s club than a political office. Since the rewards appear so slight, one cannot help wondering why anybody would choose to serve on such committees. It has been suggested that the honour of serving is sought by academics, but as the academic world comes under more pressure to move towards business-level efficiency academics have ever less time to give away. The move towards a system of professional committees would seem to be an inevitable first recommendation of this investigation.

7. Conclusion: The Need to Establish an Oppositional Science System

The citizens of the UK have lost faith in the scientific establishment. According to a recent ICM poll public trust in scientists is now lower than their trust in policemen. Only 35 per cent of those questioned said they trusted scientists ‘a lot’; 54 per cent trusted them ‘a little’; and 12 per cent did not trust them at all. The only professions to come lower in the trust ranking were politicians and journalists. On specific issues levels of lack of trust rose as high as 49 per cent on the issue of cloning animals and 40 per cent on the issue of genetically modified food (Travis, 1999).

The intellectual arguments presented in this paper are unnecessary to explain why this loss of faith has occurred: the horrors of the real world ranging from defective medical treatments receiving scientific support to the destruction of the farming industry by pusillanimous government advisers have irreparably undermined faith in the ability of scientists to offer final and absolute judgements. The hilarity which greeted the pronouncements about the devastating potential health consequences of last year’s eclipse from the government’s chief medical officer indicates the level of esteem now accorded to these once august positions.

There seem to be two problems with the system of government scientific advice: appropriate evidence does not reach the expert committees; and the members of the committees are not in a position to independently assess the evidence and pass it on to decision-makers. These problems are deep-seated and require a complete overhaul.

The official publication on this issue, The Use of Scientific Advice in Policy-Making, is entirely inadequate. It contains only 5 pages of double-spaced text, plus a list of references only one of which is not a government publication. The paper consists of a series of legalistic principles, without any analysis or evidence of understanding of the complexity of the nature of scientific research and debate. Although it does contain some principles which may improve the system¾ including the stated willingness to address issues raised by lobby groups and the importance of including experts from non-scientific disciplines¾ it does not address the question of the restructuring of scientific committees, which seems so urgent.

Another apparently cosmetic change has been the introduction of two non-scientific advisory committees in the area where most hot potatoes have recently emerged, that of genetic engineering. The Human Genetics Commission and the Agriculture and Environment Biotechnology Commission will be responsible for the formulation and delivery of advice to ministers and should ensure that ‘both scientific and non-scientific views are brought to bear’. Environmental lobbyists watch these developments with sceptical interest.

Whatever the real influence of these new citizen committees the real scientific influence will remain with the scientific advisory committees and it is here that real changes are needed. Their existing structure has grown out of the post-war culture of paternalism and secrecy: hence the obsession with reassurance, as if the emergence of news that a degenerative brain disease could be caught from eating beef might cause public panic and anarchy on the streets (and why not?). The natural response within this culture was to cover up and reassure; the seeking of the scientific truth was a secondary consideration.

While it is easy to be disparaging about the paternalism of the British decision-maker the positive side of noblesse oblige was the requirement for decency and honour. This persists in the absurd notion that corporations who want to make vast profits by selling something are none the less likely to tell the truth to a government committee deciding whether to license it. While this faith in the honour of gentlemen is quaint and touching it is incompatible with the cut-throat globally competitive world these corporations operate in and therefore in itself represents a threat to public health.

Instead of placing trust in the scientist and the producer, as the existing structure of advice has done, we propose a structure where the corporation seeking permission for its product is characterised as driven by the profit motive, with less concern for the well-being of the citizen, represented by pressure groups. The old-fashioned view might have been of two naughty children arguing, with a benign father figure (democratic governance) wisely judging between their competing claims. Given the gross inequality of power and money between corporations and pressure groups this view is well out of date. The father has long since been judged senile and packed off to an old folks’ home. Meanwhile, back at home big brother has grown up greedy and over-sized, and finds no difficulty in abusing and exploiting his younger sibling.

To defend the weaker sibling we are suggesting a new structure which is based on the opposition principle so fundamental in the UK constitution. Just as in the House of Commons the government, as protagonist, is opposed axiomatically by the Opposition, so the corporations, as developers of new processes and products, should face opposition by government scientists on behalf of the citizen. While the problems associated with an oppositional system¾ especially its engendering of an antagonistic rather than cooperative polity¾ have been emphasised in recent years its main benefit has been ignored. This is the strengthening of legislation by means of a process of bombardment to identify any weaknesses. And just as in the House the most effective route to promotion