Jump to ContentJump to Main Navigation
Food, Risk and PoliticsScare, Scandal and Crisis – Insights Into the RiskPolitics of Food Safety$

Ed Randall

Print publication date: 2009

Print ISBN-13: 9780719072307

Published to University Press Scholarship Online: July 2012

DOI: 10.7228/manchester/9780719072307.001.0001

Show Summary Details
Page of
Subscriber: null; date: 21 June 2018

Dioxin – scares without borders

Dioxin – scares without borders

(p.114) 6 Dioxin – scares without borders
Food, Risk and Politics

Ed Randall

Manchester University Press

Abstract and Keywords

This chapter deals with the contamination of animal feed with dioxin and other persistent and harmful chemicals that can accumulate in the human body. The strength of the public's reaction to news that food supplies had been contaminated, the political consequences of that news, the amplification of public anxiety about exposure to contaminated food products within Belgium and around the world, and the willingness of governments and international authorities, outside of Belgium and on the basis of quite limited information, to criticise the Belgian authorities, makes the dioxin crisis a particularly instructive and interesting case study.

Keywords:   animal feed, dioxin, food contamination, Belgium

The Belgian scare and the Belgian crisis

No single food scare story better illustrates, at least during the period of a national election, the proposition that: ‘Risk management has become a dominant concern of public policy yet the ability of governments to anticipate the strength and focus of public concern remains weak’ (Breakwell and Barnett, 2003: 80). The Belgian dioxin food scandal of 1999 turned Belgian politics upside down. The political fallout from public suspicions of government duplicity came quickly and dramatically to the government of Jean-Luc Dehane, who seemed to be on course for re-election before the dioxin in food scandal hit Belgian society. If ever there was what the American academic Cass Sunstein has called a ‘social cascade’ in a liberal democracy, which changed that country's government in a matter of weeks, the dioxin scandal was it. It was a food scandal that appeared to validate many of the observations of psychological researchers about human risk perception and the psychometric paradigm that they had developed (see Slovic, 2000 for an account of the psychometric paradigm and its development).

Such things as probability neglect, dread and the availability heuristic all seemed to be at work in Belgium in May 1999. And, just as the Belgian national psyche appeared to be recovering from the scandal and its political aftermath in June 1999, schoolchildren from all over Belgium began to report feeling ill after having drunk bottled Coca-Cola. While the national government reacted by introducing controls on the sale of Coca-Cola products in Belgium, cases of ‘poisoning’ were also reported in Northern France. Laboratory tests and independent chemical analyses failed to establish any clear cause for these poisonings and Belgians were left to ponder the possibility – indeed the probability – of ‘mass sociogenic illness’ (Nemery, Fischler, Boogaerts et al., 1999). For all those interested in risk politics and food safety and committed to more publicly accountable risk regulation the Belgian experience of 1999 should give pause for thought.

(p.115) The food scare that hit Belgium in late May 1999 rapidly transmuted into a national and international political scandal of unprecedented seriousness for Belgian society and for its political leaders. Evidence that high dioxin levels had been found in hens and eggs was announced to the Belgian public and communicated to the European Commission by the Belgian Ministry of Agriculture. The official release of test results confirmed that high levels of persistent organic pollutants (POPs) had been found in Belgian agricultural products. It was rapidly followed by an official ban on the sale of chicken, meat and eggs. The Agriculture Ministry explained that these foods were likely to have been contaminated by dioxin and related chemicals known as PCBs (polychlorinated biphenyls). With just days to go before European and national elections the Belgian government, led by Prime Minister Jean-Luc Dehane, found itself under the most intense and critical scrutiny. It was confronted by an extremely anxious Belgian public made more anxious still by a rapidly mounting volume of allegations about official incompetence and government delays in acknowledging a major contamination episode affecting Belgian agriculture; intense press speculation about the scale of the threat to public health ratcheted up public alarm.

Two weeks later the Belgian Prime Minister, his party and its coalition partner were removed from office by the Belgian public and Mr Dehane's eight years in charge had come to an abrupt and unanticipated end. The political impact of one of Europe's most memorable late twentieth-century food scares is underlined by the fact that just days before the announcement that Belgian farm products were potentially hazardous to health Mr Dehane had a clear poll lead, and good reason to believe that his Christian-Democrat and Socialist coalition would be returned to office. It was a result he had every reason to expect and to view as a reward for his government's successful management of the Belgian economy.

This chapter sets out to address three broad questions, in light of the transformation in Christian-Democrat electoral prospects and the rapid rise in public alarm about dioxin contamination of food: What is known about the seriousness of the threat to human health presented by POPs such as dioxin? What made allegations and evidence of dioxin and PCBs in Belgian farm produce so politically and socially potent? How should students of risk politics and food safety view the whole dioxin in food episode and its significance for publicly accountable risk regulation of food safety in liberal and democratic societies, and the contemporary food economy?

Dioxins, PCBs and POPs

The term dioxin refers to many tens of different chemicals that remain in the environment and that can concentrate in animal tissue, including human tissue. (p.116) In a briefing note that it issued in 2001 the European Commission claimed that only 17 of the 210 different dioxin compounds were thought to be of toxicological concern (EC, 2001). Dioxins arise as a by-product of chemical processes – some of those processes can be described as natural, such as forest fires and volcanic eruptions, and others are industrial and associated with human activity. The latter processes include making paper, burning waste, manufacturing steel and the manufacture of many different chemicals for industrial and other uses. Dioxins that are produced naturally and as a result of human activity can appear throughout the natural environment. They can be found in soil, water and in the air. When they are absorbed and accumulate in body tissues, whether of humans or other creatures, they pose a variety of health risks and have been identified as carcinogens (cancer causing agents).

Contamination by PCBs (polychlorinated biphenyls), another group of compounds that are classified as persistent organic pollutants, is typically accompanied by dioxin contamination and accumulation in fatty tissues in particular. There is a great variety of sources of information about PCBs and dioxins including the United States Food and Drug Administration and the World Health Organization (WHO). Information has been steadily growing about how dioxins and PCBs enter the environment, accumulate in and are concentrated in animal tissues and enter the food chain. There is even hope that one of the most disturbing features of POPs, including dioxins, may be addressed by science and biotechnology. Early in January 2003 New Scientist and Nature reported on the work of Michael Bunge of the University of Halle-Wittenberg in Germany who had isolated a bacterium that appeared to be capable of digesting dioxins and overcoming their toxicity. Bunge and his fellow researchers have continued with his research and in 2007 they reported on Benzoate-driven de-halogenation of chlorinated ethenes in microbial cultures from a contaminated aquifer in the journal of Applied Microbiological Biotechnology (2007).

The European Commission, in a study published in 2001, estimated that 80% to 95% of human exposure to dioxin occurred through the ingestion of food, principally foods containing animal fats from meat, poultry and dairy products. While food safety authorities are also aware that industrial accidents and releases can increase human exposure it is the ingestion of food that serves as the principal vector by which dioxin gets into the human body. However, the problems of determining what is and is not safe, when it comes to dioxins and other POPs, are considerable. This is the principal reason why it has been especially difficult to reach international agreements and settle on a common view about target or safe levels of dioxin in the general environment. Despite this the United Nations Environment Program (UNEP) successfully called for negotiations to control POPs. In May 2001 the European Union, the USA and ninety other countries became signatories of a United Nations treaty known (p.117) as the Stockholm Convention (UNEP, 2001). The Convention contained an agreement to eliminate or reduce the production, use and release of twelve key POPs. It also specified procedures that would enable governments to develop further controls on POPs.

One critical aspect of the work needed to build a scientific consensus that is of use to regulatory authorities, is reaching agreement on the amounts of dioxins known to threaten health that can be permitted in the food chain. The World Health Organization has developed the notion of a ‘tolerable daily dose/intake’ (TDI, 1999).

However, public health authorities and individual consumers face enormous (arguably insuperable) problems in trying to regulate or control exposure to environmental and food-borne dioxins because they are ubiquitous in the environment and can enter the food chain at many different points. Determining background levels of dioxin contamination and distinguishing background levels from dioxins that are being added to the environment is immensely problematic but important in terms of fashioning effective regulatory action. For regulatory authorities the challenge is to develop a range of measurements that might help to build greater confidence about the quality and reliability of official knowledge about dioxin release and dioxin-related threats to health.

Buzby and Chandran, agricultural economists who have worked for the US Department of Agriculture's Economic Research Division, refer to one incident in July 1997 when a US agency found elevated levels of dioxin in chickens, eggs, farm-raised catfish and animal feed. They noted that after a lengthy investigation the source of the dioxin contamination was found to be animal feed that had been produced using ‘ball clay’ from a Mississippi mine. The clay had served as an anti-caking agent in soya bean meal. The manufacturers of the feed had no good reason to believe that its use was associated with the introduction of highly carcinogenic material into the human and animal food chain.

While governments may be told to move quickly to stop dioxin contamination they can only act if they can identify and remove sources of contamination. It is probable that in many circumstances regulatory authorities remain entirely unaware of the presence of contamination and lack the means to track down a particular source, and therefore to control or remove it. The European Union does now have regulatory limits for dioxin in food while the USA does not, pointing to great uncertainties about measurement of dioxins in explaining its stance. The fact that some countries require certification that food has not breached dioxin limits while others do not is an obvious source of friction over trade, and along with controls on GM is something that has found the EU and the US at loggerheads. It also needs to be recognised that any form of scientifically respectable certification is likely to add considerably to the cost of food processing and ultimately to the retail price of food. Estimates of the (p.118) success that public bodies have had in reducing dioxin exposure suggest there has been a decline in both the USA and the EU since the early 1980s but there are many gaps in official knowledge and uncertainties about sources of dioxins entering the environment. Amongst the greatest of uncertainties is the scale of the threat that dioxins represent to human health. It is possible to formulate very different pictures depending on the sources that are used.

POPS/dioxins: impact on health

Lok and Powell, based at the Department of Food Science at the University of Guelph in Canada, have observed that a ‘single high-dose exposure of laboratory animals to TCDD (2, 3, 7, 8-tetrachlorodibenzo-p-dioxin) ranging from less than 1 microgram to a few micrograms per kilogram of body weight is lethal’ (Lok and Powell, 2000: 5).

Nick van Larebeke and his colleagues, at the Department of Radiotherapy, Nuclear Medicine and Experimental Cancer at Ghent University in Belgium, in their paper on the Belgian PCB and Dioxin incident of January to June 1999, provide general estimates of cancer risks associated with PCBs and dioxins during the incident. They calculated a body burden (a measure of the concentration of potentially harmful chemicals or their breakdown products found in human tissue) incurred by populations exposed to PCBs and dioxins during the incident. Their conclusion, based on assumptions about average daily intake, Belgian diets and mean body weights, was that between 44 and 8,316 cancer deaths in the Belgian population of 10 million might ultimately be expected due to the Belgian population's exposure to ‘the combination of both dioxins and PCBs’ (van Larebeke, Hens, Schepens et al., 2001: 270).

The considerable range in their estimate of deaths reflects the great uncertainty they had to take into account in attempting to calculate daily intakes of POPs. When their paper was published in March 2001, they drew particular attention to those uncertainties; uncertainties which existed despite the fact that ‘more than 20,000 PCBs and hundreds of dioxin measurements on animal feed, animal fat and different food items had been performed’.

At the core of their difficulties was the problem of establishing the additional body burden attributable to the Belgian dioxin incident in the first half of 1999. As they explained ‘no measurements of body burdens before the crisis (were) available’ and the new measurements they could call on could not be used to answer key questions about how body burdens changed before, during and after exposure in the late winter and spring of 1999. Even if before and after measurements had been available it is unlikely that they would have offered a secure basis for making reliable estimates of additional body burdens (p.119) attributable to exposure during the incident. As Nick van Larebeke explains:

(considerable) uncertainty (exists) about the extent to which consumed food was contaminated, as sampling of animal fat and food items was not performed in a systematic way, but evolved rather haphazardly during the crisis in response to many different needs and pressures, some from national law European regulatory authorities, and others were commercial in nature. Some of these demands have biased sampling to the more suspect items, others still less suspect products. (2001: 271)

Without a systematic, routine and balanced system of monitoring, which measures contamination in foods and releases into the environment as well as human and animal body burdens, authoritative and comparable data will not be available to public authorities. If regulatory authorities lack authoritative data their regulatory activities are unlikely to carry conviction or to be easily and honestly explained. The work of Nick van Larebeke is arguably the most important and significant not only because of what it adds to our understanding of contamination during the Belgian dioxin crisis but because of what it reveals about the lack of reliable and comparable data, an obvious prerequisite for informed and intelligent policy-making.

What does not seem to be in doubt is that even while there is considerable variation in sensitivity to a single-dose exposure amongst mammals to dioxins such as TCDD, low-level long-term exposure causes cancers in the organs and tissues of laboratory animals such as hamsters, rats and mice. The precise mechanisms by which TCDD and other dioxins cause cancers remain a subject of scientific dispute. Disagreements about the most likely mechanisms have been reflected in differences of view about what can be considered a safe or acceptable daily dose of TCDD (if there is such a thing) and variable estimates of the likelihood of disease at different levels of exposure. Lok and Powell note that ‘controversy has resulted in a wide range of different acceptable doses among various countries and organisations from more 0.006 to over 20 picograms per kilogram of body weight per day’. It may be helpful to note that a picogram is defined as one trillionth of a gram and that the best estimate of the amount of dioxins entering the food chain during the Belgian food scare has been put at 1 gram. In 1998, when the World Health Organisation reviewed the TDI (Tolerable Daily Intake) for dioxins it estimated that the TDI should be set at 1 to 4 picograms per kilogram of body weight.

While the greatest concern about the threat to human health posed by dioxins has concentrated on their cancer inducing characteristics it should be noted that work with laboratory animals has also implicated dioxins in reproductive and developmental abnormalities. The impact of dioxin on the development of babies and children is a source of very great concern because of the body mass of an infant and the rapidity of its growth compared with an adult. POPs have (p.120) been closely associated with the disruption of mammalian biology and development. Dioxins are known to compromise the immune system's ability to fight off infections. Given the importance of the immune response to maintaining health it is likely that dioxins are implicated in a great many different health problems which can be traced back to a weak or compromised immune system. However, the reliance on animal test results to model impact on humans and the complexity of the toxic chemistry that humans encounter in the environment means that there is almost unlimited scope for disagreement about the precise and relative weight that should be given to dioxins and PCBs in accounting for human pathology. What is available to policy makers is epidemiological data that suggest that human populations known to have spent significant periods of time in environments with an abnormally high exposure to dioxin, may not only exhibit elevated cancer rates, but other symptomatology known to be associated with the exposure of laboratory animals to dioxins.

Perhaps the last word in this section should rest with Ropeik and Gray, the authors of Risk, a Practical Guide for Deciding What's Really Safe and What's Really Dangerous in the World Around You (2000). Both authors are based at the Harvard Center for Risk Analysis. On the subject of the approach that should be adopted to control of POPs Ropeik and Gray identify a split between European nations and others, including the United States. European nations, they observe, mostly support the application of the precautionary principle – crudely defined as the view that uncertainty should lead authorities to err on the side of caution. Their own national authorities in the US have taken the view that :

such a blanket approach ignores the large body of scientific evidence that shows that risks from many POPs is actually very low at the level to which most people are exposed. In the end [referring to the Stockholm Convention – there was], a compromise [which] advocated both precaution and [reliance on] scientific evidence. (Ropeik and Gray, 2002: 453)

The origins of the Belgian crisis and its development

The proximate cause of the Belgian dioxin crisis can be traced back to events that took place in January 1999. It is now believed that fat used in the manufacture of animal feed was contaminated by a gram of a carcinogenic dioxin and up to 50 kilograms of PCBs. Precisely how the fat, which was distributed to ten separate feed mills (eight in Belgium and two others in Holland and France) by a single supplier and animal feed producer, was contaminated in the first place remains uncertain. It seems likely that oil drained from electrical transformers had been mixed – whether deliberately or inadvertently has (p.121) never been established – with used cooking oil. The oil, collected and stored by Verkest, a processor and animal feed supplier, based near Ghent in northwestern Belgium, was almost certainly the unwitting recipient of contaminated material. Verkest, in turn, supplied other animal feed manufacturers with the raw materials they needed for the production of animal feed. The contamination therefore entered animal feed production in many different places even though it is believed to have originated from a single source.

A month later one of the feed manufacturers supplied by Verkest, Da Brabender, became concerned at reports from farms that it supplied with its feed that large numbers of mother hens had become sick, chicks were failing to thrive and eggs were failing to hatch. Da Brabender reported its concerns to its insurer at the beginning of March 1999. The firm considered that there was a possibility that feed it had supplied might be the cause of the problems reported to it by its customers. The firm's insurers in turn invited a veterinary specialist to assess the situation; the veterinarian, Dr Destickere, confirmed the suspicions that had led Da Brabender to contact their insurers in the first place. Da Brabender went on to request a laboratory analysis of the feed it had produced in January and alerted the Belgian Agriculture Ministry to the problems it had identified with animal feed utilising materials supplied by Verkest.

Laboratory analyses, completed at the end of April 1999, supported Dr Destickere's hypothesis, which was that dioxin contamination of animal feed was the cause of the veterinary problems he had found. Verkest had meanwhile been given notice that it was suspected of having broken regulations governing the manufacture and distribution of animal feed.

The Belgian government was then faced, at the end of May 1999, with a stream of allegations in the media suggesting that it had known for a month or more about the contamination of animal feed and the consequent hazard to human health but had failed to alert the public. Such a charge, on the eve of a general election, was extraordinarily difficult to refute or manage. Test results which became available on 26 May showed high levels of dioxin in eggs and hens and led the government to order controls on 417 poultry farms that were thought to have purchased feed made from the raw materials originally supplied by Verkest. Within twenty-four hours poultry and products using eggs were taken from the shelves of retailers. Measures to remove these products were followed a few days later by the introduction of controls on the slaughter and movement of cattle and pigs and the removal of meat products from retail outlets across Belgium.

In the meantime both the Belgian Public Health and Agriculture ministers had resigned from office in what the Belgian Cabinet described as a measure aimed at restoring public trust. However, both ministers were reported to believe that they had handled the discovery of contaminated animal feed and (p.122) farm products appropriately. Newspaper coverage of the resignations concluded that the resignations would do little or nothing to restore public confidence and the government's decision to suspend campaign activities in advance of the impending general election simply reinforced the impression that the government was losing control and failing to cope with the crisis.

The domestic crisis of confidence in Belgium was further exacerbated by reports of an extremely adverse international reaction to the worldwide media accounts of widespread contamination of Belgian produce. The European Commission and countries right around the world in receipt of agricultural exports from Belgium were also reported, in the Belgian press, as being intensely critical of the Belgian government. Bans on food imports from Belgium were quickly announced in at least thirty countries and the German agriculture minister claimed publicly that the problems Belgium and its trading partners had found themselves facing could have been solved if other EU member states had been told earlier about what was happening.

When Buzby and Chandran (2003) reviewed the Belgian dioxin crisis for the US Department of Agriculture Economic Research Service they noted (relying on the work of a group of Belgian scientists led by Nick van Larebeke (2001)) that:

  • Approximately 500 tons of animal feed was believed to have been contaminated out of a weekly average production of 28,000 tons.

  • The contaminated feed has been distributed to many different agricultural enterprises including poultry farms, rabbit, calf, pig and cow breeding farms mostly in Belgium but also in Germany, France and the Netherlands.

  • Some 746 pig farms, 445 poultry farms, 393 bovine farms, 237 dairy farms had used feed supplied by the producers whose production was known to have been contaminated.

They also went on to note that the US Food and Drug Administration (FDA) had issued advice to interdict a range of Belgian and European agricultural products so that their freedom from contamination and safety for human consumption could be scientifically established. The products covered initially included eggs, products containing eggs and game meats, animal feeds and animal feed ingredients as well as pet foods; the list was later expanded to include other products, such as soups and cheeses, believed to contain milk. The FDA advice remained in force until 2000.

The FDA was most certainly not alone and the range of products under suspicion, as well as the number of countries introducing controls on Belgian produce, rose dramatically in the first three weeks of June 1999 (Lok and Powell, 2000: 9). International reaction as well as domestic reaction in Belgium confirmed the picture of a stampede away from Belgian produce, even Belgian chocolates.

(p.123) ‘Belgian dioxin’ in Belgian food had quickly become big news. Food markets are international and the agricultural products of EU member states are transported over vast distances. This was certainly the case, most especially for Belgian dairy products. It was inevitable that a threat to public health associated with dietary staples such as eggs and chicken would receive enormous publicity and that the release of information, which invited questions about government failure and integrity, would be particularly closely scrutinised and widely reported. As Lok and Powell put it: ‘not only health, but reputation, money, political careers were on the line’ (15).

Events moved quickly; there was drama and an endless supply of opportunities to consider who was responsible and to apportion blame. The speed with which the crisis developed, and undeniable failures in government communication with the public, meant that the hunger for reliable and detailed information could not be satisfied. Speculation took over wherever and whenever official information appeared to be incomplete or suspect. Lok and Powell, no doubt accurately but somewhat sarcastically, point out that: ‘the media performed as expected during the crisis’ (15). After all, the Belgian public had been told that what had been released into the food chain was widely believed by experts to be a ‘potent carcinogen and, even at low levels, able to persist in the body for up to 10 years’ (15).

As with other food crises, scandals and scares – and there was no shortage in Europe of other food events – news providers were able to compare with the details of the current crisis. There was considerable media mileage in considering how serious dioxin contamination was compared with Britain's BSE/vCJD crisis and endless opportunities for prognosticating about the likely or eventual outcome of the crisis compared with earlier events. With BSE still in the minds of the European public and a plentiful supply of allegations about a new cover-up, the newsworthiness of the dioxin scandal was never in doubt.

Apart from a surfeit of opportunities to replay past scandals and crises, each day seemed to bring a new breaking story about dioxin contamination and its impact on government and society. There were food bans, new advice to consumers, adverse international reactions, ministerial resignations, food recalls and farm quarantines. There were also arrests and expert commentary and analysis on every conceivable aspects of the crisis. And there was a general election close at hand.

Attempts by the Belgian government to explain that it had acted responsibly by managing the release of information, so as to avoid panic and limit damage to Belgian agriculture, were met with incredulity. As Lok and Powell put it: ‘[The Belgian authorities] looked guilty from the beginning even though the contamination was not … the government's fault’ (16). Efforts to deny a cover-up were unconvincing because the possibility of a serious threat to public health (p.124) had been known about for at least a month before an official announcement was made. The desire to allocate blame appeared to drive coverage far more strongly than the desire to establish the extent of the risks to which the public had been exposed or determine the best course of action to limit that risk. The imminence of the general election made it imperative for ministers and their supporters to send out the message that they were in control and enjoying great success in bringing matters to a safe and satisfactory conclusion, when virtually all their partners – international partners in particular – were insisting that more time was needed and very many questions remained unanswered.

Lok and Powell are almost certainly correct in asserting that:

The dioxin contamination would have still been a damaging crisis had the government promptly gone public with the crisis. However, the damage and public scrutiny would have likely been focused more on the faults of the food industry rather than on the government's mismanagement of the crisis. (17)

The failure to level with the Belgian and international public was undoubtedly the single most important reason why the Belgian dioxin crisis assumed the proportions that it did. Early and complete disclosure is almost always likely to be the best policy and to evoke the most supportive (or least hostile) response from the general public. That is not the same as saying that panic and public anger can be avoided and allayed simply by letting it all hang out. But the prospects of remaining in charge and of being seen to be in charge are greatly enhanced if the political agenda and the news agenda can be set by members of the government rather than by leaks and speculation.

In food scandals, scares and crises European governments have learnt that if they do not provide the story others can very quickly be found who will. The assessments of what has gone on (and gone wrong) appear more or less immediately in popular media and are rarely a good guide to what will emerge later, when the significance of events that generated great public anxiety and media interest are considered a little more dispassionately.


Richard North's account of the Belgian dioxin crisis is written from the perspective of the Europhobe who is deeply antagonistic towards the European Union and the role of the European Commission in food safety and food policy. North looks upon the Belgian dioxin crisis as a handy excuse and ready justification for a power grab by the European Commission. The Commission is represented as opportunist and almost wholly unconcerned with the economic consequences, for farmers and the food economy, of an unnecessary Union-wide regulatory regime. It is a view of EU politics, risk politics and the drivers (p.125) of food safety in the European Union that has attracted many supporters. It is a view of the political dynamics of scares, scandals and food crises which largely overlooks the role of the media. It also fails to take account of problems affecting food safety that are deeply entrenched in the contemporary, increasingly internationalised and intensely competitive food industry; something that has been discussed in earlier case study chapters. There have also been dramatic alterations – affecting food safety – in the balance of power between retailers and farmers, food manufacturers and the companies that supply them with essential services.

North begins his account of the Belgian dioxin crisis with a critique of the European Commission and its proposals for the expansion of the Commission's responsibilities for food safety set out in Brussels at the start of 2000. The Commissioner responsible for food safety at the time was David Byrne; Byrne is described, by North, as unveiling a White Paper on Food Safety with the Belgian dioxin scare serving as a politically useful backcloth to his announcements. The White Paper contained more than eighty proposals, including a proposal for the establishment of a European Food Authority, and plans for the comprehensive reform of European food law. This was, from North's perspective, an extraordinary and almost entirely unjustified development of EU law; the extension of a legal system which had been advancing over many years to the point where it had become involved in the regulation of almost every aspect of the work of Europe's food industries and every part of the European food economy. The new European food safety regime would, as David Byrne was often heard to say, oversee the food economy all the way ‘from farm to fork’. North had no doubt that the Commissioner's power grab, on behalf of the Commission, carried with it ‘horrendous [cost implications that] threatened to infate the price hard-pressed farmers already had to pay for their feed, and by an astronomic degree’ (North, 2001: 152).

North was most angered by the claim that the plans set out in the Commission's White Paper were a great contribution to public health and safety. He was convinced that the opportunity costs of European legislation meant that the Commission's plans were ‘anything but’ a great contribution to public safety. Despite the claims of the Commissioner, that the best interests of both consumers and producers had been considered, North described the new regulatory regime as a ‘cynical and populist demonstration of the willingness of the EU to respond to “public concerns”’. The ability of the Commissioner to present himself as a knight in shining armour – arriving in the nick of time to protect the European consumer – had, in North's view, been greatly assisted by the Belgian dioxin crisis. It may even seem possible, to readers of North's account of a power-hungry European Commission, that the dioxin crisis had in some way been dreamt up to help launch its White Paper on food safety. What (p.126) North appeared to be quite certain about was that the Belgian dioxin crisis had been misrepresented and that that misrepresentation had been used to advance the Commission's agenda for controlling ever more of Europe's agriculture.

North gives his chapter about the Belgian dioxin crisis, in his Death of British Agriculture: The Wanton Destruction of their Key Industry (2001), a deliberately provocative title. The title does not disguise, nor is it intended to, his purpose and his perspective on risk regulation in Europe and on the significance of events in Belgium. The chapter dealing with the dioxin crisis is entitled ‘The Integrationalists’.

Integrationalist is North's portmanteau term for European political leaders and officials whom he believes will seize almost any opportunity to extend the reach of the European Commission. In the case of the Belgian dioxin crisis the Commission is represented as an ambulance chaser of a particularly revolting kind; it harbours an ambition, to unify and harmonise the regulation of European agriculture, which has proven immensely destructive in almost every member state of the European Union and, from North's point of view, most importantly in Britain.

Some big questions about food scandals, scares and crises and risk politics

The events surrounding the Belgian dioxin crisis have already been presented at some length and the threat that dioxin and PCBs represent to human health has also been discussed earlier in this chapter. While North's account of the crisis and its aftermath directs attention away from Belgium and to an international stage it leaves many important aspects of the crisis largely untouched. The dynamics of the Belgian dioxin crisis, and what North views as the political exploitation of the crisis, raises very important questions for students of risk politics and food safety; questions which, it is claimed in this book, require a much deeper examination than North supplies.

Here are some of those questions:

  • Is it likely that food scares, scandals and crises will, as North suggests, simply serve as an opportunity for powerful established interests to define the public purpose in a way that serves them while ignoring more fundamental goals?

  • Do food crises, scandals and scares serve to concentrate the mind of policy makers and the general public in a way that facilitates improvements in the organisation and effectiveness of regulatory systems and food safety policy? Or, as has already been suggested in this book, do food scares, scandals and crises have only a modest impact on food safety itself and frequently fail to direct public attention to aspects of the food economy which ought to be (p.127) much more closely scrutinised?

  • Is it possible that food scares, scandals and crises, which might be expected to help strengthen public involvement and engagement with risk politics and food safety, have relatively little impact on the quality of public engagement in public-policy making about such things as food safety and even serve to blind us to what is possible?

  • Is it necessary to understand much more of the dynamics of food scares, scandals and crises in order to reveal the part that media representations of food dangers play in focusing public attention on them, and does the focus encouraged by media coverage lead to the neglect of important aspects of risk politics and the modern food economy, which – as this author believes – help to shape risk politics and food safety?

Media power, the social amplification of risk and risk politics

There can be no doubt that by making his criticisms of the European Union – and the role of the European Commission in particular – North considered that he was directing his reader's attention to economic and political realities that were shaping the European food economy; realities that enhanced the power of the European Union at the expense of farmers and others, most of whom had a much better grasp of agriculture than officials in Brussels.

An alternative view, presented here, is that the most worrying aspects of risk politics and the modern food economy have little to do with the regulatory reach of the European Union or its civil service based in the European Commission. Rather students of risk politics and food safety need to turn their attention to the role of the mass media in contemporary risk politics and food safety issues.

Risk perception and media influences have been closely studied for several decades by academic researchers such as Nick Pidgeon, Roger E. Kasperson and Paul Slovic. Their research has been conducted throughout a period when profound changes have been taking place in the structure of the media and of the food industries. It is the combination of a much more competitive and concentrated food industry, especially in the retail food sector, and heightened public sensitivity to food safety issues that has made it essential, if risk management is to be handled more intelligently and democratically, that our shared understanding of the dynamics of food politics and public opinion is improved. It is clear that the ways in which stories about food-borne hazards and public safety are conveyed by the popular media and our responses to those stories have taken on a very special significance for policy makers as well as the public they are meant to serve.

(p.128) How volatile is public opinion? How responsive is it to media accounts of risk and to the most scientific accounts of risk? To what extent is the political process buffeted by media coverage and shifts in public opinion in response to media coverage? Does coverage of risk events and the public response to that coverage weaken the capacity of our public institutions to make and operate regulatory regimes that will stand the test of time and command public respect?

It is surely necessary for all of those who are stakeholders in food safety – not just public-policy makers but also the general public (the biggest and most important stakeholder of all), to understand just how easily risk politics and food safety regulation can be blown off course. The Belgian dioxin crisis exemplifies just how quickly things can change and be changed, and how easily political choices, even the choice of a government, can be altered by a spasm of public anxiety, fuelled by media coverage of a food scare.

Media drivers, values and imperatives

The kinds of things which excite public interest and generate media interest and coverage have been closely studied by academic researchers, such as Peter M. Sandman. Sandman's account of the factors that most strongly influence the selection and presentation of environmental risks appears to translate well to the coverage of other risks, including food risks. His work is therefore an excellent starting place for anyone who wants to understand how it is possible for an event, which does not appear to be wholly exceptional, to become a cause célèbre and a catalyst for a change of government, as happened with the Belgian dioxin crisis.

Sandman's insights into the media coverage of environmental risks are also a good starting point for students of risk politics who want to move on to consider the framework developed by social psychologists such as Roger and Jeanne Kasperson. Their Social Amplification of Risk Framework (SARF) is intended to facilitate the study of the interaction between individual psychology and risk perception, the activities and role of the media, and institutional and political arrangements for communicating and managing risk. Its relevance – to all the case studies in this book – should be clear by the time readers have finished the book.

The Kaspersons – and their many colleagues – opened a door, through which others are now able to pass; all those who want to join them in considering such things as social cascades and probability neglect and the ways in which the media shapes public risk perception and public/safety policy agendas. The formation of currents in public opinion that miss out or elide important parts (p.129) of a complex risk picture, and the very existence of such things as moral panics, explicable in terms of SARF, may have great significance for contemporary risk politics and risk regulation policy making.

SARF in Sunstein's hands

The careful examination of the relationship between substantial shifts in public awareness of risk and changes in the environment in which risk regulation is managed by public-policy makers, has been a particular interest of the American academic, Cass Sunstein. Sunstein's discussion of the need to limit the impact of social cascades – what he refers to at one point in his work as ‘fear as wildfire’ – on public risk management policy will be a major theme of the Conclusion to this book and will be introduced below in an attempt to set out a more balanced account of food safety and risk regulation in Europe than has been provided by Richard North and others, whose closeness to agricultural interests may well render them victims of a very human tendency to focus on some of the most emotionally engaging parts of the risk picture.

The aim, of Sunstein and others, is to begin to supply an account of risk politics that has greater endurance and wider relevance to the dilemmas that risk-aware policy makers and a risk-aware public need to address – and to address, in this author's opinion at least – together.

Complexity, media theory and risk perception, policy and politics

There would be little point in introducing what is undoubtedly a more complex and in many ways much less immediately satisfying account of the political and social dynamics of public risk perception, in relation to food safety regulation, than that supplied by Richard North if this did not, at the same time, form part of an examination of the relationship between risk, food policy and the politics of the contemporary food economy. Much greater public understanding of the relationships between individual risk perception, the development of currents in public opinion, their connection with media coverage and with public decision making are, this author believes, prerequisites for a more mature and deliberative – in short – democratic risk politics. Understanding the dynamics of public opinion about risk is a vital step towards a thoughtful realistic appraisal of the prospects for food safety policies and for making judgements which command widespread public support and that can, at the same time, be presented as the outcome of a political process that is open as well as reasoned and reasonable.

It is, therefore, to one of the sources most often cited of unreason and un- (p.130) reasonableness in modern politics that I now turn: the role of the news values found in the mass media, most particularly the popular press, in determining what risks gets covered and how.

Sandman on media attitudes to covering risk

Peter M. Sandman, joint author of Environmental Risk and the Press (1987) and ‘Mass Media and Environmental Risk: Seven Principles’ (which can be found in Roger Bate's What Risk? (1999)), sets out to answer a question found on many different lips and a question he is keen to pose to himself: Does media coverage exaggerate environmental risk? Sandman follows that question with another: What can be done, via the media, to put risk into context? A question that is considered from a number of different perspectives in the extended discussion of popular risk perception and public participation in risk management, assessment and communication that readers will find in the Conclusion to this book.

In answer to the first question Sandman makes a series of points that are unlikely to provoke great disagreements with regular readers of the British press and that are useful in accounting for the difficulties that anyone who relies on the mass media alone is likely to experience in making balanced assessments of risk, including food-related risks.

Sandman, who is internationally recognised as a leading academic exponent and practitioner of content analysis, has come to the conclusion (not a surprising conclusion) that the amount of coverage accorded to an environmental risk is often a poor guide to the seriousness of the risk in terms of its likely impact on human health. Rather, ‘traditional journalistic criteria such as timeliness and human interest’ exert the strongest influence on what is reported and on how it is reported. When it comes to risk, most stories found in media reports focus on ‘blame’, ‘fear’, ‘anger’ and non-technical issues that have to do with what he labels ‘outrage’ (as opposed to ‘hazard’ – a term which implies the precise measurement of harm or seriousness). Sandman, who has also studied the impact of environmental risk stories on readers, goes on to assert that when ‘technical information about risk is provided in news stories [it generally] has little if any impact on the audience’.

Other characteristics of media risk coverage make it clear that ‘alarming content is more common than reassuring content’, aside from crisis situations ‘when an impulse to prevent panic’ may come into play and contain what is reported and how it is reported. The impulse to exercise restraint, in exceptional cases, may be a product of editorial and commercial influences and what Sandman describes as a heavy reliance on official sources (Sandman, 1999: 276).

Sandman's knowledgeable conclusion is that media coverage of risk is more (p.131) strongly connected with locality, immediacy and visual appeal than it is with seriousness, especially when seriousness is defined in relation to long-term impact (such things as morbidity and mortality considered over years and decades rather than hours, days and weeks). While Sandman believes that journalists have a clear preference for writing or broadcasting stories that are more likely to alarm audiences than reassure them he insists that is not the result of a wish to be sensationalist for its own sake. It is a reflection of what he calls ‘news judgements’. In the newsroom ‘missing an issue [a good story] is a much greater journalistic sin than overstating one’. Not to put too fine a point on it: the most important news stories recommend themselves in an uncomplicated way.

Who, assuming any of us were left in charge of a news room or an editor's desk, would want to publish or broadcast a story prefaced with the observation that: ‘we don't want to bring you the full story yet, we will only be able to bring you the full story in a week or a month's time, after we have done all the necessary research’? Even when reporters have taken what they believe is great care to present a story in a way that they consider balanced Sandman explains why there may be a gap, sometimes a very large gap, between the reporter's and reader's perceptions of what is and what is not truly alarming. Sandman suggests that reporters of environmental risk may believe in putting out a story about a subject such as chemical contamination, in which they explain that the threat to public health is very small, that they are behaving responsibly. It is likely that the readership includes many people who find it hard to interpret information about low levels of risk included in stories about dangers and hazards; such readers may be inclined to judge that there is a serious problem because a danger is known to be present in their locality and has been named. The simple mention of dioxin as a health hazard, is what may make readers anxious, rather than details about the degree of contamination or the number of parts per trillion of the contaminant that are likely to be ingested. Given public familiarity with notions of newsworthiness – there is after all no such thing as good news – it is even possible that concerted attempts to reassure in news items covering risk and risky events may arouse suspicion in an audience that believes it understands very well what does and does not get into the news.

Sandman is well aware that reporters have a job to do and a limited amount of time in which to do it. They rarely have the expertise to guide them through conflicting testimony about the nature and extent of the risk events they have been asked to cover. Reporters do, however, have a clear idea about what it is they want from their sources. They want to be able to draw on a source that they can claim is authoritative and, if they are dealing with a major risk story, they want to be able to say that they have reported it in a fair and balanced way.

What Sandman calls basic journalistic and editorial requirements have, in (p.132) his view, a big impact on how risks are covered. They help to explain a bias in favour of official sources and, in stories that are judged to be more complicated and important, engage in what he refers to as the ‘scavenger hunt’. The scavenger hunt is often a frantic search for a balancing opinion. This is a part of what he calls the ‘epistemology of routine journalism’.

… there is no truth (or at least no way to determine truth); there are only conflict-ing claims, to be covered as fairly as possible … [this means] tossing the hot potato of truth into the lap of the audience … a general assignment reporter on a breaking story just wants to get somebody to say how bad the situation is and somebody else to say it isn't so bad … [before moving] on to another story. (1999: 281).

Investigative reporters who want to dig deeper and build their reputation as reporters are likely, in Sandman's view, to find themselves ‘naturally more allied with their alarming sources than with their reassuring ones’ (282). It is much easier, especially in dealing with complicated subjects, to polarise and sharply distinguish points of view. That generally has two advantages over having to work harder still to find a middle ground: it helps to make what is going on easier to explain and it also assists in making a story interesting to a general audience.

The very business of making a story easier to understand, and more interesting to read, tends to introduce a bias in reporting that favours opinions over data, narrative over such things as charts and tables, and plain speaking over abstract language and terminology. Sandman concludes that the journalist's motto, when dealing with risk, is: ‘Above all focus on outrage’ (283).

None of this is meant to put news organisations or the reporters working for them in the dock. Most news organisations depend, for the revenue that keeps them in business, on advertising; advertising revenues depend on readership/audience size and this in turn depends on keeping or building audience share. Stories that are easy to understand and that engage readers’, listeners’ and viewers’ emotions have an obvious and uncomplicated attraction for most editors and reporters.

The social amplification of risk framework

John Eldridge and Jacquie Reilly are well aware of the characteristics of the media coverage described above. Nevertheless, they argue that many accounts of the place of risk in contemporary society are in danger of over-simplifying the media's role. Sandman's account appears far from simple – but Eldridge and Reilly describe a process that is more complicated still. In their account, ‘BSE in the British Media’ (Eldridge and Reilly, 2003: 138–155), they acknowledge that ‘media interest is rarely maintained in the face of ongoing scientific uncertainty’ (p.133) and recognise that even though ‘pressure groups can use the media to force an issue onto the public agenda’, this is no guarantee that the media can be ‘assumed to be automatic allies in the “democratisation of risk”’.

The ‘democratisation of risk’ is a far more demanding matter. Sustained consideration of risk, which raises and addresses questions about what is expected to happen well into the future is, in practice, difficult for news organisations to deal with as their ‘basic news principle [is one that] emphasises events of the day’ (140). Nevertheless, the idea that large numbers of people who are strongly influenced by media coverage of risks are also capable of recognising the limitations of modern news reporting, is not particularly revolutionary. Yet it is an idea that all those who entertain the possibility of making risk politics more democratic, better informed and more engaging must surely cling to. A better informed appreciation of the arbitrariness and unreliability of media accounts of risk is likely to be an important precondition for increased and better informed public participation in risk politics. Insights into risk perception and media influence, gained from social and psychological research, have contributed greatly to the development of the Social Amplification of Risk Framework (SARF).

SARF is likely to appeal to media savvy citizens, not just to academics, and it is the intellectual offspring of Jeanne Kasperson, Roger Kasperson, Nick Pidgeon, Paul Slovic and others:

[It] arose out of an attempt to overcome the fragmented nature of risk perception and risk communication research by developing an integrated theoretical framework capable of accounting for findings from a wide range of studies, including: from media research; from the psychometric and cultural schools of risk perception research; and from studies of organisational responses to risk. (Pidgeon, Kasperson and Slovic, 2003: 13)

The framework was designed to help all of those who are interested in the relationship between individual risk perception, media, politics and the formation of public opinion, to consider how the way in which risk events are communicated affects not only how they are perceived, but also how great an impact they have on society generally and on public policy in particular. The idea that different risk events – a plane crash or a release of dioxin that contaminates the whole of the food chain – are mediated by culture, a variety of different communication channels and social and political institutions, and that public concern can be stepped up or down (amplified or attenuated) in the process was intended to empower social researchers and liberate them from the shackles of any single disciplinary view of human relations and popular communication.

The Kaspersons and their colleagues regularly employed the analogy of a stone dropped into a pond, to explain the way in which what may seem to be (p.134) isolated events can send out ripples that have an impact at some distance. If we change the analogy, exchanging the pond for a complex piece of electrical equipment, one that connects thousands, even millions, of other pieces of electrical equipment, and that forms a network, then SARF can be compared with the study of an electrical system in which signals can gain and lose strength, and coherence and signal strength can be altered, stepped up and down, at a great many different points.

When millions of different social actors are part of a great network that influences how they communicate with one another, and what they communicate about, we have a model of risk communication that is able to integrate many different elements, from individuals and the groups to which they belong, to the organisations and institutions that own and control important parts of the communication network through which information and opinions are transmitted.

Much of the work on SARF has concentrated on the amplification rather than on the social attenuation of risk communication. It has concentrated on what have been referred to as moral panics (Cohen, 1972) and social cascades (Sunstein, 2005). The Kaspersons and their colleagues were searching, as John Eldridge and Jacquie Reilly have explained, for a conceptual framework ‘capable of integrating the technical analysis of risk and the cultural, social and individual response structures that shape the public's experience of risk’ (Slovic, 2000: 234). While much remains to be done in developing SARF it has helped social and psychological researchers investigate the ways in which some risk events trigger what appear to be disproportionate public reactions and others seem to disappear without trace. Slovic has observed that:

apparently minor risks or risk events as assessed by technical experts, sometimes produced massive public reactions, accompanied by substantial social and economic impacts and sometimes even by subsequently increased physical risks … (2000: 233)

SARF in perspective

Eldridge and Reilly, in their contribution to the volume Social Amplification of Risk, in a chapter dealing with ‘BSE and the British Media’, report that they were able to frame much, though not all, of their research into media coverage of BSE and its impact on public opinion using the SARF. They were, nevertheless, left with some very interesting and important questions about the limitations and possible weaknesses of the framework. They say that they resisted the proposition, which may attract some users of SARF, that there is a way of identifying the particular risks that emerge from the studios of the most widely watched television stations or that are expressed in the columns of the most (p.135) widely read newspapers, that trump all other accounts of risk, including those put forward by experts and policy makers whose concern with hard data rarely seems to be a match for the opinions and feelings of the public at large, or the most convincing media voices that they hear most often. At one point in their article on the British media and BSE, Eldridge and Reilly appeal to the wisdom and insight of the sociologist Robert Merton. Merton wrote:

Total subjectivism leads us astray by failing to provide a theoretical place for systematic concern with the objective constraints upon human action. The social, demographic, economic, technological, ecological and other constraints are not all caught up in social definitions. To ignore these constraints is mistakenly to imply that they do not significantly affect both the choices people make and the personal and social consequences of those choices. (Merton, 1976: 176)

Students of risk politics and food safety who are persuaded that we all – or almost all – get carried away in social cascades from time to time, or are drawn into moral panics, are understandably reluctant to treat the product of complex interactions over riskiness, between the watching and reading public and the media, as a sound basis for political judgements about risk. Nevertheless, we may be emboldened by what Merton had to say about objectivity and subjectivity in human perceptions and choices, and by what has emerged from studies influenced and shaped by SARF, to invite policy makers and the public alike to reflect on the ephemeral nature of the media-influenced communication and assessments of risk, which typically depend upon incomplete and often seriously unbalanced accounts of risk. Reflections on the fickleness of media-driven risk awareness may even assist us in presenting the case for forms of communication and public participation that encourage greater deliberation and that make us less prone, than reports appearing in the mass media do, to accepting instant judgements.

Academic researchers and others who have been attracted by SARF, and drawn upon it in their work on the formation of public opinion and popular risk perception, may find that they have a role that extends beyond that of observers and commentators. The ability to reflect on how the media influences opinions and behaviour is one that can be fostered and shared beyond communities of academic researches. Just as cinema audiences have learnt to understand the manipulations of flm makers, the general public may have a capacity, that has often been underestimated, for reflecting upon the role that SARF, though it may not be known by such an acronym, plays in risk politics, risk regulation and food safety. Something that could be characterised as a necessary, though by no means sufficient, condition for a more sophisticated and participative risk politics.

(p.136) Sunstein on fearing the laws of fear in public-policy making

Some academic commentators are less hopeful than this author about the popular capacity for reflection and balanced judgement. They fear public vulnerability to the media distortion and suspect that there is a general inability, outside of expert and highly risk-aware communities, to make balanced judgements about risk regulation. One academic, more than any other, has succeeded in communicating how the insights derived from the psychometric paradigm and those contained in SARF, should make us wary of attempting to extend and deepen public involvement in risk politics. Cass Sunstein, who gave a series of lectures in Cambridge in the spring of 2004, brilliantly expounded the case for recognising what he labelled the ‘laws of fear’ and for keeping much of public policy and risk regulation at a safe distance from the general public.

Sunstein's reasons for wanting to maintain a very substantial domain in which public policy is made deliberatively, and at arm's length from the general public, can be explained relatively straightforwardly. In the fourth chapter of the book based on his Cambridge lectures, The Laws of Fear (2005), he explores what he refers to as ‘fear as wildfire’.

Sunstein has no quarrel with the many social scientists, especially social psychologists such as Paul Slovic and anthropologists such as Mary Douglas, who have argued that the ways in which we perceive risk and respond are powerfully influenced by our social situation and our culture. The incomplete and culturally conditioned picture that most of us have of risks and of our vulnerability to social influences alarms him. Sunstein makes his point, in telling fashion, by recounting the extraordinary levels of fear and the extraordinary behaviour that followed media coverage of murders in the Washington DC area in October 2002.

The Washington murders, attributed to a sniper or snipers targetting drivers refuelling their cars at petrol stations, resulted in ten deaths and three critical injuries in just three weeks. There were frequent, and what Sunstein takes to be reliable, reports of people living in the Washington area driving hundreds of miles to avoid having to use petrol pumps in Washington DC. In Sunstein's eyes this provides a classic illustration of what is known as probability neglect. The very remote likelihood of being shot by a sniper was causing Washingtonians to accept a substantially greater risk of death and injury by getting into their cars and travelling great distances. The psychological impact of deaths caused by what turned out to be two snipers could be attributed, in part, to the availability heuristic. Our tendency to focus on events that are literally the most available in judging what is most dangerous: events which we can bring to mind most easily, because we have an immediate, dramatic and emotionally potent knowledge of them.

(p.137) Sunstein was not convinced, however, that either probability neglect or the availability heuristic could explain just how greatly the sniper shootings in Washington in 2002 had affected the behaviour and outlook of Washingtonians. There was something more and he believed that he understood what it was, a social phenomenon widely reported on and understood by social scientists.

Whether it was social scientists writing about moral panics or about social cascades it was clear – to Sunstein – that individual and cultural predispositions were a powerful influence on which risks were judged to be most pressing and salient, and which risks could be most easily overlooked. Culture and individual disposition, Sunstein argued, helped to explain a striking difference between citizens of the US and citizens of France. While the French worried about genetically modified crops, which did not appear to be a major concern for most US citizens, US citizens appeared much more concerned than their French counterparts – judging by the annals of opinion research in the United States – about the risks associated with nuclear power and nuclear waste disposal.

Sunstein had little doubt that, attuned to the predispositions and cultural frame of reference of the audience for their news coverage, news organisations focused on risks that they expected to resonate with the anxieties of their readers, listeners and viewers. Cass Sunstein was exploring a meshing of insights drawn from the psychometric paradigm, SARF and cultural theories of risk. Into this brew Sunstein stirred the differential ability of political leaders, campaigners and others, many with an axe to grind, to win and hold the attention of the public as a whole. There is what he calls:

A vicious circle involving the availability heuristic and media incentives, with each aggravating the other, often to the detriment of public understanding. (Sunstein, 2005: 103)

It is, he argues, the psychological and cultural threats to public understanding, exposed and exploited by contemporary media practices, which make it hard for most of us to come to sound judgements about risks, costs and benefits. Sunstein finds this particularly worrying. When mechanisms that reinforce an unsound and unbalanced assessment of risk are given free rein, then the opportunity for balancing arguments to be heard can be dramatically reduced. Societies that have institutions that are insufficiently robust to limit the possibility of social cascades and their impact on public policy-making, and which are drawn to the most extreme interpretations of the precautionary principle, will end up with risk regulation that is quite literally petrifying. Sunstein paints a picture of risk regulation overwhelmed by combinations of the availability heuristic, probability neglect and cultural and personal predispositions, which leave little room for reason and scientific data to balance risk, cost and benefit in fashioning a reliable risk regulation regime. The problem he identifies, in the (p.138) public communication of risk, overwhelmed by fear as wildfire, is that:

Individuals and societies may be fearful of non-existent or trivial risks – and simultaneously neglect real dangers … Law and policy ought not to reflect people's blunders; democracy should not mechanically follow citizens’ fears, or for that matter their fearlessness. (105–106)

Sunstein does recognise a key role for values, values that can be clearly articulated and that bear on real choices. He accepts that social values have a vital role to play in informing trade-offs between rival objectives that require us to balance risk, cost and benefit. But the places in which he believes the properly informed balancing of risks, costs and benefits can take place appear to be locales (scientific, political and cultural) from which the general public are largely absent, generally self-excluded and likely to remain absent. Sunstein's assertion, about the public's absence from places where public policy is best made may appear, to many, to be well justified. Given the public over-reaction – both inside and outside of Belgium – to the discovery of dioxin contamination of food, a contrary view may be hard to present. Although we should not, despite Sunstein's anxieties about the extent of public participation in the detailed process of risk management, lose sight of the fact that the public reaction to what appeared to be belated announcements about the discovery of dioxin contamination in food in Belgium, was amongst other things a reaction to what was widely perceived to be a calculated decision to keep the public in ignorance.

Sunstein's conception of high-quality and appropriate deliberation about risk, deliberation that takes place in the absence of those whose interests and welfare are under discussion, may seem to be a contradiction in terms. However, if social cascades and moral panics threaten the quality of public decision-making, then those who wish to exclude the public also need to be aware of the possibility that social cascades and moral panics will be fuelled by actions, even well-intended actions, by risk communicators and managers that treat the public, if not the public interest, as peripheral to the process of risk regulation itself.

What seems uncontentious is that deliberations about risk and food safety need to be better informed not only about the technical aspects of particular risks, such as those that arise from such things as dioxin contamination, but about the relationship between such risks and the structure of the modern food economy. It is far less certain that Sunstein's general argument, about the strength of the case for keeping the public largely at the margin of risk politics, will prove to be either wise or practicable.

(p.139) The New Scientist and the public understanding of risk and science

In its remarkable coverage of the Belgian dioxin scandal the New Scientist magazine set out to offer insights and provide an informed perspective on risk in simple and direct language; an approach that was essentially unmatched elsewhere.

The magazine's Deborah MacKenzie wrote that she found the ambivalence of the Belgian public easy to understand. Belgian people were angry at their government's failure to issue immediate warnings but were well aware that the release of dioxins was not the government's fault and that there was no obvious way of determining what the impact of dioxin on their health had been and would be in the future. The limited ability of the authorities to make sound judgements about the damage that had been done and the mortality and morbidity that was to come was not something that could be corrected easily, quickly or inexpensively. The public, by and large, in Belgium and elsewhere, understood that increased levels of public safety generally come with a substantial price tag and depend on expertise and information that is not readily available to government or the public at large.

In the immediate aftermath of the dioxin contamination it had proven difficult to explain to the Belgian public, or indeed the public in any part of the world, that the release of dioxins into the environment and contamination of food with dioxins was not explicable in terms of any single event, however shocking. The need for work to measure both the background levels of dioxins and releases from particular events thought likely to endanger public health, by threatening to contaminate or actually contaminating the food supply, was something that simply could not be explained or, more important, addressed successfully in the midst of a great national crisis. The possibility, indeed the likelihood, that there are features of the food economy that substantially increase the likelihood of dioxin contamination is something that needs to be explained carefully over a long period of time.

If the general public, with an increasingly sophisticated knowledge of cooking and the preparation of food, can appreciate the many different qualities of food and ways of preparing it, it ought not to be judged unreasonable to assert that the appreciation of the strengths and weaknesses of a modern food economy can also be much better explained and more widely understood.

From MacKenzie's point of view one of the most worrying aspects of the dioxin release in Belgium, and the consequent food contamination and ingestion of carcinogens, was the impact that dioxin and other POPs had had on the most vulnerable members of society. The New Scientist had made a very simple and obvious point, a point that is surely central to public deliberation about the regulation of food safety. If the accumulation of POPs takes place over (p.140) very long periods of time and acts in ways that are extremely hard for vulnerable individuals to detect then it needs to be studied consistently over a long period of time. The ability to make secure and informative measurements over long periods is something that is primarily dependent on government action and that necessarily entails choices, which are best made in ways that do not depend on any single headline or on the level of public outrage at any particular government.

In its editorial, entitled ‘Recipe for Disaster’, the New Scientist wrote not only about the science but also about the risk of politics of dioxins and PCBs in food (NS, 1999). The New Scientist was in no doubt that what had happened in Belgium was simply and straightforwardly a:

consequence of a large-scale, fast-throughput, competitive industry that keeps us in the West … well supplied with cheap food. Large-scale integration means efficiency, but [it] also [means] that a single mistake can spread right across the system. (NS, 1999)

Those characteristics of the modern food economy are every bit as relevant and salient to public discussion of food safety regulation as the undoubted existence of social cascades.

Well-informed public decision-making not only needs to address the possibility, indeed the reality of social and availability cascades, it also needs to address the problems of probability neglect and even more obvious forms of neglect: neglect of important aspects of any risk regulation regime aimed at maintaining and enhancing food safety. To do this in a way that has any prospect of commanding public support requires that the public are involved and consulted about the full range of issues that need to be addressed – and the costs and opportunity costs of public action and inaction.

We can all understand that it is easy to blame others when outrage is the strongest emotion that we feel. We can all understand that outrage is a poor guide to a safer future and that policy may change little even if policy makers are changed, unless other things change at the same time. Kicking one lot of public-policy makers out and replacing them with another is most unlikely to contribute much, of itself, to the design or operation of a risk regulation regime. Rather, as the New Scientist editorial suggested, when reflecting on the Belgian dioxin affair:

The bottom line is a food industry where innocent mistakes can happen on a massive scale and not be detected until too late … looked at [in this] way the Belgian fasco is not surprising …

The New Scientist was certain that the Belgian authorities had a duty, in light of what had happened, to collect as much information as possible over a period of thirty years, so that ‘officials have some idea of the kind of doses people have (p.141) received’ and the health consequences of those doses. The ability to undertake continuous measurements of this kind, and to employ the analytical capacity to make sense of the data that are produced, requires a politics of anticipation and appropriate judgements about the need for precautionary steps.

The greatest challenge in risk politics is to find a way of achieving a balance between precaution and reaction, in the knowledge that while this balance is vulnerable to social cascades, to waves of unreason and unreasonableness, it can also be determined in a way that is not only more deliberative but also more robust if it engages the public in ways and over periods of time that are not dictated by the most widely, dramatically and feetingly reported risk events.