Food Selection and Risk Perception

Recent crises associated with food safety issues in the affluent world are bringing about profound changes, temporary or permanent, in attitudes about food, selection patterns and food production. While experts evaluate risks using various probabilistic tools, for example epidemiological data (mortality and morbidity rates associated with a given occurrence or practice), lay people, rather than evaluating risk in a probabilistic fashion, perceive it in ways that are often described as «irrational» by experts, industry and authorities. This is particularly true when it comes to possible dangers associated with food. Perception of risk by the consumer can at least in part be predicted. Relevant dimensions include risk configuration; certain types and configurations tend to increase fear and outrage. Other factors are associated with psychological and cognitive features, what anthropologists call «magical thinking», social and cultural factors. Not all groups and cultures are equally fearful or worried, and not about the same risks. In 1996, Europe (and the world) experienced the first BSE crisis following the announcement in Britain that Bovine Spongiform Encephalopathy (BSE) could probably be transmitted to humans as New Variant Creutzfeldt-Jakob disease (nvCJD). Across the continent, beef consumption dropped by 20 to 60 per cent, and did not resume its previous course until several months later (8). In October 2000, a second BSE crisis caught French authorities unprepared. There had been no new, alarming scientific finding. Yet one particular incident triggered the reaction: a cow was diagnosed with BSE just as it was to enter the slaughterhouse. By the rules in force in France, all animals in the herd were to be culled and destroyed. But in this instance it turned out that several had already been slaughtered and marketed before they could be identified. A number of supermarket chains recalled significant quantities of meat. Even though the infected animal had been removed before actually entering the human food chain, and in spite of the fact that so-called Specified Risk Materials are systematically disposed of, the incident spurred considerable media attention and aroused great public emotion. Within days, many mayors across France banned beef from municipal school cafeterias. After several days of media and political turmoil, the government was forced to enact a general ban on meat and bone meal in all animal feed. (Until then it had been banned for ruminants only.) French consumption of beef fell by nearly fifty per cent compared with the same period the previous year. In midFebruary it was still thirty-five per cent lower than just before the outbreak. By the end of the next summer, consumption levels were up, but were still ten per cent lower than the previous year. On July 28, 1987, Monitor, a program on the West German television network ARD, aired a report on the North Sea fishery. Ten million viewers were treated to close-ups of Anisakis simplex—herring worms—taken from the entrails and flesh of a herring, wriggling about on a knife blade, then under a microscope. A young man recounted that he had been infected by this parasite after eating herring and that, as a result, twelve centimetres of his large intestine had had to be surgically removed. Researchers employed by Monitor found live worm larvae in jars of herring purchased in supermarkets. The program anchor concluded that the existing system of self-regulation in the German fishing industry was inadequate to guarantee public safety and health (15). 137 138 Claude Fischler Food Selection and Risk Perception The market crashed overnight. Auction prices plunged by half. Retail fish sales fell by fifty to eighty per cent, depending on the region. The program’s producer expressed surprise; he said that he had expected a drop of less than ten per cent and that the purpose of the report had simply been to exert some pressure to reform monitoring and regulatory procedures. With a view to reassuring consumers, a fishery expert pointed out that, over the course of eighteen years, only sixty cases of worm contamination in humans had been noted out of 7.5 billion fish meals consumed in Germany, a statistic that hardly labeled the herring worm as a major threat to public health. There are great differences between the crisis of 1987 and that of 1996. The threat posed by the herring worm could be identified, understood and assessed in terms of probabilities. Transmissible spongiform encephalopathies in general, and nvCJD in particular, pose an unknown risk. The most precise calculations—if they can be called that—put the total number of possible victims at somewhere between 100 or so (the number of known cases today) and 140,000 (22). Moreover, these diseases are a major scientific challenge, as some of their features seem to contradict some of the most fundamental laws of biology. But the two crises also have much in common. Both had social effects that officials and the media called a “panic” or “scare”, words that imply irrationality. The lay public and the experts assessed the risks in very different ways. As a basis for their assessments, professionals knew (in the case of the herring worm) or tried to estimate (in the case of BSE) the number of clinical cases and possible morbidity and mortality levels. As for TV viewers, they saw a writhing, revolting parasite, enlarged 1 At this point, there is still no complete consensus on transmission of Bovine Spongiform Encephalopathy, Creutzfeldt-Jakob Disease and other transmissible spongiform encephalopathies (TSEs) but most of the scientific community accepts that prions play at least a part in the process. Prusiner's theory, which earned him a Nobel Prize in 1997, accounts for the otherwise difficult to explain fact that TSEs are transmitted both genetically and by infection. The prion is, according to the theory, the pathogenic form (PrPsc) of a protein that is normally present in organisms (PrPc). The gene coding for the normal protein can be altered by a mutation, thus coding for pathogenic prions instead, which results in the "sporadic" form of the disease. Yet normal prions seem to have the property of turning into pathogenic prions by contact, through a mechanism that is still in part hypothetical. In consequence, if fed animal tissue containing PrPsc prions, an animal can in turn develop the illness. on-screen to Loch Ness monster dimensions. They saw mountains of cow carcasses being bulldozed by heavy machinery; they saw truckloads of slaughterhouse waste and semi-identifiable guts. They were not about to start estimating the probabilities of the risk. The cognitive and physiological mechanisms in play were those of disgust and fear, and these brought about, literally, a visceral rejection of the food associated with, or one might say contaminated by, the repulsive stimulus. This helps us understand the phenomenon: the most deadly risks, quantitatively speaking, do not necessarily cause the most profound anxiety or have the hugest media and economic repercussions, especially when it comes to food. Over the past few decades, work by psychologists and social scientists has revealed the split between risk evaluation by experts and risk perception by the lay public. Paul Slovic (20) showed that although nuclear power posed the greatest risk in the view of members of a sample of lay Americans, it was only in twentieth place for a group of experts in risk assessment. Why the disparity? Risk evaluation (by experts) and risk perception (by the lay public) are not generated in the same way, and do not result from the same kind of reasoning or mental mechanism. The factors that influence the public’s perception of risk can be divided into two groups: the form and particular configuration of the risk itself, and the psychological, cognitive, social and cultural dimensions of the subject. The Form of the Risk Several features of risk have been identified as able to produce “dread and outrage factors” and thus to heighten emotion and outrage in individuals and groups. Here are a few: A nearby, concrete risk that can be represented, vividly imagined or observed generates more fear than a distant, abstract risk. A deliberate risk, one the subject has decided to be exposed to, is less problematic than an imposed risk. The risks associated with skiing are high, but they result from a personal decision and are more likely to be accepted as such; risk perception is, therefore, diminished. On the other hand, a risk taken without 139 the subject’s knowledge or without the subject having decided to take it— especially if the risk stands to benefit someone other than the risk-taker—makes for a powerful outrage factor. Note that studies on perceptions of genetically modified (GM) organisms show that, for the present, a good part of the public perceives neither benefits to them nor control over possible risks (1–4). The inability to control an identified risk is a source of anger and extra anxiety. The possible danger linked with drinking tap water is a classic case: it is very hard to avoid using tap water completely and thus to protect oneself from the risks associated with it. Another example can be drawn from the mad cow crisis. While it is relatively easy to avoid eating bovine offal that has been identified as infectious (possible perceived control), it becomes extremely difficult to protect oneself (difficult or impossible perceived control) when one discovers that all manner of beef byproducts are to be found in a wide variety of unsuspected products, such as gelatin in candies, cosmetics and surgical sutures. Risk perception is also affected by whether the causes of the risk are human or natural. In principle, “natural” risks are thought to arouse less outrage than those taken as the result of a human action. In fact, it can be observed that we seem to be more readily interested in identifying the guilty parties than in a rigorous analysis of the deterministic complexities, even in the case of natural disasters. This often prompts the media and public opinion to point the finger of blame at entities ranging from the government and politicians to multinational corporations. Finally, risks associated with familiar technologies, such as railways, cause less outrage than those linked with new, poorly understood technologies, such as genetic engineering. Characteristics of the Subject Cognitive psychology has shown that probabilistic reasoning is counterintuitive. Even people with training in statistics sometimes commit errors linked to cognitive bias (14, 21). Risk is a probabilistic concept, and in discussions related to any crisis situation it is difficult, to say the least, to avoid reference to probabilistic reasoning or to statistical data. 140 Claude Fischler Food Selection and Risk Perception