Safe Food: The Politics of Food Safety (5 page)

The panel’s responses to the EPA surely constitute the most thorough evaluation of a food allergen ever conducted and provide a vivid example of how difficult it is to make policy decisions based on science that is incomplete and uncertain (which so often is the case). Panel members said they were “uncomfortable with the available data” and did not have enough information to decide whether the StarLink protein could cause allergic reactions. They knew that proteins are strings of amino acids arranged in a particular sequence, and that whether a protein provokes an allergic response depends on how that sequence folds—its structure and shape. Only some proteins are allergenic, but it is not yet possible to predict the structural features that induce allergic reactions. The panel members had to make educated guesses about the size, digestibility, and stability to heat of the Cry9C protein, and about the prevalence of this protein in the food supply.

One reason the Cry9C protein is toxic to insects is that they cannot easily digest it—break it down—to its constituent amino acids; the structure
of the protein survives the digestive processes more or less intact. The Cry9C protein also is relatively stable to heat, so cooking might not destroy its ability to cause allergic reactions. Furthermore, preliminary feeding studies showed that the Cry9C protein appeared intact in the blood of rats and provoked immune responses, meaning that rats could not digest it and destroy its allergenicity. No such studies had been conducted in humans, however. Thus, panel members could not dismiss the possibility that the StarLink protein
might
be allergenic to humans. They judged the StarLink protein to have a “medium” likelihood of being allergenic, mainly because its potential to induce allergic reactions could not be disproved. Because processing and cooking were likely to destroy some of the Cry9C proteins, and the amounts were quite small to begin with, they judged Cry9C to have a “low” probability of actually causing allergic reactions in the population. These judgments supported the EPA’s precautionary decision not to allow StarLink to enter the human food supply.
⁵

A further complication is the question of whether people actually experience allergic reactions when they eat StarLink products. As it turns out, this connection is not easy to prove. Just because people feel sick after eating a food does not necessarily mean that the food—and not something else—caused the illness. Finding the StarLink
gene
in a food does not necessarily mean that the protein it specifies will cause allergic reactions. Like other genes, the StarLink gene is made of DNA (deoxyribonucleic acid), and its constituent components are common to all living species (see appendix). DNA and genes do not induce allergic reactions, but they specify the structure of proteins. Proteins (but not all of them) cause allergies. To prove that the StarLink protein is allergenic, scientists have to show that people reporting allergic reactions ate foods containing the StarLink corn protein and displayed immune responses to the StarLink protein in their blood. To investigate these matters, the FDA had to develop new testing materials and methods, and quickly. By June 2001, 63 people complained to the FDA about allergic reactions to StarLink and agency scientists collected food and blood samples from about 10 of them.

Using the new methods, FDA scientists tested the food samples but could not detect the StarLink
gene
in any of them. They also failed to find the StarLink protein in the foods, although the test was inconclusive in one sample. In the meantime, scientists from the Centers for Disease Control and Prevention (CDC) tested the blood samples for evidence of immune responses to the StarLink protein; they found none. These results
led the agencies to conclude that the reported illnesses must have been caused by something other than an allergic reaction to the StarLink protein.
⁶

With these results in hand, the EPA Scientific Advisory Panel met again in July 2001 to continue debating issues related to StarLink allergenicity. By this time, the EPA had canceled the StarLink registration, thereby prohibiting further plantings. The U.S. Department of Agriculture (USDA) joined with Aventis to buy the remaining mixtures of conventional and StarLink corn to use for animal feed or industrial uses. Corn handlers, millers, and food processors began testing to see whether their stocks contained the StarLink Cry9C protein, and began selling off the commingled corn.

In the meantime, EPA panel members continued to raise questions about the reliability of the FDA and CDC testing methods and said that they still could not exclude the possibility that StarLink might be allergenic. They saw no reason to change their previous conclusion that the Cry9C protein had a medium chance of being allergenic but a low chance of actually causing allergic reactions in the population. Instead, they said it was time to ask
political
questions: “What went right? What went wrong? What have we learned? How did Cry9C penetrate the human food supply? Why was the adulteration detected by a public interest group rather than through a more formal surveillance program (e.g., Federal agencies or regulated industry)?”
⁷

The answers to such political questions depend on the point of view—and, therefore, the interests—of the various stakeholders in food safety: the food industry, the government, consumer advocacy groups, and the general public. The StarLink affair revealed how these interests affect opinions and actions related to safety matters.

We can begin by looking at the reactions of the food industry—in this instance, the companies that produce, process, and sell StarLink corn or its products. As indicated in
figures 1
and
2
, large national and international corporations own many of the companies involved in the StarLink chain of production and distribution. These companies are businesses that must respond to the demands of directors and stockholders, and it seems likely that their managers had more immediate matters to worry about than whether corn intended for animal feed was commingling with conventional corn.

Aventis officials behaved as if they had no doubt that the EPA would approve StarLink for human consumption and would allow it to remain in the food supply. They began with denials and finger-pointing, starting with an attempt to discredit the accuracy of the GeneticID test. When subsequent testing confirmed the presence of the Cry9C gene in supermarket foods, Aventis “volunteered” to give up its right to plant StarLink, reportedly because the EPA threatened to revoke its registration.
⁸
The company also tried another tack; it petitioned the EPA to allow StarLink to remain in supermarket foods for four more years until virtually all commingled products would be sold. Aventis officials argued that the amounts in food were too small to harm consumers and that having to remove foods containing StarLink from corn supplies and supermarket shelves would greatly disrupt the food system. Indeed, disruptions were likely to be considerable, since the commingled corn for the 2000 crop amounted to 124 million bushels, and Japan and South Korea had rules forbidding any genetically modified corn from entering their countries. For all of these reasons, the Grocery Manufacturers of America and other food industry trade associations strongly supported the Aventis petitions.

Using yet another tactic, Aventis asked the EPA to set a “tolerance” limit for StarLink—a level below which regulatory agencies would ignore traces of the Cry9C gene or protein in food. Aventis warned corn processors that StarLink was so thoroughly commingled in the corn supply that the only way to deal with that situation was to
accept
it: “Will there ever be an end to this? Unfortunately, as of right now, the answer is ‘no’—there will never be an ‘end’ as long as there is zero tolerance for Cry9C in food.”
⁹

These events led critics to ask the questions raised in any political scandal: What did Aventis and the EPA know, and when did they know it? Reports soon trickled out that both company and government officials knew—perhaps as early as 1997 and certainly by 1998—that StarLink was commingled with conventional corn. At a meeting late in 2000, I heard an official of the EPA say—unfortunately not for direct quotation—that Aventis had worked hard to lobby the White House Office of Science and Technology Policy, the State Department, and the FDA, USDA, and EPA during the months prior to the taco shell disclosure in an effort to convince federal officials that StarLink was not going to cause safety problems. Because Aventis officials acted as if StarLink were demonstrably safe, they were vulnerable to criticism from consumer groups like Friends of the Earth: “Aventis can’t possibly have enough information to conclude that StarLink is safe at any level in our food.”
¹⁰

Other companies in the StarLink chain joined Aventis in further denial and blame. Officials of Garst Seed said that farmers knew they were supposed to separate StarLink from other corn, and “it’s unfortunate some customers say they weren’t informed about the program. . . . But we worked hard to get that message out.”
¹¹
Farmers, however, denied they had been told to segregate StarLink and filed lawsuits for damages. Operators of grain elevators also denied hearing anything about the need for crop segregation, and at least half of them had forwarded commingled corn for unapproved uses. Overall, the various companies in the chain of production and distribution assumed that their customers would not much care about this issue. As an analyst for J. P. Morgan explained, “If you’re eating at Taco Bell, health consciousness is not high on your list of concerns.”
¹²

The government also is a major stakeholder in food safety, and its responses reflected the peculiar way in which regulatory authority is distributed among no less than three major agencies—the EPA, FDA, and USDA (see
chapter 1
). EPA officials criticized Aventis for claiming innocence about how StarLink might have gotten into the human food supply, for insufficiently informing growers about the need for crop segregation, and for flagrantly ignoring the terms of the restricted registration. The FDA at first seemed unconcerned; StarLink corn was the EPA’s problem, and the taco shells, which do fall under FDA jurisdiction, seemed unlikely to be harmful. One FDA official reassured the
New York Times
, “This is not a case where we have illnesses or health problems.”
¹³
When the FDA had to ask Friends of the Earth for a sample of the taco shells in order to conduct its own after-the-fact testing, however, it seemed clear that the agency was giving “inadequate oversight and attention to a serious matter of public health.”
¹⁴
The secretary of the USDA blamed Aventis: “Some might argue that the StarLink episode will lead to greater government involvement. . . . It’s important to remember that this problem may not have occurred had industry complied with the terms of its license.”
¹⁵
Nevertheless, the USDA agreed to spend $20 million to buy back commingled seed in an effort to prevent disruption of the corn market.

Consumer advocacy groups used the potential allergenicity of StarLink to bolster their demands that genetically modified foods be tested before entering the food supply and labeled so people can protect themselves against foods to which they might be allergic. They viewed the events as evidence that neither government nor industry were looking out for the public interest. Representatives from Friends of the Earth and Consumers Union argued, “There is no way the taxpayer should bail out
Aventis for the genetic pollution they created,” and “EPA should not reward Aventis for their failure to follow the law.”
¹⁶
Even business commentators were dismayed: “Almost everybody involved screwed up. . . . The promises made by StarLink’s inventors proved worthless, falling prey to managerial inattention, corporate mergers, blind faith, misplaced hope, woeful ignorance, political activism, and probably greedy farmers too, if you can imagine such a thing.”
³
Whether or not StarLink really is allergenic (a food safety issue) its unlabeled presence in processed foods did nothing to encourage trust in the food supply, and these events revealed the markedly different ways in which the various stakeholders view matters of food safety risk.

With StarLink products recalled and class action suits settled, we now turn to the food safety interests of the general public. As consumers, we want food to be safe—or safe enough—and we expect the food industry and government to make sure that it is. We also are part of the political equation. As stakeholders in the food system, however, our influence depends on the extent to which we recognize the political forces at work in safety matters. Enhancing that understanding is a principal aim of this book. If the StarLink episode teaches us anything, it is that ensuring food safety is a matter of politics as well as science. In conveying this lesson, the StarLink story illustrates several of the themes that recur throughout the chapters that follow.

The first theme is the fragmented, overlapping, and confusing distribution of authority among the federal agencies concerned with food safety: the EPA, FDA, and USDA. All three agencies were in some way responsible for making sure that StarLink did not get into the human food supply, yet the system failed to ensure that food companies followed rules designed to protect public health. We will see how this divided authority complicates federal oversight of microbial contaminants in food, genetically engineered foods, and protection of the food supply against potential threats of bioterrorism.

A second theme is the food industry’s promotion of economic self-interest at the expense of public health and safety. We have just seen how the developers of StarLink assumed that the corn was safe to eat, made little effort to keep it out of the human food supply, and blamed other parts of the food distribution chain for its appearance in taco shells. The StarLink affair is just one example of what
Sierra
magazine calls “Brave New Nature—What Happens When Biology Meets Big Business?” (see
figure 3
). This book provides further examples of situations in which food companies deny responsibility and blame others in matters of food safety, and oppose, resist, and undermine food safety guidelines, following them only when forced to do so by government action or public opinion.