Safe Food: Bacteria, Biotechnology, And Bioterrorism Part 2

Despite the warnings, Mr. Espy accepted a variety of favors from lobbyists for Tyson Foods, then the world's largest chicken-processing company (and now the largest producer of beef as well). Early in 1994, federal investigators accused Mr. Espy of violating the Meat Inspection Act by accepting-or permitting a companion to accept-airline travel, tickets to sporting events, a small scholars.h.i.+p, and other gifts worth about $12,000 from Tyson Foods, plus similar gifts from other meat and poultry companies. The results of that investigation forced Mr. Espy to resign from his position as USDA secretary, and a later investigation by a special prosecutor led to his indictment by a federal grand jury. Eventually, Mr. Espy was acquitted of all charges, largely because the prosecutor could not convince a jury that the gifts were intended intended to influence the USDA. Although it is difficult to imagine what other purpose the gifts might have served, the Supreme Court also ruled that Mr. Espy was ent.i.tled to accept them because they were not directly linked to regulatory matters. to influence the USDA. Although it is difficult to imagine what other purpose the gifts might have served, the Supreme Court also ruled that Mr. Espy was ent.i.tled to accept them because they were not directly linked to regulatory matters.33 Federal prosecutors also accused Mr. Espy's chief of staff, Ronald Blackley, of interfering with USDA attempts to regulate poultry safety. A USDA staff member told a reporter that Mr. Blackley had been surprised to learn that the agency was working on poultry rules: "He said to take [them] out of the computer. . . . We were a little sh.e.l.l shocked . . . wondering if we had all heard what we thought we'd heard. We put a stop to all poultry activity."32 At the time, the scandal made the USDA's ongoing efforts to improve poultry safety much more difficult. Some critics charged that the department's proposed rules for poultry inspection were simply "an effort to prove that . . . Espy was not beholden to poultry interests."34 When the agency decided not to go forward with the plan, officials had to deny that they had made this decision just to please the poultry industry. The Espy scandal, neither the first nor the last of its kind, was unusual only in that the favors were so visible and the issues so important. This particular USDA secretary had the opportunity and the ability to convert his department's century-old inspection system to one better equipped to deal with microbial pathogens. Tyson Foods' donation of tickets to sporting events demonstrated that even small favors produce substantial benefits if given at the right time, in this case just when the USDA was trying to get poultry producers to test for When the agency decided not to go forward with the plan, officials had to deny that they had made this decision just to please the poultry industry. The Espy scandal, neither the first nor the last of its kind, was unusual only in that the favors were so visible and the issues so important. This particular USDA secretary had the opportunity and the ability to convert his department's century-old inspection system to one better equipped to deal with microbial pathogens. Tyson Foods' donation of tickets to sporting events demonstrated that even small favors produce substantial benefits if given at the right time, in this case just when the USDA was trying to get poultry producers to test for Salmonella Salmonella and other pathogens. If nothing else, it worked greatly to Tyson Foods' advantage to keep Mr. Espy preoccupied with responses to legal challenges from a special prosecutor. As if the political nature of this situation were not transparent enough, one of President Clinton's last acts in office was to grant presidential pardons to Mr. Blackley and six food company executives and lobbyists who had been convicted of attempting to corrupt Mr. Espy. Reportedly, the White House and other pathogens. If nothing else, it worked greatly to Tyson Foods' advantage to keep Mr. Espy preoccupied with responses to legal challenges from a special prosecutor. As if the political nature of this situation were not transparent enough, one of President Clinton's last acts in office was to grant presidential pardons to Mr. Blackley and six food company executives and lobbyists who had been convicted of attempting to corrupt Mr. Espy. Reportedly, the White House invited invited defense lawyers to request the pardons, and granted them just hours before George W. Bush took office as president in January 2001. defense lawyers to request the pardons, and granted them just hours before George W. Bush took office as president in January 2001.35 USDA REQUIRES PATHOGEN TESTING: E. COLI E. COLI O157:H7 IN GROUND BEEF, 1994 O157:H7 IN GROUND BEEF, 1994.

By the early 1990s, USDA officials had argued for two decades that the decision in APHA APHA v. v. Butz Butz meant that the department did not have legal authority to set limits on microbial contaminants in meat and poultry because pathogens like meant that the department did not have legal authority to set limits on microbial contaminants in meat and poultry because pathogens like Salmonella Salmonella were "inherent" in raw meat. As late as 1993, the administrator of USDA's Food Safety and Inspection Service (FSIS), H. Russell Cross, explained to a congressional committee: "At the present time, meat and poultry inspection laws do not define raw meat and poultry containing bacteria as adulterated." were "inherent" in raw meat. As late as 1993, the administrator of USDA's Food Safety and Inspection Service (FSIS), H. Russell Cross, explained to a congressional committee: "At the present time, meat and poultry inspection laws do not define raw meat and poultry containing bacteria as adulterated."36 As noted earlier, the USDA could have interpreted As noted earlier, the USDA could have interpreted APHA APHA v. v. Butz Butz as giving the department considerable lat.i.tude to do whatever seemed necessary to protect the public, including setting performance standards-allowance limits verified by testing-for pathogens in meat. While Mr. Espy's legal difficulties were front-page news, he chose Michael Taylor to become administrator of FSIS. Mr. Taylor, a lawyer, moved to the USDA from the FDA; there, his previous employment with Monsanto raised conflict-of-interest questions about his role in setting policy for regulation of genetically modified foods (see as giving the department considerable lat.i.tude to do whatever seemed necessary to protect the public, including setting performance standards-allowance limits verified by testing-for pathogens in meat. While Mr. Espy's legal difficulties were front-page news, he chose Michael Taylor to become administrator of FSIS. Mr. Taylor, a lawyer, moved to the USDA from the FDA; there, his previous employment with Monsanto raised conflict-of-interest questions about his role in setting policy for regulation of genetically modified foods (see chapter 7 chapter 7). His actions at the USDA raised no such questions. In late September 1994, six weeks after a.s.suming leaders.h.i.+p of FSIS, Mr. Taylor gave his first public speech in his new job to an annual convention of the American Meat Inst.i.tute. He said that it was high time for everyone involved in meat production and processing "to be driven as much by public health goals as by productivity concerns." FSIS intended to take advantage of "the tools of microbiology to ensure that preventive controls are in place to reduce the risk of harmful contamination and to verify that those controls are working." He announced that FSIS would soon propose regulations requiring installation of science-based HACCP systems in every meat and poultry plant. "Raw ground beef contaminated with E. coli E. coli O157:H7," he said, "poses a serious risk to public health, and contaminated lots should be excluded from commerce." The USDA intended to require the destruction or reprocessing of contaminated meat and "we expect companies who encounter contaminated lots of raw ground beef . . . to take similar action." O157:H7," he said, "poses a serious risk to public health, and contaminated lots should be excluded from commerce." The USDA intended to require the destruction or reprocessing of contaminated meat and "we expect companies who encounter contaminated lots of raw ground beef . . . to take similar action."37 If that challenge was not enough to bring his audience to rapt attention, he explained that FSIS would be taking these actions on the basis of the department's revised interpretation of APHA APHA v. v. Butz: Butz: To clarify an important legal point, we consider raw ground beef that is contaminated with E. coli E. coli O157:H7 to be adulterated within the meaning of the Federal Meat Inspection Act. We are prepared to use the Act's enforcement tools, as necessary, to exclude adulterated product from commerce. Finally, we plan to conduct targeted sampling and testing of raw ground beef at plants and in the marketplace for possible contamination with O157:H7 to be adulterated within the meaning of the Federal Meat Inspection Act. We are prepared to use the Act's enforcement tools, as necessary, to exclude adulterated product from commerce. Finally, we plan to conduct targeted sampling and testing of raw ground beef at plants and in the marketplace for possible contamination with E. coli E. coli O157:H7. This sampling program . . . will serve as an example and an incentive for those commercial enterprises that produce, process, and market raw ground beef to control their processes and conduct their own tests. O157:H7. This sampling program . . . will serve as an example and an incentive for those commercial enterprises that produce, process, and market raw ground beef to control their processes and conduct their own tests.37 Furthermore, because E. coli E. coli O157:H7 is infectious at very low doses, FSIS would consider O157:H7 is infectious at very low doses, FSIS would consider any any level of contamination of ground beef with these bacteria to be unsafe, adulterated, and subject to enforcement action. The agency, however, would restrict this "sample, test, and destroy" approach to just this one pathogen, level of contamination of ground beef with these bacteria to be unsafe, adulterated, and subject to enforcement action. The agency, however, would restrict this "sample, test, and destroy" approach to just this one pathogen, E. coli E. coli O157:H7, and to just this one product: ground beef. O157:H7, and to just this one product: ground beef.38 Food safety advocates admire Mr. Taylor for his courage in delivering this speech to an audience expected to be unsympathetic if not downright hostile. They also appreciate his skill in s.h.i.+fting USDA food safety policies to those more favorable to public health, especially at a time when the department's leaders.h.i.+p was in such deep trouble. Indeed, he needed courage. His speech caused consternation in the cattle, meatpacking, and grocery industries. Meat producers and processors understood that if the USDA considered E. coli E. coli O157:H7 an "adulterant," they would break the law if they sold foods containing this pathogen. They would be vulnerable to criminal prosecution. As a representative of the American Meat Inst.i.tute told the press, "the new USDA policy has the perhaps unintended consequence of creating rampant, irresponsible, criminal litigation." O157:H7 an "adulterant," they would break the law if they sold foods containing this pathogen. They would be vulnerable to criminal prosecution. As a representative of the American Meat Inst.i.tute told the press, "the new USDA policy has the perhaps unintended consequence of creating rampant, irresponsible, criminal litigation."39 Industry lawyers instructed their clients not to do their own testing of ground beef for Industry lawyers instructed their clients not to do their own testing of ground beef for E. coli E. coli O157:H7 because finding it would expose them to legal liability. Rosemary Mucklow, then the executive director of the Western States Meat a.s.sociation, said, "How can FSIS treat O157:H7 because finding it would expose them to legal liability. Rosemary Mucklow, then the executive director of the Western States Meat a.s.sociation, said, "How can FSIS treat E. coli E. coli in hamburger meat as an adulterant subject to enforcement strategies, while not applying the same standard to salmonella in broilers. . . . Automated chicken lines allow birds to leave the plant as Inspected for Whole-someness with . . . grossly unacceptable defects. . . . Such gross policy interpretation favoring the poultry industry and disfavoring the beef industry is a travesty indeed." in hamburger meat as an adulterant subject to enforcement strategies, while not applying the same standard to salmonella in broilers. . . . Automated chicken lines allow birds to leave the plant as Inspected for Whole-someness with . . . grossly unacceptable defects. . . . Such gross policy interpretation favoring the poultry industry and disfavoring the beef industry is a travesty indeed."40 We will encounter further commentary from Ms. Mucklow later in these pages. We will encounter further commentary from Ms. Mucklow later in these pages.

In the meantime, the American Meat Inst.i.tute-which had opposed the safe-food-handling labels-now used them to complain that the proposed testing program would cause cause food safety problems. Microbial testing would "mislead consumers with promises of a safer food supply, and as a result they may relax their own cooking and handling standards." food safety problems. Microbial testing would "mislead consumers with promises of a safer food supply, and as a result they may relax their own cooking and handling standards."4 The Food Marketing Inst.i.tute also s.h.i.+fted responsibility to consumers in its argument against the initiative: "It is essential that nothing dilute the consumer message that the proper cooking of meat eliminates food-borne pathogens." The Food Marketing Inst.i.tute also s.h.i.+fted responsibility to consumers in its argument against the initiative: "It is essential that nothing dilute the consumer message that the proper cooking of meat eliminates food-borne pathogens."41 The two trade a.s.sociations and five others quickly filed suit to block the pathogen testing plan, based on this wonderfully convoluted argument: because the USDA had done nothing to control The two trade a.s.sociations and five others quickly filed suit to block the pathogen testing plan, based on this wonderfully convoluted argument: because the USDA had done nothing to control E. coli E. coli O157:H7 since the first outbreak in 1982, the present situation could hardly be considered an emergency. This transparently self-serving argument prompted the O157:H7 since the first outbreak in 1982, the present situation could hardly be considered an emergency. This transparently self-serving argument prompted the New York Times New York Times to note the industry's "odd way" of promoting public health: "Trying to give their obstructionist lawsuit a respectable veneer, the plaintiffs voice concern that the spot-inspection program could mislead consumers into relaxing their own safe handling and cooking practices. . . . It is not consumers the lawsuit seeks to protect but the industry's right to sell tainted beef." to note the industry's "odd way" of promoting public health: "Trying to give their obstructionist lawsuit a respectable veneer, the plaintiffs voice concern that the spot-inspection program could mislead consumers into relaxing their own safe handling and cooking practices. . . . It is not consumers the lawsuit seeks to protect but the industry's right to sell tainted beef."42 Among the many ironic aspects of this dispute, the trade a.s.sociations' lawsuit turned out to be heard by the very same Texas district court judge, James Nowlin, who had ruled against the USDA's proposals to require food-handling labels on procedural grounds just a year earlier. This time, the court surprised observers by ruling in favor of the USDA. Its rationale: because ordinary cooking temperatures could not kill E. coli E. coli O157:H7, the USDA had good reason to consider these bacteria as adulterants and test for them. This decision at last permitted the USDA to redesign its ancient food inspection system and start testing for this one harmful pathogen. Industry groups, however, saw the decision as mandating a program that "fails to protect consumers, wastes tax dollars and violates the law," and they vowed to "maintain our course of legal action to stop it." O157:H7, the USDA had good reason to consider these bacteria as adulterants and test for them. This decision at last permitted the USDA to redesign its ancient food inspection system and start testing for this one harmful pathogen. Industry groups, however, saw the decision as mandating a program that "fails to protect consumers, wastes tax dollars and violates the law," and they vowed to "maintain our course of legal action to stop it."43 The trade a.s.sociations' lawsuit had one additional-and unantic.i.p.ated-consequence. It mobilized the families of children killed by E. coli E. coli O157:H7 to form their own group-Safe Tables Our Priority (STOP)-to lobby for more rigorous meat regulations. The group picketed a meeting of the American Meat Inst.i.tute and held a press conference to accuse meat producers of obstructing safety efforts: "My 6-year-old son Alex deserves to be alive today. . . . I hold the meat inst.i.tute personally responsible" and "It's time to stop blaming consumers for not cooking and give them a clean product." O157:H7 to form their own group-Safe Tables Our Priority (STOP)-to lobby for more rigorous meat regulations. The group picketed a meeting of the American Meat Inst.i.tute and held a press conference to accuse meat producers of obstructing safety efforts: "My 6-year-old son Alex deserves to be alive today. . . . I hold the meat inst.i.tute personally responsible" and "It's time to stop blaming consumers for not cooking and give them a clean product."44 Another consumer group, the Safe Food Coalition, proposed a "simple household solution" to the problem of the industry's intransigent refusal to test for Another consumer group, the Safe Food Coalition, proposed a "simple household solution" to the problem of the industry's intransigent refusal to test for E. coli E. coli O157:H7 and its persistent avoidance of accountability: obtain proof of responsibility. "Tired of being a victim? . . . Weary of subjecting your family to a game of Russian Roulette every time you buy a package of hamburger meat? . . . [When] unpacking groceries, tuck the supermarket receipt and a small lump of hamburger in a ziplock bag. Toss this in the freezer. . . . In five seconds, at virtually no cost, you've got accountability. . . . This simple act gives control back to you and tells industry loud and clearly that we're not going to take it any more." O157:H7 and its persistent avoidance of accountability: obtain proof of responsibility. "Tired of being a victim? . . . Weary of subjecting your family to a game of Russian Roulette every time you buy a package of hamburger meat? . . . [When] unpacking groceries, tuck the supermarket receipt and a small lump of hamburger in a ziplock bag. Toss this in the freezer. . . . In five seconds, at virtually no cost, you've got accountability. . . . This simple act gives control back to you and tells industry loud and clearly that we're not going to take it any more."45 In this instance, the political context permitted the USDA to hold its position and test one product (ground meat) for one pathogen (E. coli O157:H7). When the department attempted to extend testing requirements to other forms of meat and other pathogens, it again met with fierce resistance, as O157:H7). When the department attempted to extend testing requirements to other forms of meat and other pathogens, it again met with fierce resistance, as chapter 3 chapter 3 will reveal. will reveal.

USDA POISED TO PROPOSE HACCP, 1994.

By the early 1990s, the long history of collusion between meat producers, Congress, and the USDA seemed to have entered a new phase in which public interests held greater influence. Meat producers' protection of the century-old inspection system no longer seemed credible. Although the industry contended that testing for microbes is unnecessary because so few samples are contaminated, this argument ignored a key point: even a small level of contamination can do great harm when the number of animals is large. For example, if just 0.2% of cattle are contaminated with E. coli E. coli O157:H7, 74,000 beef carca.s.ses might be infected and, therefore, hundreds of thousands of pounds of hamburger. O157:H7, 74,000 beef carca.s.ses might be infected and, therefore, hundreds of thousands of pounds of hamburger.46 For this reason alone, the pathogen reduction component of HACCP (which necessarily includes performance standards and testing for microbes) seems thoroughly warranted. For this reason alone, the pathogen reduction component of HACCP (which necessarily includes performance standards and testing for microbes) seems thoroughly warranted.

USDA's historical reluctance to change its inspection and pathogen control systems derives directly from the agency's conflicting mandates: to ensure the safety and quality of foods under its jurisdiction and, at the same time, to promote their marketing and consumption. The long-term collusion between the department and the meat industry impedes progress. Over the years, the breakdown of the agricultural establishment, the emergence of new food pathogens like E. coli E. coli O157:H7, and the appointment of USDA officials interested in the health effects (as well as the economic effects) of agricultural products, paved the way for more vigorous efforts to inst.i.tute HACCP with performance standards for controlling pathogens. O157:H7, and the appointment of USDA officials interested in the health effects (as well as the economic effects) of agricultural products, paved the way for more vigorous efforts to inst.i.tute HACCP with performance standards for controlling pathogens.

In the 20 years between 1974 and 1994, resistance to HACCP with pathogen reduction came from many sources: federal agencies unwilling to confront powerful const.i.tuents, industry groups willing to accept HACCP only without government oversight (especially of pathogen levels), consumer groups suspicious of the industry's commitment to safety standards and the government's ability to enforce them, and inspectors unwilling to change the nature of their work. By 1994, advocates feared that more lives of children would need to be sacrificed before Congress, the USDA, and the industry would take action to keep dangerous bacteria out of meat. Even the threat of financial liability did not seem severe enough to induce industry action. The inst.i.tution of HACCP rules appeared inevitable to all but the most determined segments of the meat industry, but whether the rules would include requirements for pathogen performance standards and testing remained open for debate. Chapter 3 Chapter 3 explains how that debate developed. explains how that debate developed.

CHAPTER 3.

ATTEMPTING CONTROL OF FOOD PATHOGENS, 19942002

DESPITE THE BARRIERS DISCUSSED IN CHAPTER 2 CHAPTER 2 AND THE AND THE objections raised by businesses likely to be affected by the new regulations, government agencies were eventually able to inst.i.tute HACCP (Hazard a.n.a.lysis and Critical Control Point) systems designed to prevent harmful microbes from getting into food. This chapter describes how that happened, mainly with respect to HACCP controls for beef. Beef industry protests were more vehement and often more effective than those of other industries, and interactions of beef trade a.s.sociations with the U.S. Department of Agriculture (USDA) and with Congress left more visible traces. Because most outbreaks of microbial illness derive from foods regulated by the Food and Drug Administration (FDA), this chapter also explains how the FDA tried to require the industries under its jurisdiction to inst.i.tute HACCP plans, how those industries opposed the plans, and how that agency's systems-once in place-operated in practice. Opponents of HACCP often framed their objections in scientific terms: because cooking kills most food microbes, government intervention is unnecessary. When outbreaks did occur, food producers, processors, and retailers accused each other of causing them, and all blamed government inspectors and consumers. We will see that food companies were not alone in their objections to HACCP requirements. Meat inspectors opposed the new regulations because HAACP changed their work from examining animals to examining paper. In addition, some food safety advocates agreed with the inspectors' contention that HACCP gave too much control of production to industry and allowed foxes to guard chickens, as it were. objections raised by businesses likely to be affected by the new regulations, government agencies were eventually able to inst.i.tute HACCP (Hazard a.n.a.lysis and Critical Control Point) systems designed to prevent harmful microbes from getting into food. This chapter describes how that happened, mainly with respect to HACCP controls for beef. Beef industry protests were more vehement and often more effective than those of other industries, and interactions of beef trade a.s.sociations with the U.S. Department of Agriculture (USDA) and with Congress left more visible traces. Because most outbreaks of microbial illness derive from foods regulated by the Food and Drug Administration (FDA), this chapter also explains how the FDA tried to require the industries under its jurisdiction to inst.i.tute HACCP plans, how those industries opposed the plans, and how that agency's systems-once in place-operated in practice. Opponents of HACCP often framed their objections in scientific terms: because cooking kills most food microbes, government intervention is unnecessary. When outbreaks did occur, food producers, processors, and retailers accused each other of causing them, and all blamed government inspectors and consumers. We will see that food companies were not alone in their objections to HACCP requirements. Meat inspectors opposed the new regulations because HAACP changed their work from examining animals to examining paper. In addition, some food safety advocates agreed with the inspectors' contention that HACCP gave too much control of production to industry and allowed foxes to guard chickens, as it were.

To gain some insight into the basis of this conflict, I wanted to observe a HACCP system in action. Shortly after the USDA's final HACCP rules for meat went into effect in the late 1990s, the owner of a meatpacking plant in New York State agreed to let me visit as long as I did not identify the plant by name.

His company cooks meats under HACCP plans similar to the USDA model ill.u.s.trated in figure 5 figure 5 ( (page 69), and his plant ill.u.s.trates both the strengths and weaknesses of HACCP systems. Production practices followed the prescribed plan to the letter (a strength), but HACCP plans require mountains of paperwork and the attention of a full-time employee (at considerable added expense). Because some of the products include vegetable as well as meat ingredients, they fall under the regulatory requirements of three agencies: the FDA, the USDA, and New York State. Inspections vary in frequency-the USDA daily, New York State four times a year, and the FDA once a year-and are conducted according to the unique rules and reporting requirements of each agency. In practice, the multiple authorities mean that plant officials must fill out three distinct sets of reporting forms (a time-consuming and expensive nuisance). The on-site USDA inspector I met at the plant checked temperature records but seemed entirely uninterested in the production process (a serious weakness, as I will explain). One plant employee confided to me-shades of Upton Sinclair-that "someone could be butchering a dog in front of them [the inspectors], and they wouldn't have a clue."

Because the first of the three critical control points was to measure the temperature of products after they were cooked (see figure 5 figure 5), the managers learned that the ovens were not heating properly. They identified a succession of faults in the engineering of the ovens and tinkered with them until the problem was fixed (a strength). After the products were cooked, however, they were immediately transferred to open racks in a refrigeration room and chilled-uncovered-by cold air blasted in from a ceiling unit. The temperature of the products dropped quickly, as required by the plan (a strength). Unfortunately, the plan did not account for the ability of harmful bacteria like Listeria Listeria to flourish at cold temperatures in the cooling system and to contaminate the uncovered products to flourish at cold temperatures in the cooling system and to contaminate the uncovered products after after they were cooked (a weakness). The plant managers seemed unaware of the potential hazard. Because they followed the HACCP plan so scrupulously, they did not think the uncovered products posed problems (a further weakness). Within weeks of my visit, the company had to recall thousands of pounds of products because some had been found to be contaminated with they were cooked (a weakness). The plant managers seemed unaware of the potential hazard. Because they followed the HACCP plan so scrupulously, they did not think the uncovered products posed problems (a further weakness). Within weeks of my visit, the company had to recall thousands of pounds of products because some had been found to be contaminated with Listeria Listeria.

From this experience, it seemed obvious that HACCP plans can prevent contamination but that diligence in following them is not enough; the plans also must be thoughtfully designed and overseen, and verified by testing. The role of the on-site USDA inspector was particularly striking. He was not involved in promoting the plant's microbial safety. As a 20-year USDA veteran, he had been trained to inspect animals, not paperwork, and was unaware of the peculiar characteristics of newly emergent bacterial pathogens. He seemed much in the tradition of the inspector described in The Jungle The Jungle a century earlier, as we have already seen. a century earlier, as we have already seen.

With this understanding of HACCP systems in action, we can now return to the political battles of the mid-1990s that eventually enabled federal agencies to require some industries to follow HACCP plans. By late 1994, with the legal obstacles out of the way, federal agencies could begin the formal-and glacially tedious-rule-making process to require HACCP controls for one food industry or another: proposing rules in the Federal Register; Federal Register; requesting, collecting, and dealing with public comments; rewriting the proposals; and, eventually, issuing final rules that would go into effect some years later. The USDA proposed rules for meat and poultry, and the FDA proposed rules for some of the foods under its jurisdiction. The two agencies approached the task in quite different ways, particularly in their decisions about whether HACCP plans, performance standards, and requirements for pathogen testing should be required or voluntary. The FDA was the first to place its notices in the requesting, collecting, and dealing with public comments; rewriting the proposals; and, eventually, issuing final rules that would go into effect some years later. The USDA proposed rules for meat and poultry, and the FDA proposed rules for some of the foods under its jurisdiction. The two agencies approached the task in quite different ways, particularly in their decisions about whether HACCP plans, performance standards, and requirements for pathogen testing should be required or voluntary. The FDA was the first to place its notices in the Federal Register Federal Register, and we begin with this agency's approach to HACCP rules.

THE FDA TRIES HACCP, ONE FOOD AT A TIME.

Although up to 80% of outbreaks are caused by foods regulated by the FDA, this agency has had a difficult time figuring out what to do about them. The FDA first proposed to develop HACCP controls for seafood in January 1994 and, in late summer of 1994, asked for public comment on whether and how HACCP might be extended to "land foods." In approaching HACCP regulation in this manner, the FDA was responding not only to the need to control pathogens but also to its own internal difficulties. The resource constraints on the FDA's food safety program seemed unlikely to improve. It also seemed unlikely that the agency would ever be able to inspect, sample, and a.n.a.lyze more than a tiny fraction of the foods for which it was responsible. In deciding how to regulate food safety without funds or personnel, the FDA floated a trial balloon: Although the agency has reached no final conclusions about how its regulatory programs should be revised to make food as safe as possible, FDA has tentatively concluded that the improvements in the agency's current food safety a.s.surance program should be based on a state-of-the-art, preventive approach known as HACCP. . . . The agency has tentatively chosen a HACCP approach because HACCP addresses the root causes of food safety problems in production, storage, transportation, etc., and is preventive.1 The FDA was careful to note that it expected to engage in further conversations with industry and other groups and that it planned to work closely with the USDA to make sure the rules of the two agencies were consistent. It asked for public comment on whether HACCP should be mandatory for all of the food industries under its jurisdiction or just for certain segments of those industries. It also asked for input on how mandatory HACCP plans might apply to the entire chain of food distribution, from production to retail sale.

As a further reflection of this cautious approach, the FDA asked companies to volunteer volunteer to develop pilot HACCP systems. The idea was to study the plans and use them as a basis for deciding how to proceed further. As it turned out, several firms agreed to be guinea pigs for this purpose: Alto Dairy, Campbell Soup, ConAgra, EarthGrains (Sara Lee), Pillsbury (General Mills), and Ralston Foods. Their pilot plans-which involved products such as cheese, frozen dough, breakfast cereals, salad dressing, fresh and pasteurized juices, bread, and flour-demonstrated that HACCP controls worked well to help companies identify safety problems and correct them. This was only to be expected given Pillsbury's experience 30 years earlier. Furthermore, the companies reported a decline in the frequency of product recalls (also to be expected with reduced contamination), along with unantic.i.p.ated benefits in improved production efficiency, employee "owners.h.i.+p" and partic.i.p.ation, and customer satisfaction. None of the volunteers, however, included microbial testing as a component of their plans. The pilot plans demonstrated that HACCP would reduce pathogens and be good for business but also suggested that companies would not test for harmful bacteria unless forced to do so. to develop pilot HACCP systems. The idea was to study the plans and use them as a basis for deciding how to proceed further. As it turned out, several firms agreed to be guinea pigs for this purpose: Alto Dairy, Campbell Soup, ConAgra, EarthGrains (Sara Lee), Pillsbury (General Mills), and Ralston Foods. Their pilot plans-which involved products such as cheese, frozen dough, breakfast cereals, salad dressing, fresh and pasteurized juices, bread, and flour-demonstrated that HACCP controls worked well to help companies identify safety problems and correct them. This was only to be expected given Pillsbury's experience 30 years earlier. Furthermore, the companies reported a decline in the frequency of product recalls (also to be expected with reduced contamination), along with unantic.i.p.ated benefits in improved production efficiency, employee "owners.h.i.+p" and partic.i.p.ation, and customer satisfaction. None of the volunteers, however, included microbial testing as a component of their plans. The pilot plans demonstrated that HACCP would reduce pathogens and be good for business but also suggested that companies would not test for harmful bacteria unless forced to do so.2 While these studies were underway, the FDA proposed HACCP rules for a few additional foods that seemed especially hazardous. It began with HACCP for seafood and sh.e.l.lfish (proposed in 1994 to take effect in 1997), for raw sprouts and eggs (to take effect in 1999), and, as discussed below, for fresh juices (2000). The FDA also initiated another voluntary HACCP experiment, this time for dairy products, and it proposed safe-handling instructions for eggs: "Eggs may contain harmful bacteria known to cause serious illness, especially in children, the elderly, and persons with weakened immune systems. For your protection: Keep eggs refrigerated; cook eggs until yolks are firm; and cook foods containing eggs thoroughly."3 But by 2009, the FDA required HACCP only for seafood. For all its other foods, HACCP was But by 2009, the FDA required HACCP only for seafood. For all its other foods, HACCP was voluntary voluntary. The FDA's food-by-food approach to HACCP, its lack of requirements for microbial testing, and its scarcity of inspectors left industries with many opportunities to avoid installing such plans or adhering to them. For example, only 44% of seafood processing firms had implemented HACCP plans by 1999, and more than half of all inspections revealed serious problems with the plans five years after their implementation. By 2001, the sh.e.l.lfish industry's chief safety strategy-education of consumers-had failed to reduce the illnesses and deaths caused by eating raw products, and neither the industry nor its FDA regulators had imposed preventive measures. This experience reinforced suspicions that voluntary approaches would not work.4 USDA'S POLITICAL BATTLES In contrast, and rather a surprise in view of its past history, the USDA moved quickly to introduce HACCP under the more consumer-friendly leaders.h.i.+p appointed by President Bill Clinton. By the mid-1990s, some segments of the meat industry were asking the department to inst.i.tute HACCP regulations, if for no other reason than to rea.s.sure the public that meat was safe. The Food Safety and Inspection Service (FSIS) began to develop HACCP rules for meat and poultry through a method previously unimaginable for this agency: it openly consulted stakeholders. The FSIS held information briefings, scientific conferences, public hearings, federal and state conferences, agency meetings, and a professional forum to listen to points of view. Its 1995 proposed rules differed from those of the FDA in several critical respects, most notably in their emphasis on requirements for pathogen testing. Indeed, the department called the plan Pathogen Reduction: HACCP Pathogen Reduction: HACCP-a critical distinction. The USDA plan established performance standards and required the companies to prove by daily sampling and testing that pathogenic contaminants did not exceed levels specified in the standards.5 True to form, some meat industry groups objected. An official of the Armour company, for example, told a congressional committee that HACCP was an imperfect system that did not address the real problem-consumer education: "There is a concern that HACCP has been oversold and public expectations may be unrealistically high. In particular, HACCP cannot guarantee the absence of enteric pathogens on raw meat or poultry. . . . Food safety is a shared responsibility involving industry, Government, and consumers. Public education on safe handling of foods continues to be a key factor in preventing foodborne illness."6 Untrue to form, the American Meat Inst.i.tute pet.i.tioned the USDA to require HACCP for all meat and poultry plants: "We believe so strongly in HACCP's benefits for meat and poultry safety that we think it should be mandated for our segment of the industry." This group's unexpected support of HACCP is explained by its a.s.sumption-erroneous, as it happened-that industry industry inspectors would replace those of USDA. Later, when meat industry a.s.sociations realized the implications of pathogen testing-that products found contaminated would be considered adulterated and unfit to sell-they tried to block the proposals. Meat producers and processors much preferred a "virtual" safety system: HACCP without pathogen reduction and entirely voluntary compliance. inspectors would replace those of USDA. Later, when meat industry a.s.sociations realized the implications of pathogen testing-that products found contaminated would be considered adulterated and unfit to sell-they tried to block the proposals. Meat producers and processors much preferred a "virtual" safety system: HACCP without pathogen reduction and entirely voluntary compliance.7 Congress Demands "Negotiated Rulemaking," 1995 Before proposed regulations become final and go into effect, they are supposed to be held open for a specified period of time for public comment. The comment period for the proposed Pathogen Reduction: HACCP rules occurred at a time when especially conservative Republicans had taken control of Congress and were attempting to reduce regulatory burdens on industry. Meat and poultry lobbyists took advantage of this favorable situation to urge Congress to block the proposed rules. They used the usual argument: home cooks are responsible for most episodes of foodborne illness, and oversight of industry is unnecessary. Despite the false premise (most outbreaks derive from foods prepared outside the home), the lobbying succeeded in several respects. First, Congress extended the comment period to give the industry more time to organize opposition. Next, industry lobbyists convinced some members of Congress to amend the appropriations bill to delete funding for HACCP implementation. Finally, while discussions of this funding amendment were in progress, the lobbyists also convinced Congress to order the USDA to partic.i.p.ate in "negotiated rulemaking," a process that would require the department to work closely with meat producers to make the regulations mutually acceptable. To observers offended by the idea that the industry would define its own regulations, the purpose of negotiated rule-making seemed clear-to postpone or eliminate HACCP. Representative George Brown (Dem-CA), explained: "The House Agriculture Committee would like to write more industry-friendly legislation and cut the USDA regulations off at the pa.s.s."8 The congressman who introduced the anti-HACCP funding amendment, James Walsh (Rep-NY), chaired the appropriations subcommittee for agriculture. Mr. Walsh seemed to be acting on behalf of the meat industry-a lawyer for the National Meat a.s.sociation had partic.i.p.ated in drafting his amendment.9 In a further action, Senator Robert Dole (Rep-KS), then majority leader and already campaigning for president, introduced a regulatory reform bill that would require federal agencies to review new regulations likely to cost industry more than $50 million annually, and to demonstrate that the benefits of such regulations would outweigh their costs. One purpose of the Dole bill was to stop government from regulating food safety. It contained provisions to (1) eliminate rules for pathogen testing, (2) postpone seafood inspection, (3) repeal the Delaney clause in the Food, Drug, and Cosmetic Act (which precluded use of carcinogenic food additives), (4) permit use of some carcinogenic pesticides, and (5) privatize approvals of food additives. In a further action, Senator Robert Dole (Rep-KS), then majority leader and already campaigning for president, introduced a regulatory reform bill that would require federal agencies to review new regulations likely to cost industry more than $50 million annually, and to demonstrate that the benefits of such regulations would outweigh their costs. One purpose of the Dole bill was to stop government from regulating food safety. It contained provisions to (1) eliminate rules for pathogen testing, (2) postpone seafood inspection, (3) repeal the Delaney clause in the Food, Drug, and Cosmetic Act (which precluded use of carcinogenic food additives), (4) permit use of some carcinogenic pesticides, and (5) privatize approvals of food additives.

Such blatantly consumer-unfriendly legislation was ripe for satire, and figure 7 figure 7 presents one such pointed commentary, in this case, from political cartoonist Garry Trudeau. Consumer advocate Ralph Nader observed that the Dole bill represented nothing less than a "big business takeover of the U.S. government in its health and safety responsibilities." Nevertheless, after contentious debate, the Senate pa.s.sed various amendments to the Dole bill as part of the Republicans' Contract with America. As if to soften the bill's evident purpose, one such amendment expressed "the sense of the Senate that nothing in the bill is intended to delay the timely promulgation of any regulations that would meet a human health or safety threat." presents one such pointed commentary, in this case, from political cartoonist Garry Trudeau. Consumer advocate Ralph Nader observed that the Dole bill represented nothing less than a "big business takeover of the U.S. government in its health and safety responsibilities." Nevertheless, after contentious debate, the Senate pa.s.sed various amendments to the Dole bill as part of the Republicans' Contract with America. As if to soften the bill's evident purpose, one such amendment expressed "the sense of the Senate that nothing in the bill is intended to delay the timely promulgation of any regulations that would meet a human health or safety threat."10 Mr. Walsh's industry-driven appropriations amendment was also under consideration, but the New York Times New York Times urged opposition: "By voting to defeat Mr. Walsh's amendment today, the Appropriations Committee would send a welcome signal that it cares more about protecting const.i.tuents' health than about pleasing the meat and poultry industries." Consumer advocates from the Was.h.i.+ngton, DCbased Center for Science in the Public Interest (CSPI) wrote that the Walsh proposal was "just a smoke screen to give businesses free rein to do business as usual-even if that means killing innocent children." urged opposition: "By voting to defeat Mr. Walsh's amendment today, the Appropriations Committee would send a welcome signal that it cares more about protecting const.i.tuents' health than about pleasing the meat and poultry industries." Consumer advocates from the Was.h.i.+ngton, DCbased Center for Science in the Public Interest (CSPI) wrote that the Walsh proposal was "just a smoke screen to give businesses free rein to do business as usual-even if that means killing innocent children."11 Late in June, the House committee pa.s.sed the Walsh amendment, making it clear that it was doing so to give "meat packers a chance to win relief from new food-safety regulations."12 This meant that if the Senate also pa.s.sed the amendment, the USDA would not be able to issue HACCP rules until it completed its "negotiated rulemaking" conversations with meat and poultry processors. This possibility inspired further editorial comment in the This meant that if the Senate also pa.s.sed the amendment, the USDA would not be able to issue HACCP rules until it completed its "negotiated rulemaking" conversations with meat and poultry processors. This possibility inspired further editorial comment in the New York Times New York Times: [image]

FIGURE 7. The political cartoonist Garry Trudeau had this to say about Senator Robert Dole's attempt to deregulate the meat industry. Mr. Dole was expected to run for president in the next election. (Doonesbury, August 20, 1995, 1995 G.B. Trudeau. Reprinted with permission of Universal Press Syndicate. All rights reserved.) Two things will happen to anyone who takes a close look at the way meat is processed and inspected in this country: they will wonder how it is that even more people are not made sick by tainted meat, and they will get sick to their stomachs themselves. . . . Naturally, the meat industry and its stooges in the Republican Party have ganged up on the Agriculture Department (and the American consumer) to make sure the new inspection system never sees the light of day. . . . [Negotiating rulemaking] is like negotiating prison rules with convicts. . . . Children will continue to die in excruciating pain because the meat they ate was contaminated, and because unscrupulous Republicans in Congress fought aggressively to keep it that way.13 At this point, Mr. Walsh suddenly withdrew his amendment, attributing this surprising retreat to a revelation that Congress could in fact work out its differences with the USDA: "We got the personal commitment of the secretary to create the dialogue we sought."14 Alternative explanations seem more likely, however. Pressure from advocacy groups was surely a factor, especially a campaign organized by the families of children who became ill or died after eating contaminated hamburger. Advocates took credit for the amendment's withdrawal as a "resounding victory for public health and an unmasking defeat of good ol' boy politics." This last was a reference to press accounts that Mr. Walsh had accepted $66,000 in donations from meat and agricultural interests. Alternative explanations seem more likely, however. Pressure from advocacy groups was surely a factor, especially a campaign organized by the families of children who became ill or died after eating contaminated hamburger. Advocates took credit for the amendment's withdrawal as a "resounding victory for public health and an unmasking defeat of good ol' boy politics." This last was a reference to press accounts that Mr. Walsh had accepted $66,000 in donations from meat and agricultural interests.15 It also seems likely that members of Congress, not wanting to be viewed as destroyers of public health and killers of innocent children, suggested that Mr. Walsh would face a difficult floor fight if he pursued his anti-HACCP agenda. Ultimately, the Dole bill also failed to pa.s.s. Thus, one unantic.i.p.ated-and positive-result of Mr. Walsh's amendment was to unite food safety advocates and encourage them to press for an independent food agency that would not be subject to such cra.s.s political pressures, an issue discussed in greater detail in It also seems likely that members of Congress, not wanting to be viewed as destroyers of public health and killers of innocent children, suggested that Mr. Walsh would face a difficult floor fight if he pursued his anti-HACCP agenda. Ultimately, the Dole bill also failed to pa.s.s. Thus, one unantic.i.p.ated-and positive-result of Mr. Walsh's amendment was to unite food safety advocates and encourage them to press for an independent food agency that would not be subject to such cra.s.s political pressures, an issue discussed in greater detail in chapter 4 chapter 4.16 The Last Attempts to Derail HACCP, 19951996 With the demise of the Walsh amendment, the USDA released the nearly 200 pages of "final" rules for Pathogen Reduction: HACCP for meat and poultry products. As with the earlier drafts, these would need to be made available for public comment before going into effect in July 1996. The rules required large firms to develop, install, and implement HACCP plans by the beginning of 1998, small firms by 1999, and very small firms by 2000. To help companies figure out how to proceed, the department created 13 model plans and provided detailed instructions for developing and using them (figure 5 in in chapter 2 chapter 2 is based on one such model). is based on one such model).17 The published Pathogen Reduction: HACCP rules revealed that political pressures succeeded in achieving at least one compromise. Although the USDA originally wanted meat and poultry companies to be responsible for The published Pathogen Reduction: HACCP rules revealed that political pressures succeeded in achieving at least one compromise. Although the USDA originally wanted meat and poultry companies to be responsible for Salmonella Salmonella testing, it now said that federal inspectors would test for testing, it now said that federal inspectors would test for Salmonella Salmonella "on an unannounced basis." Companies would have to test for the "on an unannounced basis." Companies would have to test for the generic generic form of form of E. coli E. coli (as a marker of fecal contamination) in just a small number of samples: 1 out of every 300 beef carca.s.ses, 1,000 hogs, 3,000 turkeys, and 22,000 chickens. (as a marker of fecal contamination) in just a small number of samples: 1 out of every 300 beef carca.s.ses, 1,000 hogs, 3,000 turkeys, and 22,000 chickens.5 This time, meat processors used the comment period to press Congress to eliminate requirements for Salmonella Salmonella testing. Their congressional sympathizers introduced an amendment to the Farm Bill that would create an "independent" oversight panel of food, meat, and poultry scientists with broad powers to review FSIS decisions on HACCP procedures, standards, and practices. The amendment required the USDA to submit proposed rules to the panel and then allow 90 days for public comment. At the very least, this plan would further delay the regulations. The policy director of the House Agriculture Committee explained that the purpose of the panel was testing. Their congressional sympathizers introduced an amendment to the Farm Bill that would create an "independent" oversight panel of food, meat, and poultry scientists with broad powers to review FSIS decisions on HACCP procedures, standards, and practices. The amendment required the USDA to submit proposed rules to the panel and then allow 90 days for public comment. At the very least, this plan would further delay the regulations. The policy director of the House Agriculture Committee explained that the purpose of the panel was scientific: scientific: "The H.A.C.C.P. rule is purported to be science-based, but it seemed clear that some of the regulatory decisions have no scientific validity. The regulations should never have gotten to this stage." "The H.A.C.C.P. rule is purported to be science-based, but it seemed clear that some of the regulatory decisions have no scientific validity. The regulations should never have gotten to this stage."18 Reporters, however, viewed the panel as yet another political tactic to allow the meat industry to avoid having to test for pathogens. Eventually, this amendment failed to get enough votes to be included in the final bill. Reporters, however, viewed the panel as yet another political tactic to allow the meat industry to avoid having to test for pathogens. Eventually, this amendment failed to get enough votes to be included in the final bill.

Sorting out the political forces for and against Pathogen Reduction: HACCP is especially complicated because USDA inspectors also opposed the regulations. They had experienced a huge increase in "tasks not performed" just since the year before, meaning that their workloads had increased to the point where they could not complete a.s.signments. One inspector, for example, complained that he was supposed to inspect 16 meat plants within a six-hour period. Once HACCP went into effect, workloads would be reduced (a benefit), but the inspectors did not like the idea that their jobs would change from inspecting animals and meat products-useless for preventing microbial contamination as that might be-to inspecting paperwork. The on-site USDA inspector at the meatpacking plant I visited, thoroughly disenchanted with having to deal with that paperwork, was counting the days until he could retire. Old-time inspectors quipped that because HACCP minimizes risk but can never a.s.sure absolute safety, its initials really should stand for "Have a Cup of Coffee and Pray."19 The General Accounting Office (GAO), however, continued to press for HACCP, also invoking science as the rationale. For years it had been issuing reports urging Congress to require oversight of food safety based on science (meaning a.s.sessment of hazards at critical points and evaluation by microbial testing) rather than sensory perception (poke-and-sniff). GAO officials complained that federal agencies repeatedly ignored their warnings and were still inspecting foods by obsolete systems that could not possibly address modern microbial hazards.20 In the case of meat and poultry regulation, reason eventually won out over politics, the benefits of HACCP prevailed, and the USDA issued its final set of regulations in July 1996-an election year in which the political climate had s.h.i.+fted to one more friendly to consumer interests. Perhaps as a sign of that change, the White House Office of Management and Budget (OMB), which typically opposed expansion of government regulations, praised the new rules: "For years, we have had the Government doing the work, the inspectors in the plants, and you hear stories of cursory checks and that's it. . . . This is an attempt to get away from Government micromanaging the process and instead saying to the regulated ent.i.ty, 'you figure out how to do it, you're responsible, and we'll do some testing to make sure there are performance standards.' "21 An alternative reason for the OMB's blessing may have been financial. USDA economists calculated that the economic benefits of Pathogen Reduction: HACCP would outweigh its costs even under the most conservative estimates. Although the new rules were expected to cost industry more than a billion dollars over a 20-year period, the economic benefits to society would exceed that amount even if just 5% of foodborne illnesses could be prevented. Economists thought that if HACCP could achieve a 90% reduction in illnesses caused by just the six most common food pathogens, the 20-year savings in medical costs and lost productivity would be $170 billion or more.22 Furthermore, the meat and poultry industries would share financial benefits in the form of reduced recalls and liability, enhanced consumer confidence, and robust sales. Furthermore, the meat and poultry industries would share financial benefits in the form of reduced recalls and liability, enhanced consumer confidence, and robust sales.

By mid-1996, the HACCP schedule was in place. The E. coli E. coli O157:H7 testing of ground meat proposed in 1994 was due to start in January 1997. Large companies were to install Pathogen Reduction: HACCP plans by January 1998, and smaller companies by January 2000. From then on, all meat and poultry companies were to follow such plans and test for generic O157:H7 testing of ground meat proposed in 1994 was due to start in January 1997. Large companies were to install Pathogen Reduction: HACCP plans by January 1998, and smaller companies by January 2000. From then on, all meat and poultry companies were to follow such plans and test for generic E. coli E. coli, endure USDA testing of ground meat for E. coli E. coli O157:H7 and spot-checking for O157:H7 and spot-checking for Salmonella Salmonella, and meet performance standards for pathogen reduction. Food safety would depend on how carefully they designed and implemented the plans, and how well the USDA enforced them. For most foods regulated by the FDA, however, HACCP remained voluntary. As we will now see, events soon revealed serious gaps in the regulations, and indicated additional needs: for extension of HACCP rules to all food products at all stages of production, for federal authority to recall contaminated products, and for ways of countering the culture of entrenched resistance to government oversight so prevalent in the meat industry.

THE PRODUCT GAP: ODWALLA APPLE JUICE, 1996.

In October 1996, an outbreak of E. coli E. coli O157:H7 made it clear that O157:H7 made it clear that all all foods needed to be produced under Pathogen Reduction: HACCP because lapses could be catastrophic, not only for the victims, but also-temporarily, if not permanently-for the companies responsible for them. This outbreak put about 70 people in hospitals, made 14 children dreadfully ill, and resulted in the death of one child. foods needed to be produced under Pathogen Reduction: HACCP because lapses could be catastrophic, not only for the victims, but also-temporarily, if not permanently-for the companies responsible for them. This outbreak put about 70 people in hospitals, made 14 children dreadfully ill, and resulted in the death of one child.23 The unexpected feature of the outbreak was its source: apple juice. Investigators used genetic techniques and diet histories to trace the infections to freshly bottled apple juice produced by Odwalla, a California company specializing in "natural" foods. The company included windfall apples, those that had fallen off the trees onto the ground, among the lots pressed to make the juice, and investigators suspected that the fallen apples must have come in contact with animal manure containing E. coli E. coli O157:H7. Although apple juice is naturally acidic, its acids are not strong enough to kill this hardy microbe. Odwalla did not pasteurize its juices; its managers believed that temperatures high enough to kill most bacteria would alter the flavor of the juice and reduce vitamin content (which pasteurization does, but only slightly). The managers also believed-gravely in error-that the acidity of the solutions used to wash the apples and of the juice itself would kill harmful bacteria. O157:H7. Although apple juice is naturally acidic, its acids are not strong enough to kill this hardy microbe. Odwalla did not pasteurize its juices; its managers believed that temperatures high enough to kill most bacteria would alter the flavor of the juice and reduce vitamin content (which pasteurization does, but only slightly). The managers also believed-gravely in error-that the acidity of the solutions used to wash the apples and of the juice itself would kill harmful bacteria.24 Investigators made other disheartening observations. Just before the outbreak, the company had relaxed its standards for accepting blemished fruit. It overruled warnings from its own in-house inspector not to use the batch of apples responsible for the outbreak without special precautions. At the time, Odwalla was expanding rapidly and having difficulty meeting production demands. Its stock price was declining. These pressures also contributed to the company's failure to follow its own established procedures.25 To their great credit, Odwalla officials quickly took full responsibility for the failure of their safety systems and issued a recall. They paid medical expenses for the people who had become ill and an indemnity of about $250,000 to the family that had lost a child. Eventually, they settled more than a dozen civil suits at a cost of more than $12 million-just for the families of the five children who had been most injured. They also paid in other ways. In the first criminal conviction recorded in a large-scale outbreak of foodborne illness, Odwalla officials pleaded guilty to violating federal food safety laws, paid a $1.5 million fine, and were placed on probation for five years.26 Odwalla's corporate policy includes an explicit statement of social responsibility to employees and customers. Its officials immediately admitted wrongdoing and wrote checks. They also took action to improve production practices. Odwalla now flash-pasteurizes its juices (high temperature, short time), uses a HACCP plan, and actively promotes its quality-control efforts. These actions restored consumer confidence. By 1999, sales had almost returned to former levels, and by 2001 the company was comfortably profitable. The actions also restored investor confidence. In 2000, Odwalla merged with Fresh Samantha, another fresh juice company. A year later, ironically, this producer of fresh, "healthy" juices was purchased by Coca-Cola, the world's largest soft drink company, in a deal said to be worth $181 million.27 Among the many lessons of the Odwalla outbreak is the vulnerability of the nation's supply of fruit and vegetables to cross-contamination from infected animals, pointing even more forcefully to the need to prevent foodborne illness at its source. The Odwalla incident induced the FDA to demand a warning label on unpasteurized juices. Manufacturers would have to prove that their production practices achieved a "5-log" (100,000-fold) reduction in the number of dangerous contaminants in their juice products. Otherwise, juice labels would have to display this statement: "WARNING: This product has not been pasteurized and, therefore, may contain harmful bacteria that can cause serious illness in children, the elderly, and persons with weakened immune sys