NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Food Forum. Managing Food Safety Practices from Farm to Table: Workshop Summary. Washington (DC): National Academies Press (US); 2009.

Cover of Managing Food Safety Practices from Farm to Table

Managing Food Safety Practices from Farm to Table: Workshop Summary.

Show details

3The Complexities of Food Safety and Some Strategic Approaches Being Taken

Moderator Janet Beauvais1 opened the second session, Strategic Approaches to Outbreak Control, with a comment on the importance of sharing best practices and integrating not just nationally but also internationally. Indeed, the theme of the global nature of food safety and the importance of international integration figured much more prominently in this session than it did in the first session, with the first presentation, Robert Brackett’s Industry Perspective on Managing Risks in a Global Economy, revolving around the necessity of adopting a global supply chain management approach toward food safety risk management. Brackett, Senior Vice President and Chief Science and Regulatory Affairs Officer of the Grocery Manufacturer Association (GMA), argued that even “made in the U.S.” food products are global products, with suppliers for many products coming from multiple countries worldwide. A key message of Brackett’s presentation was that because of the global nature of our food supply, no single company (or country) holds entire responsibility. Ensuring that our food products are safe requires that all components of the global system—from producer to consumer—be functioning the way they are intended to function. Accordingly, GMA has developed and adopted a “Four Pillars of Food Safety” program with a focus on global “supply chain management” and foreign-supplier quality assurance.

Shifting gears a bit, rather than revolving around the complexities of the global nature of the food supply chain, the second presentation of this session, Russell Flowers’s Technological Improvements in Outbreak Prevention, revolved around the complexities of the technological nature of outbreak prevention. Flowers, of Silliker, Inc., considered the range of technologies available or being developed for outbreak prevention and the knowledge required to minimize or prevent contamination events. He emphasized the importance of data and re-iterated earlier claims made during the day that even as testing technology advances; testing is not prevention. One of his key messages was that rather than expecting some of the new molecular and other advanced detection technologies to “save us,” we need to better utilize the tools and knowledge that we already have. He pointed to plant facility layout as an area where improvements could be made.

Returning to the global theme, the third presentation of the session, Julie Caswell’s Roles and Responsibilities of Industry and Government in Managing Relationships with Global Food Suppliers, revolved around the reality that risk management for food safety is a complex endeavor not just because of the inherent risks associated with food production but also because of the diverse nature of our international food supply chains. Caswell provided a conceptual overview of the range of combined public-private approaches to food safety management being used in the United States and elsewhere. Caswell commented that, unlike Canada and the United Kingdom, the United States makes food safety decisions on a risk-by-risk basis. She asked whether it might be possible and beneficial to develop a more comprehensive approach that would allow for more predictive, rather than reactive, food safety decision making.

This chapter provides summaries of these three presentations and the discussion that followed.


Presenter: Robert Brackett2

Robert Brackett began by remarking that globalization is not something that is coming. It is here. It has changed our view of (1) what food is, (2) what food safety is, and (3) how businesses practice and manage food risks. He addressed each of these in turn, with most of his presentation focusing on the last component: how businesses have changed the way they practice and manage risks, with a growing emphasis on global supply chain management.

As an example of how globalization is changing our view of what food is, consider a loaf of bread that is “made in the U.S.” but which is comprised of seven ingredients that could have come from any number of different places on any given day:

  1. Wheat gluten from France, Poland, Russia, the Netherlands, or Australia
  2. Honey from China, Vietnam, Brazil, Uruguay, India, Canada, Mexico, or Argentina
  3. Calcium propionate from the Netherlands
  4. Guar gum from India
  5. Flour enrichments from China
  6. Beta-carotene from Switzerland
  7. Vitamin D3 from China

What ends up being a U.S. product is truly a global product. This loaf of bread is not a unique example, Brackett noted. It is very common, and it complicates how food safety risks are addressed. For example, from 2003 to 2007, U.S. imports of fruits and fruit preparations increased 18 percent; grains and feeds, 40 percent; vegetables and vegetable preparations, 23 percent; oilseeds and products, 63 percent; dairy products, 14 percent; tree nuts and preparations, 18 percent; and poultry and poultry products, 54 percent. The only flat trend has been for livestock and meats, with a slight negative (−1 percent) change between 2003 and 2007. Overall, the last several years have seen a 29 percent increase in U.S. food product imports. Today, more than 20 percent of all U.S. imports are food products, amounting to more than eight million shipments annually. Moreover, about 40 percent of all trade in agriculture, fisheries and forestry occurs between developed and developing countries, which presents its own subset of challenges.

While the specific implications of globalization for food safety are uncertain and continually evolving, Bracket emphasized that one thing is certain: globalization of the food supply means that all countries must share responsibility for food safety throughout the entire food supply chain, from producers to consumers. Ensuring that our food products are safe requires every component of the global food supply system be functioning the way it is intended to function. No single company or country holds entire responsibility.

Brackett made a few comments about consumer confidence and referred to some data analyzed by the FMI showing that the number of consumers who are either “completely” or “somewhat confident” that supermarket food is safe has gone down from 82 percent in 2006 to 66 percent in 2007. According to similar data collected by the GMA in 2008, 52 percent of consumers surveyed reported that they were “more concerned” with the safety of foods produced in the United States compared to two or three years ago. Only 4 percent are less concerned, and 44 percent reported no change in their level of concern. The numbers are more alarming for imported foods. Compared to two or three years ago, 68 percent of consumers surveyed reported being “more concerned” about the safety of imported foods. Only 3 percent were less concerned, 28 percent reported no change in their level of concern, and 1 percent replied that they “don’t know.” Obviously, globalization is changing the confidence people have in their food supply. Not having confidence in the global food supply means that not only are consumers not trusting food manufacturers or food in general, but they also have less confidence in regulatory and other government agencies and are less likely to react to government advice.

Brackett then went on to discuss the nature of the risks being faced—both what those risks are and why they exist. He identified three main categories of risk:

  1. Microbiological
    1. e.g., Salmonella in jalapeño peppers
    2. e.g., Cyclospora in raspberries
  2. Chemical
    1. Unapproved pesticides, which have been particularly problematic in dietary supplement ingredients
    2. Heavy metals, such as lead in candy
    3. Unapproved chemotherapeutics, such as chloramphenicol being detected in honey and fish
    4. Undeclared allergens
  3. Physical, including anything from glass to rocks

He listed several reasons why these risks exist:

  1. Non-adherence with GAPs/Good Manufacturing Practices (GMPs)/Good Aquaculture Practices (AqPs)
  2. Poor sanitation (e.g., often something as simple as not having a screen over a window or having appropriate toilet facilities)
  3. Poor water quality
  4. Economic adulteration (e.g., the use of melamine to obtain an economic advantage, the use of an ingredient of lesser quality)
  5. Counterfeiting (i.e., not only of the products themselves but also certificates indicating that an imported product has met a particular standard or audit)
  6. Intentional contamination
  7. Terrorism
  8. Industrial sabotage

The remainder of Brackett’s presentation focused on how to manage these risks, with an emphasis on “supply chain management.” He began by reiterating Daniels’ point that audits are not enough and that food safety risk management requires a higher level vision or strategy. To this end, GMA has developed a comprehensive set of risk management foundational elements (many of which Daniels had identified in his presentation as being important for the produce industry in particular):

  • Industry responsibility: Food safety and consumer protection is ultimately the responsibility of industry, regardless of regulations or legal mandate, although industry is not the only sector with responsibilities.
    • Industry needs to establish preventive—not reactive—food safety programs, even in the absence of regulation (i.e., we must go beyond what is expected by the regulatory authorities).
    • Regulatory agencies need to provide incentives, not disincentives, for industry to enhance their safety programs and increase compliance, such as decreased inspection frequencies.
    • We need more collaboration between industry and regulatory agencies so that appropriate risk management decisions can be made down the line rather than companies being afraid to reveal too much information because of the likelihood of facing a punitive action or having information shared with their competitors.
    • Industry needs to manage the supply chain in a way that they have not done in the past.
  • Adequate funding: Regulatory agencies need to have the appropriate amount of funding in order to do their job.
  • Adequate authority: Similarly, regulatory agencies need adequate authority in order to do their job.
  • Adequate and effective training: Both the private sector plants and public sector regulators need adequate and effective training.
  • Risk-based: Decisions and strategies need to be risk-based so that the appropriate resources are being focused where they will do the most good.
  • Leveraged resources: The federal government needs to leverage state and industry resources more than it has been doing.
  • Research: We need to know what the risks and consequences of our actions are.

Before elaborating further on one singularly important component of this set of foundational elements—the need for industry to manage the supply chain in a way that they have not done in the past—Brackett briefly described GMA’s Four Pillars of Food Safety program, a strategic plan developed in 2007 for addressing industry responsibility for a safe global food supply. The four pillars are:

  1. A mandatory foreign supplier quality assurance program, whereby suppliers must have written food safety plans that can be verified by both the importer and government.
  2. A voluntary qualified importer food safety program to serve as an incentive to those companies willing to share data and ensure that they are going above and beyond what regulatory agencies expect (i.e., the incentive is that products will be imported more quickly).
  3. Capability building: foreign focus: While there are some government programs already actively engaged in this type of activity, the private sector should be contributing as well. In fact, some companies are doing this already, for example by ensuring that products are safe before being exported instead of after being imported.
  4. Capacity building: U.S. broader focus: The regulatory agencies need to have the resources and authorities they need in order to do their job.

Brackett defined “supply chain management” as “due diligence to assure that products received from suppliers meet required regulatory, legal and contractual standards of safety.” He listed several reasons why supply chain management is important:

  • Legal considerations
    • Companies must comply with regulations and laws.
    • Liability poses another legal restraint.
  • Financial considerations
    • As learned from the melamine crisis, cheaper is not necessarily less expensive since, in the long run, being cheap could cost not just a company but an entire country considerably more than it would have if the more expensive, higher-quality ingredient had been chosen initially.
    • Both commercial and country brand values are important considerations.
      • Many companies would rather discard millions of dollars worth of food in order to save their brand and their reputation.
      • Countries themselves have brands to protect: when a country is associated with a foodborne illness, it often becomes very difficult to reassure customers that the country is doing anything right from that point onward.
  • Consumer expectations and an implied contract with consumers that products are safe
    • Consumers expect and assume that the products they purchase are going to be safe and, when things go wrong, they put responsibility and blame on both the company that sells the product and the exporting country.

GMA has issued a document, the GMA Food Supply Chain Handbook, published in five different languages (English, Spanish, Mandarin, French, and Russian), that includes a checklist of things that buyers should expect of their suppliers in an effort to “raise the [safety] bar.” While the handbook includes many elements that one expects (e.g., HACCP, GMP, Good Handling Practices [GHP], GAP, and GAqP), it also emphasizes employee training, U.S. regulatory compliance, the importance of recall programs and the importance of product testing.

In conclusion, Brackett emphasized three key points:

  1. International trade in foods and agricultural products will continue to increase. We will have to address this. There is no way we can go backward at this point.
  2. Managing the supply chain will be essential to assuring safe products and consumer confidence. We will have to find new and different ways of doing this and not just rely on audits, certificates and other traditional tools.
  3. The government and private sector MUST cooperate if we are ever going to solve the problem of food safety and continue to provide safe products for American consumers.


Presenter: Russell Flowers3

Rather than run through a long list of new technologies and processing techniques, Russell Flowers remarked that the focus of his presentation would be on microbial ecology during the food production process and opportunities for technological applications that disrupt that ecology. He noted that there are situations where technological opportunities for prevention exist but are not being utilized.

But first, what is it that we are preventing? The answer: contamination of foods with pathogens. We do this in three ways, which Flowers referred to as the “3 Ks”:

  1. Keep them out. This is particularly important for products that do not undergo a lethal process. There are several ways to do this:
    • Focus on the raw materials/ingredients by practicing GAP with both plants and animals, building quality relationships with suppliers instead of relying on a purchase product mentality, and certifying critical (i.e., at risk) ingredients.
    • Focus on the process environment (e.g., as the meat industry has done with Listeria, which Huffmann addressed in his presentation) by improving plant and equipment design and practicing GMP and GHP.
    • Prevent recontamination of processed product.
  2. Kill them. Heat is often used for this purpose, as Zink elaborated earlier. Ideally, all products would have a kill step in their in-use containers with no further contamination subsequent to the kill step. There are several technologies available for this purpose:
    • Traditional technologies for killing pathogens include the use of heat, pH/acidity, sanitizers/biocides and irradiation.
    • Newer technologies available for killing pathogens include the use of pulsed electric fields, pulsed light, high power ultrasonics, cold plasma, ohmic heating, UV light, microwave, high pressure, and new biocides acceptable for direct food contact. To date, application of most of these newer technologies is limited to certain types of products.
  3. Keep them from growing. In cases where pathogens are present, the goal is to keep those pathogens from growing. Again, there are several ways to do this:
    • Through formulation and packaging (e.g., pH, aw, Eh).
    • Through temperature control (either high or low but not in the middle).
    • Through moisture control.

Flowers identified several tools available for accomplishing the 3 Ks: GAP, GMP, GHP, HACCP, cleaning and sanitation, plant and equipment design, and personnel and maintenance practices.

Flowers then posed two questions: (1) What are some of the problems we face, and (2) how might technology be used to solve some of these problems? First, the problems:

  • Contamination of pathogen-free tissue during harvest and primary processing. This is a critical issue for the meat industry in particular.
  • Contamination of raw material where there is no kill step in the process or preparation. There are several examples of raw material ingredients that don’t have kill steps and for which this is a common problem:
    • E. coli O157:H7 in fresh meats.
    • Salmonella and E. coli O157:H7 in produce.
    • Salmonella in dry ingredients for chocolate, dry blend diet drinks, etc.
    • L. monocytogenes in fresh produce and meats.
    • Campylobacter in fresh poultry.
  • Process failure: This is often due to a HACCP deviation or recontamination subsequent to the kill step and before packaging.
  • Post-process contamination. Again, there are several examples of common post-process contamination problems:
    • L. monocytogenes in cooked meats. (Flowers noted that as Huffman elaborated, the meat industry has made significant progress in controlling L. monocytogenes in RTE meats, but the process is not foolproof.)
    • L. monocytogenes in pasteurized dairy products.
    • Salmonella in dry milk products. (You won’t find a long list of Salmonella outbreaks in dry milk products, Flowers commented, but that doesn’t mean that contamination events are not happening prior to products reaching the market. One of Flowers’s responsibilities as a practicing microbiologist is to consult with food companies and test products early on, to prevent outbreaks.)
    • Salmonella in dry cereals and pet foods. (These are usually due to moisture control issues, which lead to products being re-contaminated after the kill step.)
  • Microbiological testing needed to quantify and manage risks.

Most importantly, Flowers emphasized, the microbiological safety of many products is dependent upon preventing the introduction of pathogens during harvest/slaughter and in the post-kill process and packaging environment. In the post-process environment, this means not only preventing the growth of pathogens but also preventing the spread of pathogens from contaminated areas to produce and product contact areas. In order to accomplish this, the ideal food plant layout has a separate raw processing area (where all the cutting, sorting, etc., are conducted), a kill step between this area and the next, and then a separate packaging area (see Figure 3-1). Flowers explained that this ideal food plant layout, with complete separation between the raw and processed product, would minimize one of the most important sources of contamination: the introduction of pathogens in the post-kill process and packaging environment. In actuality, this is not a common layout; most existing facilities have a lot of back-and-forth movement of product and situations where cooked product that has already gone through the kill step is subject to recontamination. Flowers mentioned that, as Huffman alluded in his presentation on Listeria in RTE meat products, there are ways to manage this problem, but it requires diligence and a redesign of the plant and the process flows. Some of the industry is doing this very well, Flowers said, but some is not. Unfortunately, one company’s failure to manage this problem can affect the reputation of the entire industry. Flowers mentioned that the same problems occur in home and food service operations, where there is a lot of back-and-forth movement of foods and potential exposure of cooked products to contaminated areas.

FIGURE 3-1. The ideal food plant layout, as presented by Flowers.


The ideal food plant layout, as presented by Flowers.

Flowers then addressed the second question: How do we combat this microbial growth? First, consider what microorganisms need for growth: moisture, temperature, time, nutrients, and the absence of inhibitors. Preventing growth or recontamination in those areas requires controlling each factor:

  • Moisture control
    • This limits not only growth but also distribution (e.g., the aerosols created when a forklift travels along a wet floor or when cleaning in one area of the plant creates aerosols that can travel to another area of the plant).
  • Temperature control
    • Reducing the temperature can slow or stop growth.
    • Elevating the temperature accelerates growths but only until it reaches a lethal limit, beyond which it can stop growth.
  • Time, nutrients, and inhibition control through cleaning and sanitation
    • The ongoing process of cleaning and sanitation starts the clock and limits time, which is important even in refrigerated environments where Listeria and other pathogens can grow. In many plants, however, sanitation schedules are set according to shift changes (e.g., every other shift) and not according to what the data indicate the best frequency would be. If the frequency is not fine-tuned, even if a single cleaning and sanitation is effective at that point of time, nutrients and microbes can build up quickly between cleanings.
    • Routine cleaning and sanitation remove nutrients.
    • Cleaning and sanitation also introduce inhibitors.

Next, Flowers listed several technical steps necessary for minimizing the severity of any single contamination event:

  • Must identify and characterize the causative agent.
    • There have been many advances in test methodology allowing for faster, more sensitive detection (within 8 to 24 hours). Other technological advances are in the pipeline.
    • Molecular advances have led to the availability of genotyping kits for use by industry. Flowers noted that the technologies upon which these newer industry-available techniques are based have in fact been around “for a while,” in academic and government labs. For example, PulseNet, a CDC program, is a national network of labs contributing to a pulsed-field gel electrophoresis [PFGE] database for use in the identification of pathogenic strains.
  • Need to determine the source of contamination of process failure.
  • Must have traceability of the product and its ingredients (i.e., not necessarily traceability for recall but traceability of each ingredient in a product).
  • Must be able to differentiate between good, suspect, and bad products or product lots, which require having “break points” as well as good outbreak and recall management.
  • Must be able to clean up a plant and document effectiveness.
  • Must establish and validate preventive measures for the future.
  • Must be able to provide verification for initial product runs after a contamination event.

More generally, the industry needs to develop data necessary for understanding the microbial ecology of the process environment; and it needs to use those data to educate employees about microbial ecology, develop criteria for pathogen testing, improve manufacturing facilities, and improve and verify process flow. Also, routine testing is important. Flowers noted that Silliker has encountered many problems that could have been prevented through routine testing. While testing generates large quantities of negative information, it does identify events that require corrective actions.

Flowers elaborated on the need for break points. While challenging, lots must be defined, good data records kept, and break points established. Accomplishing these tasks is particularly difficult for continuous processes, like milk drying and flour operations which can run for a month or more without stopping and where it is difficult to draw a line. It is also a problem for rework situations (e.g., a carcass coming back into a plant). The issue is often further complicated by the lack of reliable test data in the event of an incident or outbreak.

Flowers listed a set of six steps that companies should take:

  1. Plan ahead for a potential contamination event.
  2. Develop rational break points.
  3. Minimize lot sizes to the extent economically possible.
  4. Establish processes and related documentations (e.g., records retention policies).
  5. Develop a sampling plan (i.e., one focused not so much on lot acceptability but process verification).
  6. Test using validated method(s) and credible (i.e., third-party accredited) laboratories. (Flowers noted that while this may seem a biased remark, given his affiliation with Silliker, in fact companies invalidate testing data.)

In his final remarks, Flowers emphasized that while new technologies for killing pathogens are being developed and validated, these technologies will not prevent contamination. While the new testing and traceability technologies are going to continue to improve with respect to speed and sensitivity and while the genetic characterization of pathogens is becoming easier and more automated, we are not going to see any time in the near future a “gun we can shoot at a carcass” indicating the presence of E. coli or a particular strain of E. coli. Even with this new testing technology, testing is not preventative; its best use is as a way to verify ongoing control. Rather than focusing so much on new technologies that “are going to save us,” Flowers urged better utilization of the tools and knowledge that are already available and adoption of new technologies as they become available. The entire industry needs to be doing this, not just the vast majority. Even just a small portion of companies not making the effort can cause damage to the reputation of the entire industry.


Presenter: Julie Caswell4

Julie Caswell began her presentation by noting “the obvious”:

  • First, food safety is a joint effort, with the private sector producing food safety but with government providing oversight and regulation. The private market not only responds to the public regulatory system (e.g., by finding ways to reduce compliance costs) but also, in some cases, gets ahead of regulation (e.g., by differentiating products and adding value based on safety or through first mover advantage). Therefore, the risk management environment encompasses a mixed mode of private and public responsibility.
  • Second, in most countries around the world, there has been an ongoing shift in the food safety regulatory approach, from a “command and control” approach to a more “performance” approach that puts the responsibility for food safety production more directly and strongly on the food business operators themselves and which has led to a rapid development of private standards.
  • Third, when examining food safety, we are dealing with a very complex and demanding policy space that involves both public and private sector incentives and controls.

Caswell expressed the importance of taking a “big picture” look at control and management of food safety risks, noting three key features worth examining:

  1. The complexity of risk management for food safety:
    1. The set of risks is complex, with varying sources of origination, transfer, and magnification along the supply chain.
    2. Supply chains are diverse, with both international and domestic food supply chains shifting quickly and in a decentralized manner and often in a way that cannot be picked up on very rapidly at a systematic level.
  2. The need to focus on a mix of private (business and consumer) incentives and public (regulatory) controls.
  3. The need to step back and consider generic approaches to the public versus private roles and responsibilities in food safety and risk management and determine which generic approaches are being used and under what circumstances.

Caswell elaborated on the last bullet point and described a range of generic approaches to private versus public involvement (see Figure 3-2 for schematic). On one end of the spectrum, there is no intervention, with public agencies having no policies around food risk or risk management and with industry relying solely on private controls. On the other end of the spectrum is direct regulation, whereby public agencies are prescribing what companies must do (e.g., with respect to labeling) and prohibiting certain actions, products, and processes. Between these two ends are a range of private/public mixes, including self-regulation, such as the use of industry voluntary codes of practice and farm assurance schemes; information and education, whereby government has the ability to generate and communicate information to consumers and make consumers aware of food safety issues; co-regulation, where regulation is the responsibility of a public-private partnership, with statutory or government-backed codes of practice (a popular mode of regulation in the European Union, particularly in the United Kingdom); and incentive-based structures, where the amount of regulation is in response to what companies are doing, for example inspection frequencies being conducted based on how well a company has performed in the past. Caswell argued that we do not spend enough time thinking about which of these options best describe our food safety regulations. For example, in which box does regulation of Listeria control in RTE meats fit?

FIGURE 3-2. A spectrum of generic approaches to private versus public involvement in global food safety risk management, as described by Caswell.


A spectrum of generic approaches to private versus public involvement in global food safety risk management, as described by Caswell.

In fact, any single box in Figure 3-2 encompasses a range of regulatory approaches. For example, direct regulation approaches range from labeling requirements (at the lower end of intervention) to the use of target, performance or product/process standards (a medium level of intervention) to requiring prior approval before a product can enter the market (at the high end of intervention). Even within each of these different levels of intervention, there is variation. For example, using a standards approach (i.e., the medium level of intervention) requires setting a standard (“standard setting”), deciding how the standard is going to be implemented (“process implementation”) and then enforcing and monitoring the standard (“monitoring and enforcement”). Moreover, regulation at each of these sub-levels could be either national or international and determined by either private or public interests. In short, Caswell said, the distinction between public and private regulations is less discrete than often assumed, with most markets having a mix of co-existing public and private safety regulations and considerable interrelationships and dependencies between the two.

Knowing which box and specific regulatory approach is being used and whether it is the right approach in any given food safety situation enables us to ask whether we might achieve better results with respect to both food safety effectiveness and economic efficiency if our regulatory approach were different (i.e., if a different mix of public and private roles and responsibilities might be more effective).

As an example of the variation that exists even within a single regulatory approach, or option, Caswell pointed to a study by Spencer Henson on the rapid development of private standards in food safety management and the range of standards being employed (Henson, 20085). Private standards used to be set predominantly on a business-to-business basis, with individual companies setting their own individual standards. These individual-level standards were either national (e.g., Nature’s Choice by Tesco in the United Kingdom, Field-to-Forks by Marks and Spencer in the United Kingdom, Filière Agriculture-Raisonnée by Auchan France) or international (e.g., Wal-Mart and Nestle). Henson has shown that, over time, private standard-setting has evolved into a meta-standard approach, where joint standards are used among a group of suppliers or retailers. Examples of these “collective” standards include, at the national level, the Dutch HACCP, the British Retail Consortium (BRC) Global Standard, Assured Food Standards, Qualität und Sicherheit (the “QS system”) and Integrate Keten Beheersing, and at the international level, the International Food Standards, Safe Quality Food (SQF) 1000/2000/30006 and GLOBALGAP (formerly EUREPGAP).

As another example of the variation in mixed private-public approaches being utilized, Caswell and colleagues conducted a study for the Food Standards Agency in the United Kingdom looking at private/public mixes across the United States, Canada, the United Kingdom, and Australia (Fearne et al., 20057). In the United States, they examined three areas of food safety regulation: Listeria in RTE meats, the introduction of biotechnology, and animal identification. They examined standards setting, process implementation, and enforcement and monitoring (i.e., the three components of the “standards” approach that Caswell had described previously). As illustrated in Figure 3-3, they found a mix of public and private involvement among these three components and across all three areas of regulation.

FIGURE 3-3. Examples of the wide range of public–private mixes with respect to food safety regulation among different areas of food safety regulation (the left-hand column in each table describes the area of food safety regulation considered), as described by Caswell.


Examples of the wide range of public–private mixes with respect to food safety regulation among different areas of food safety regulation (the left-hand column in each table describes the area of food safety regulation considered), as described (more...)

Also as shown in Figure 3-3, food safety programs in the United Kingdom and Canada employ a range of public-private mixes but are more consistent in their approach than we are in the United States. The United States is very reactive in its approach, Caswell explained. We make most decisions on a risk-by-risk basis, rather than assigning responsibility for the three different components (i.e., standard setting, process implementation, and enforcement and monitoring). This raises the question: Is there a way for the U.S. regulatory structure to step back and adopt a more comprehensive approach and philosophy toward the public–private mix across a range of risks, rather than dealing with each risk individually? Is there any consensus on the best overall approach to food safety production? How should we assign those responsibilities?

Caswell argued that having a comprehensive approach would mean having a set of criteria for dealing with risks (e.g., if the risk involves a, b, and c, then public–private mix x should be used) and dealing with risks on a meta-basis rather than on a risk-by-risk basis. She admitted that while it would be very difficult to develop this generic approach, it would be very conducive to the discussion of food safety management if our discussions moved in that direction. We need some sort of comprehensive roadmap for managing different classes or risks and knowing which approach(es) to take. Caswell concluded by referring to Taylor’s keynote presentation, reminding the audience of Taylor’s comments on the need to establish institutional roles and responsibilities. She emphasized that, while considering those roles and responsibilities, we need to think more generically about the private-public mixed mode of regulation and which approaches work best under which risk circumstances.


The first question was directed to Caswell. Audience participant Sandy Hoffmann of Resources For the Future asked if, in Caswell’s generic schematic of mixed private-public roles and responsibilities, a next-step would be to think about the characteristics of the risks or the market system and how those “map into alternative loci of control.” Caswell answered yes. While the public health risk associated with any situation is always going to be the guiding principle, other considerations would include market risk, governmental capacity to be effective in that area, the scope and comprehensiveness of private standards and trade impacts.

Ned Groth commented on how the workshop lacked perspective from an exporting developing country (i.e., exporter to the United States) and that it is a very important perspective to consider. At other food safety meetings, developing country representatives often echo concerns about the “silo effect.” That is, the development of standards and safety systems is often pursued either entirely within the private sector or entirely within the public sector and without adequate consultation with other stakeholders, particularly when those other involved parties are halfway across the world. Government personnel do see each other at Codex, WHO, FAO, or other meetings, Groth noted, where they are able to establish bilateral or multilateral collaborative efforts. Likewise, within the private sector, companies talk to suppliers in other countries, with a lot of one-on-one, back-and-forth interaction within individual problem areas. But how well are we doing at having international discussions where all stakeholders from all countries, both those importing and exporting, developed and developing, are at the table and addressing these issues in an integrated, synthesized way?

Brackett and Caswell both offered responses. First, Brackett agreed that it is an important issue. He noted that while there is communication and shared training between developed country importers and developing country suppliers, there is often a lack of communication between those importers/suppliers and government regulatory agencies. Similarly, academics and government representatives communicate frequently at professional meetings, Codex meetings and elsewhere. But he doesn’t know of any single construct that puts all three sectors—industry, government, and academia—together. That said, some international organizations and private companies currently have proposals in place to do just that, but nothing like it exists yet.

Caswell agreed that it was a very interesting and also very challenging question. The challenge, she said, is that private standards develop very rapidly and are having a significant impact on the market. In fact, one of the “glories of private standards” is that they develop and move rapidly, quickly responding to new and changing situations. In some ways, she said, we don’t want food safety regulation to be in a “Codex mode,” where a meeting occurs but the decision to act on a particular situation doesn’t happen until a few years later. Making the effort to integrate and coordinate public and private parties—and across countries—requires a means of capturing the “dynamic ability of private standards to evolve quickly.” A good intermediate step might be the development of meta-standards, or industry-wide standards. The individual private parties set the standards; those standards give those companies market power.

Mike Robach9 elaborated on Caswell’s comments on private standards by emphasizing that it is very important when talking about the proliferation of private standards to consider how those standards are sometimes “forced through the supply chain without appropriate scientific and public vetting.” He noted that Cargill has worked with a number of organizations, including FAO, OIE, and others to standardize its processes and now utilizes these standards in its operation of 850-plus food processing plants in 66 countries, regardless of location. Robach also reiterated that Codex is slow and needs to become “more nimble” and able to adapt to technological change. He suggested that the private sector could facilitate that. Robach echoed the call made by many other workshop participants throughout the day: we need to have all stakeholders at the table—not just inter-governmental agencies and the private sector but also national governments and civil society (i.e., the appropriate consumer input). Not only do all of these stakeholders have a role to play, but their participation also ensures that moving forward is a transparent process.

Another audience participant concurred that standard-setting, while perhaps facilitated by the private sector, should be a public process. She noted the extent to which private industry has been requesting public standards from the government over the years, particularly during times of crisis. A good example is the egg marketing and grading program, which is run through AMS (Agricultural Marketing Service) but was requested by that industry. Likewise with the seafood grading and inspection program run by the National Marine Fisheries Service. Often when industries are in crisis, they turn to government for standards.

Getting back to some of the specifics of Caswell’s talk, another audience member asked to what extent food safety management is really shifting from a “command and control” to a performance or process risk-based system. Or, is the shift from an approach where government issues “command and control” standards to one where the proliferation of private standards has led to a private “command and control” standards? Brackett commented that the movement, at least philosophically, is away from “command and control” to the use of more performance-based standards. He pointed to FDA’s HACCP regulations for juice and eggs as an example: there has been a “log reduction” of what is expected with no expectations of how that is to be accomplished. Utilizing performance-based standards allows industry to be more creative in its approach, he argued. That said, there are still residuals of the old “command and control” system, for example with milk pasteurization, and there are still very specific guidelines about how that must be done. As technology continues to change, however, it is going to push industry even more toward embracing a performance-based approach. Technology is changing too fast to dictate which particular method should be used. Caswell noted that it is in fact possible for the government to be operating in a “performance” mode with companies responding in a “command and control” mode and that these two modes of operation would not be inconsistent.

A couple of technical questions were directed to Flowers. First, a question was asked about his comments about break points and the convention of conducting a sanitation step every third shift in the meat industry: Is it technologically feasible to have extended runs in the meat processing environment, where the potential food safety benefits would be greater than have been realized? The questioner commented that the meat industry has been living with this “dogma” for years and years—that it is appropriate to conduct a complete clean-up on a nightly basis. But is this an area of food safety regulation that needs to be re-addressed? Flowers agreed that in other industries, like the dry food industry, shutting down, cleaning up, and re-starting sometimes creates additional problems and that continuous runs sometimes have fewer microbiological problems. He said it was “possible” that the same might be true of the meat industry and that it would be a matter of verifying (i.e., that your temperature and other post-process contamination control mechanisms are working). However, as Flowers had emphasized in his presentation, he emphasized again that the meat industry relies on data and reacts to data. So while the continuous nature of the dry food runs could be extended to refrigerated plants as well, he said, “I would really have to see that data” before answering that question.

Another audience member asked Flowers about the need for laboratory certification with respect to verification testing. The questioner commented that a number of bills moving in that direction (i.e., requiring lab certification) have been drafted for consideration in the 110th Congress and then asked, are there good international or other models that can be relied on as Congress moves forward on this issue? Flowers clarified that the issue is not certification but accreditation of laboratories and said that, yes, there are good models available, such as the ISO standard for accreditation of laboratories. The issue with those standards, however, is that they are applied to all types of laboratories and are very vague with respect food laboratories. We need much more specific standards. Flowers mentioned a group that he chaired a number of years ago—the Food Laboratory Accreditation Working Group (FLAWG), involving USDA, FDA, and industry members—and how the group wrote some very specific guidelines to accompany ISO Guide 1702-5 with respect to how food laboratories in particular should comply. Flowers also noted that some of the ISO standards, such as a requirement for proficiency testing, have been modified and “watered down” over the years.

Finally, Caswell was asked which policies and changes in regulation over these past two decades have had the greatest impact on reducing food-borne illnesses. Caswell said, “We don’t know.” The questioner commented that it would be a useful research question.



Ms. Janet Beauvais is the Director General of the Health Canada’s Food Directorate.


Robert E. Brackett, PhD, is Senior Vice President and Chief Science and Regulatory Affairs Officer at the Grocery Manufacturer’s Association, Washington, DC.


Russell S. Flowers, PhD, is Chairman and Chief Science Officer, Silliker Group Corp., Homewood, IL.


Julie A. Caswell, PhD, is Professor of Resource Economics at the University of Massachusetts, Amherst.


Henson, S. J. 2008. The role of public and private standards in regulatin international food markets. Journal of International Agricultural Trade and Development 4(1):63–81.


SQI is a food quality certification program managed by FMI.


Fearne, A., M. Garcia, J. A. Caswell, S. Henson, and Y. Kharti. 2005. Exploring Alternative Approaches to the Traditional Modes of Food Safety. Final Report, Imperial College, London. Prepared for the United Kingdom Food Standards Agency under Control D03004.


This section summarizes the discussion that followed the three presentations of this session.


Robach, of Cargill Inc., was a panelist in the third session but an audience member during this session.

Copyright © 2009, National Academy of Sciences.
Bookshelf ID: NBK26431


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...