CHAPTER

17

Performance Standards for Food Safety

by Sandra A. Hoffmann and Alan J. Krupnick

Mr. President, we urge you to adopt broad and consistent use of product performance standards as the centerpiece of foodborne pathogen regulation. Performance standards help assure product safety, maintain consumer confidence, and provide economic incentives for industry to find more efficient means of meeting food safety goals. This action would build on and strengthen already accepted principles of food safety regulation. In some instances, it will require legislative change; in other cases, sufficient legal authority already exists. Our recommendation draws on and is in large part in agreement with recent recommendations from the 2003 National Academy of Sciences Committee on Review of the Use of Scientific Criteria and Performance Standards for Safe Food.¹ Establishing performance standards is one element of the overall structural reform recommended in Chapter 16. Even if the full menu of reforms suggested there cannot be implemented, it should be possible to implement clear performance standards.

Background

Since at least the early 1980s, major advisory bodies, including the U.S. General Accounting Office (GAO) and the National Academy of Sciences (NAS), have made repeated calls for establishing a clear relationship between public health outcomes and food safety regulatory requirements. Much of the U.S. statutory food safety inspection mandate dates from the early 1900s and now serves very little public health purpose. For instance, the visual inspection of animals and carcasses required by the meat and poultry inspection acts are incapable of detecting or preventing pathogen contamination that leads to thousands of deaths and hundreds of thousands of hospitalizations each year in the United States.

These same advisory bodies have insisted on the need for greater flexibility to allow food processors to take advantage of rapid advances in biological sciences and food process technology. This insistence is a reaction to use of a command-and-control approach to food safety regulation, such as mandating a particular food safety technology or manufacturing process. In the 1990s, the U.S. Department of Agriculture (USDA) and the Food and Drug Administration (FDA) moved, within the limits of statutory and political constraints, to implement a more modern process standard approach to assuring product safety. They required meat, poultry, seafood, and juice processing plants to develop and implement a risk-management process: agency-approved Hazards Analysis and Critical Control Point (HACCP) plans. In developing these plans, plant management must access their production process, identify critical points where hazards could enter the system, and develop and implement a plan to control those critical points. However, most HACCP plans do not say what the plants should be trying to achieve through this process. The use of such plans was intended to encourage manufacturing processes that prevent product contamination with pathogens and limit pathogen growth. The major goal of these reforms was to facilitate adoption of modern quality-control processes. An equally important goal was to allow industry the flexibility to choose how to meet these goals.

But use of HACCP plans is falling short from the perspective of providing the flexibility needed to encourage efficiency and innovation. A plant’s plan must be preapproved by the regulatory agency before it can be implemented. The structure of requiring government preapproval of production processes at each plant limits a plant manager’s ability to take advantage of new technology and the learning that occurs in implementing a production process. Although plans are to be revised on a periodic basis, preapproval inherently imposes constraints on firms’ flexibility to find better ways to control food safety hazards. Furthermore, experience with environmental regulation suggests that plants will tend to follow agency guidance regarding choice of process to reduce their regulatory risk. This behavior significantly reduces the actual flexibility provided by a process standard approach like HACCP. In addition, both the FDA and Food Safety Inspection Service (FSIS) are statutorily required to regulate at the plant level. This plant-by-plant approach makes it impossible to cost-effectively manage risk along the entire supply chain. For example, it may be more cost-effective to minimize salmonella in poultry flocks than it is to decontaminate poultry carcasses.

Furthermore, current HAACP regulations still are not tied to desired public health outcomes. Plants are required to implement plans that reduce pathogens to an “acceptable level.” But the regulations do not define what an acceptable level is or how it is related to public health outcomes. A recent NAS committee noted that current regulations give “little or no guidance on the level of hazard control expected by the government from an adequately designed and implemented HACCP plan.” The GAO and the interested public have criticized HACCP regulation as providing inadequate accountability for food safety. Plants can be in compliance with agency-approved process control plans, but because the plans are not tied to achieving expected health outcomes, no way exists to assure that they will meet public health goals.

Tying HACCP regulation to product performance standards based on public health impact could address both of these shortcomings. In 2003, an NAS committee recommended requiring food safety product performance standards tied to explicit public health goals as a way of both assuring that food safety regulation achieves its public health purpose and providing incentives for innovation and efficiency.

The Product Performance Standard Approach

A system of performance standards should include several elements.

Accountability Tied to Public Health Goals

Food safety product performance standards need to be tied directly to accomplishment of a clear public health goal. The public health goal could be a desired reduction in incidence of foodborne disease or an acceptable risk of illness. EPA uses such a goal—not exceeding a one-in-a-million risk of cancer—in pesticide registration. The public health goal should be translated into measurable food safety objectives (FSO). The 2003 NAS report defines an FSO as “the maximum frequency and/or hazard in a food at the time of consumption that is considered tolerable for consumer protection.” By using data, scientific modeling, and where necessary, professional judgment, it is possible to determine the concentration of pathogen in product (the performance standard) at any given point in the supply chain that can be reasonably expected to result in a contamination level at the time of consumption that meets the FSO.

Incentives for Efficiency and Innovation

The decision of where in the food supply chain to impose product performance standards will affect the incentive they provide for efficiency and innovation. In theory, the greatest efficiency gains can be achieved by imposing a single performance standard as close to final consumption as possible. For example, an E. coli O157:H7 standard could be imposed on ground beef as it enters the retail chain. This would allow market forces to create incentives for wholesalers, shippers, processors, and producers to take control efforts at the most cost-efficient points in the whole production–processing–marketing chain. It also would create incentives for firms to innovate and create processes, procedures, and technologies that reduce costs.

Final suppliers, such as McDonald’s, already can demand specific safety levels from its meat suppliers. This and other experiences with “channel captains” in the food industry demonstrate that it is possible for product standards placed at the end of the supply chain to provide incentives for cost-effective reduction in pathogen risk throughout the supply chain.²

Given current statutory constraints, NAS suggests establishing a series of product performance standards for transportation and retail, processing, and on-farm. This is less advisable than a single standard as far forward as possible in the chain, perhaps one at the processing plant and another at delivery to retail outlets. If stage-specific performance standards were adopted without consideration of their impact on the entire food supply-and-demand chain, the criteria selected may be inefficient and performance standard–based regulation could create real disincentives for firms to adopt effective food safety controls.

Market forces will tend to promote cost-saving technological change. However, it is less clear that there is as strong a market incentive for safety-enhancing technological change. Performance standards can enhance such incentives if tied to public health goals and FSOs that require continuous improvement.

Science-Based Enforcement

A product performance standard approach to food safety regulation must include a science-based approach to enforcement. Compliance with performance standards can be scientifically verified through either product sampling or established statistical process controls. Modern industrial practices use advanced statistical methods for assuring that a production process meets a given standard for its products. In addition, the regulatory system should allow for adoption of advances in process control verification.

An Ability to Adjust to Rapid Change

Biological sciences, food consumption patterns, and food production practices in the United States all are changing very rapidly. In the past 20 years, we have experienced the impact of emerging pathogens, such as the deadly E. coli O157:H7, and have learned that acute infections from foodborne pathogens can have serious long-term health effects. In response, a food safety regulatory system organized around product performance standards must be continually updatable. Public health goals, FSOs, and the performance standards derived from them need to be reviewed on a regular basis without need for regulatory change. However, experience with chronic delays in the U.S. EPA’s program of periodic reviews of the scientific support for pesticide registrations warrants examination in designing such a system.

Congressional Support for Data Collection

In order to set public health objectives and to know whether the regulatory system is achieving the goal of protecting public health, good surveillance data that are reliably collected over time are needed. Data also are lacking to conduct some of the risk assessments needed to translate public health goals into FSOs and performance standards. Congress needs to provide adequate resources to develop and maintain data support necessary to design, monitor, and enforce performance standards.

Obstacles

Several obstacles exist to the adoption of performance standards. These include legal, market, technical, and political obstacles, as well as a lack of enforcement resources.

Legal Obstacles

Statutory authority to promulgate product performance standards may be lacking. The attempt by FSIS to use mandatory salmonella product performance standards as a criteria for judging the adequacy of meat processing plants’ HACCP plans was struck down by the 5th Circuit Court of Appeals in 2001 as being outside the agency’s statutory authority. Bills were introduced in the 108th Congress to give the USDA authority to use mandatory performance standards, but they were not referred out of committee (S. 2013, H.R. 3956).

Placing product standards at the end of the supply chain could require substantial legislative reform. Current federal law compartmentalizes responsibility for food safety both to particular parts of the supply chain and to particular foods. For example, FSIS has jurisdiction over meat and poultry slaughter and processing plants. A frozen pizza could be regulated by several federal agencies. State government has traditionally had jurisdiction over inspections and testing of products in warehouses and at retail. Local governments also have jurisdiction over retail establishments. Creation of a food safety regulatory system built around product performance standards placed on final products could necessitate revisiting these relationships, although it may be possible to introduce performance standards through FDA’s Model Food Code, which is revised regularly and has been adopted by all but one state. Use of product performance standards at the point where the product leaves the processing stage and enters the marketing stage of the food supply chain would require much less extensive legislative reform. This portion of the food supply chain currently is under federal jurisdiction.

Market Obstacles

The size of firms in the food industry varies greatly. The design of product performance standards needs to consider the influence of industry concentration on the transmission of signals about product quality requirements upstream from the retailer to the shipper, the processor, and the farm. What works in an industry dominated by a few large retailers, such as McDonald’s, may not work in one without large retailers.

As in other regulatory areas, attention needs to be paid to impacts on small businesses, even though they make up a small fraction of total sector output. Small businesses could be given longer deadlines for compliance, exemptions, or more lenient requirements. Information-sharing initiatives also could help raise management and technical sophistication at small firms.

Another central concern is the impact on the food budgets of the poor. The prices of lower-quality products may rise more than the cost of others, as these products now tend to be produced to lower levels of safety. In this case, the poor may no longer have the option of trading off safety for cost. However, the efficiency gains and incentives for innovation provided by performance standards may well reduce prices, putting the poor in a position of getting more safety for a lower price.

Technical Obstacles

Performance standards can work only if they are effectively enforced. Monitoring and measuring contamination are challenging tasks. Currently, we lack data on baseline contamination and efficacy of different interventions. But a performance system creates incentives to improve data. The retailer, as well as the government, has particular incentives to produce data. Indeed, if the enforcement system is designed appropriately, each stage in the farm-to-table chain has incentives to make sure either that the food they pass on is clean or that it is very clear who has the responsibility to clean it.

Despite rapid advances in testing technology, meaningful delays still exist between the time a sample of a product is taken in a plant and the time results are available. By then, the product often has been packaged and even marketed. As a result, as currently conducted, pathogen tests provide a “scorecard” for plant performance rather than a direct indicator of product safety. Faster testing is required.

Potential Political Obstacles

Current inspection laws mandate highly labor-intensive inspection regimes. A major obstacle to past efforts to make major reforms of meat inspection statutes has been resistance from meat and poultry inspectors’ unions. Consumer opponents of HAACP view it as “deregulation.” Some may have the same reaction to reliance on performance standards to provide incentives to producers. But other consumer groups see product performance standards as a means of assuring accountability for product quality.

Regulated firms may prefer the security of knowing that they are complying with a process standard to the responsibility of having to assure safe product. As experience with environmental regulation suggests, firms will have to be convinced that the agency has conferred on them real compliance flexibility before they will chance a departure from traditional approaches to food protection.

Lack of Enforcement Resources

Historically, the distribution of resources for enforcement across federal agencies has been uneven. Of particular concern, FDA has had a very low level of funding for inspection activities. The greatest gains in security, efficiency, and innovation from use of food safety performance standards would result from their comprehensive application across the food supply. This kind of new approach to food safety regulation would require reexamination of the distribution of enforcement funding across the federal food safety system.

Conclusions

A shift from reliance on process standards to reliance on product pathogen performance standards could enhance the effectiveness and efficiency of federal food safety regulation. The greatest gains in enhancing the effectiveness and efficiency of the U.S. food safety system would come from creation of performance standards for final consumer products. Such a shift will require legislative action and a major change in the structure of food safety regulation. Smaller, but still significant, gains could be had by using performance standards to regulate pathogens in products as they leave processing plants. This also would require legislative authorization, but it would necessitate a much more modest change in the food safety regulatory structure.

It is critical that product performance standards be based on measurable public health goals, as recommended by the NAS. Without a direct tie to public health outcomes that can be revised and updated in response to changing scientific knowledge, we are in danger of re-creating the same rigidities handed to us by the 1906 Congress.

The following principles³ outlined by USDA economist Elise Golan encapsulate our recommendations for the design of food safety product performance standards: Regulate as close to the end user as possible to encourage upstream innovation. Choose strict, not merely feasible, standards to encourage efficiency and innovation. And use criteria for compliance verification that are informative, reliably measurable, and flexible.

S.A.H.

A.J.K.

Notes