Thursday, August 25, 2016

Label limbo: Use by when? According to whom?



By Michaela Oldfield
MSU Global Food Law Fellow

You know the scene. There’s that container of yogurt idling in your fridge. It had a "best by" date of a week ago, maybe more. Is it just idling, or is it festering? Should you eat it, or toss it?

This is a conundrum faced by consumers every day, and what do to about their confusion has been getting some recent news coverage. There is a movement happening to address the use of dates in food labels, because they are creating confusion that leads to consumers throwing away perfectly safe food. Even Wal-Mart is getting in on the action.

This issue of labeling is part of a larger, much more complex issue of how to address food waste. Here’s some of the other issues implicated in food waste reduction efforts:

  • Why is food being wasted? Because of cosmetic standards, because of food safety standards, because there’s too many calories in our food system for how many people we have? The food waste solution depends on what you see as the cause of the problem.
  •  Is this about reducing waste to improve sustainability, make more food available, or save consumers money? Is reducing food waste the right approach to tackle that particular problem?
  •  Food “waste” can actually be redirected into food “recovery” when it is donated to anti-hunger organizations. But who bears the cost of collecting, distributing and disposing of unusable foods? And is it just to expect low-income individuals to rely on recovered food for their nutritional and caloric needs?
  • The best way to reduce waste is to not produce or purchase more than can be consumed, which reduces the availability of food for recovery - how might food waste reduction initiatives operate at counter purposes with each other?
  •  If we reduce food waste, does that reduce food demands, depressing prices or causing problems for the food industry that needs to sell more food?

There are more issues. Those are just a few that have occurred to me as I’ve been considering the issues and implications of food waste in the U.S. While focusing on date labeling makes sense because its low hanging fruit, these complications make me a little concerned about the strategy. It runs the risk of making it seem that a simple switch by industry and behavioral change by consumers will be a sufficient quick-fix.

Nonetheless, given the recent coverage of food date labels, here’s a brief intro/refresher on some food date labeling law.

Consumption dates on foods in the United States are governed by the Food, Drug and Cosmetic Act, 21 U.S.C. §341 et seq. However, contrary to what many consumers might think, most foods are not required to have any sort of “consume by” or expiration date. For the most part, dates must be used in a manner that is not ‘misbranded”, 21 U.S.C. § 331(a), meaning they cannot be “false or misleading in any particular”, 21 U.S.C. § 343(a)(1). The one exception is that infant formula must be labeled with “use by” dates. 21 C.F.R. § 107.20.

Other than that, it’s up to companies to decide what dates to use and even whether to put any sort of dates.

This issue came up a few years ago in United States v. Farinella, 558 F.3d 695 (7th Cir. 2009), in which the FDA prosecuted a man for misbranding salad dressing when he changed the “best when purchased by” date on the bottles.

The prosecutor in the case took “best when purchased by” to be the same thing as expiration dates, id. at 697, and prosecuted the man on the grounds that it was a safety issue. But that’s not really what those date labeling terms mean.

There’s no consensus on exactly what “best when purchased by” or “best when used by” dates mean 558 F.3d at 698. They’re generally about freshness and taste, but it’s up to the manufacturer to decide what these dates should be based on their knowledge of the product’s shelf life. Id. And in theory, anyone in the distribution chain could make a judgement about the products quality decline and relabel accordingly (though that could constitute a breach of contract between the manufacturer and distributor). Id. at 697.

Expiration dates can have a specific meaning. The FDA’s regulation of infant formula requires “use by” dates which are calculated “on the basis of tests or other information showing that the infant formula, until that date, under the conditions of handling, storage, preparation, and use prescribed by label directions, will: (1) when consumed, contain not less than the quantity of each nutrient, as set forth on its label; and (2) otherwise be of an acceptable quality (e.g., pass through an ordinary bottle nipple).” 21 C.F.R. 107.20(c).

Some states also regulate the “use by” and other labeling dates, with considerable variation as to whether it’s for safety or other fair-marketing reasons. (See the NRDC report the Dating Game for an inventory of states’ labeling laws.)

In the case of United States v. Farinella, since it was being prosecuted under the FDCA, Farinella had done nothing violating that law and so his conviction was overturned.

This confusion over the date labeling arose more recently in Chase v. Bros. Int'l Food Corp., 3 F. Supp. 3d 49 (W.D.N.Y. 2014). In that case, Chase complained to his employer about the date labels being changed, was fired for not signing a non-disclosure agreement that would have restricted him making the information public, and subsequently brought a false termination claim against his employer under the whistleblower protection provisions of the Food Safety Modernization Act, 21 U.S.C.A. § 399d(a).

Whether the employee’s activity was protected under the whistleblower provision turns in part on whether the employee had a reasonable belief that the law was being violated. “It is well settled that an employee's mistaken belief that an employer has violated a law may nevertheless be reasonable provided that the totality of the circumstances establishes that a reasonable employee considering the same facts could have concluded that the employer had violated the law.” 3 F.Supp.3d 49 at 54. Consequently, the labeling date confusion exposes food manufacturers to some potentially difficult employment situations.

So long story short: there’s no consistent federal requirements on what labeling dates should be. Representative Pingree has introduced legislation to try to address this. But in the meantime, so long as the terms and dates are used in a not-misleading manner, food companies can apply dates as they deem appropriate.

If FDA agents and food company employees can’t understand these laws, it’s no wonder that consumers are confused too!

Thursday, August 11, 2016

The dose makes the poison, but who defines poison?



By Michaela Oldfield
MSU Global Food Law Fellow

One of the many highlights of IFT16 was an all-star panel (organized by yours truly) on toxicology research, risk communication, and the Generally Recognized as Safe standard.

Our first speaker was Dr. Michael Holsapple, the founding director of the MSU Center for Research in Ingredient Safety (CRIS). Dr. Holsapple is a toxicologist who is establishing CRIS to fill in the gaps in addressing food safety. There are three major components of food safety: allergens, microbial risk and toxicology. While there are university research centers that address allergens and microbial risk, CRIS is the first center focused on determining the toxicological effects of ingredients. Because its work is not limited to food, the center is planning to conduct research on all consumer product ingredients.

Dr. Michael Holsapple, founder of the Center for Research
in Ingredient Safety at Michigan State University.
The quote from Dr. Holsapple that sticks with me is “the dose makes the poison,” originally coined by the Swiss German physician and alchemist Paracelsus – also known as the founder of toxicology. It struck me because while perhaps true, the issue I see the food industry grappling with is who defines when a dose becomes a poison? Industry, regulators or consumers? Legally it’s the FDA, but how it works in practice is complicated.

Our second speaker was Charlie Arnot, of www.foodintegrity.org, speaking about how to communicate risk science to consumers and regulators. His talk was important because even if industry and regulators have determined an ingredient is safe, consumers may not buy products if they don’t trust or believe “the science.

So how can industry communicate science to consumers so that they understand why a particular dose is not considered poisonous.

First is to understand why consumers no longer trust the food industry or regulators. Mr. Arnot noted there have been significant social shifts in the last 45 years that have eroded trust in institutions. Most notably for me were the emergence of television (and now social media) that are changing how consumers obtain and process information, and major violations of trust by institutions and leading public figures such as Clinton, Nixon, Lehman Brothers and quite a few others.

At the same time, the food industry has gone through massive consolidation, integration and industrialization, so that it is now also viewed as an institution that is suspect.

Second is to understand how trust is built and communicated in today’s society. The top variable influencing whether a consumer trusts a message is whether they perceive shared values with the messenger. But the believability also depends on other elements of the message, such as the openness, transparency, and honesty of the message being communicated, as well as “outrage factors” such as familiarity, consumer’s control of the risk, and their sense of fairness at being exposed to it.

There are quite a few other factors I’m not going to identify because I want to discuss GRAS and how the legal complexity of food ingredients and food additives regulations confounds efforts by industry and scientists to build trust with consumers.

“In its broadest sense, a food additive is any substance added to food. Legally, the term refers to "any substance the intended use of which results or may reasonably be expected to result – directly or indirectly – in its becoming a component or otherwise affecting the characteristics of any food." This definition includes any substance used in the production, processing, treatment, packaging, transportation or storage of food. The purpose of the legal definition, however, is to impose a premarket approval requirement.”

If a substance is a food additive, it must go through pre-market approval before it can be used in food. 21 USC 348(a)(2). However, the definition of food additives, at 21 USC 321(s), is complex because it’s also designed to remove commonly used substances, such as salt, from having to go through pre-market approval. So in full, the definition reads:

(s)The term “food additive” means any substance the intended use of which results or may reasonably be expected to result, directly or indirectly, in its becoming a component or otherwise affecting the characteristics of any food (including any substance intended for use in producing, manufacturing, packing, processing, preparing, treating, packaging, transporting, or holding food; and including any source of radiation intended for any such use), if such substance is not generally recognized, among experts qualified by scientific training and experience to evaluate its safety, as having been adequately shown through scientific procedures (or, in the case of a substance used in food prior to January 1, 1958, through either scientific procedures or experience based on common use in food) to be safe under the conditions of its intended use; except that such term does not include:

(1) a pesticide chemical residue in or on a raw agricultural commodity or processed food; or
(2) a pesticide chemical; or
(3) a color additive; or
(4) any substance used in accordance with a sanction or approval granted prior to September 6, 1958, pursuant to this chapter, the Poultry Products Inspection Act [21 U.S.C. 451 et seq.] or the Meat Inspection Act of March 4, 1907, as amended and extended [21 U.S.C. 601 et seq.];
(5) a new animal drug; or
(6) an ingredient described in paragraph (ff) in, or intended for use in, a dietary supplement.
So the substances in (1)-(6) are all excluded from the definition.

What is important for this discussion is the Generally Recognized as Safe, or GRAS, provision of the sentence. A substance is a food additive if it is not GRAS. That has to go through a clear pre-market approval process, 21 C.F.R. Part 170, which in itself is complicated enough.

But what is GRAS, and what does that really mean for food ingredients? For a substance to be GRAS, and therefore not a food additive, it must be “generally recognized, among experts qualified by scientific training and experience to evaluate its safety, as having been adequately shown through scientific procedures …to be safe under the conditions of its intended use.”

The FDA’s regulations on GRAS determination standards are in 21 C.F.R. §§ 170.30 and 170.35. I’m not going to go into them, because they’re too complicated to cover in a blog. Instead, I will cover the comments that Prof. Neal Fortin offered on GRAS, because these are important to my discussion of who defines the poison threshold in practice.

  1. Loss of GRAS status is self-implementing
The provision in the Food Drug and Cosmetic Act is self-executing, so it doesn’t require the FDA to approve a substance as GRAS.  If the substance is generally recognized by experts as safe for its intended use, the substance is exempt from the additive approval process.

There are two ways a substance can be considered ‘generally recognized as safe’ – through a long history of use in food prior to 1958, or based on scientific data. Either way, the assessment must be based on the expertise of scientists, but scientists can base their assessments on general use prior to 1958; after 1958 they must base their assessment on scientific evidence. The regulations establish what evidence is needed to establish this. While companies do not have to go through a pre-approval, in choosing to use a substance, they must do their own self-assessment of whether a substance is GRAS according to those regulations. They may choose to notify the FDA, or they may keep their assessment on record in the company.

So while this allows for an easier process for companies to use a substance, it also means that loss of GRAS status is also self-implementing. As in, when new scientific evidence emerges that contradicts or calls into question whether experts generally recognize a substance is safe for its intended use, then it no longer has GRAS status.

  1. There is no such thing as a GRAS substance
There is no substance that is universally GRAS. GRAS is only for the product’s intended use. The case exemplifying this is United States v. An Article of Food, Coco Rico, Inc., 752 F.2d 11 (1985). In this case, the company took potassium nitrate, which is approved for use in curing meat, and put it in a soda concentrate. The company could offer no affirmative evidence that  potassium nitrate is generally recognized among experts as safe for use in soda; only assertions that they didn’t know of any evidence that it was unsafe. Consequently, it was not exempt from the food additive definition, so it was considered a food additive, which was unsafe because it had not gone through the pre-approval process, and therefore the concentrate was adulterated.

So basically, the GRAS status of a substance needs to be determined for every intended use.

  1. Long history of use never creates GRAS status
On a related note, a long history of use never creates GRAS status because it’s the intended use that matters. So, for instance, even though caffeine has had long use in coffee and sodas, and had a GRAS status for those uses, when it’s put to a different use – say, caffeinating alcoholic beverages – it is not GRAS until there is general recognition among experts that it is safe for its intended use. Since this is a new use, that is different from what was done prior to 1958, that GRAS determination must be based on scientific studies, not just the fact that caffeine was commonly used in other products prior to 1958.

  1. GRAS proof of safety standard is more stringent than the food additive standard
Finally, the standard of “generally recognized” is actually more stringent than the food additive standard which requires consensus. This means that studies questioning the safety of a produce – even if they don’t prove it is unsafe – can destroy a substance’s GRAS status for a particular use.
So practically what does this all mean for toxicology research, science communication, the food industry, and the dose makes the poison?

There are legally complex standards for a substance to be GRAS that rely on complex scientific evidence. The knowledge of the law and science often resides with experts who are parts of institutions – government and the food industry – that consumers no longer feel they can trust, and understanding these complexities requires expertise that is largely beyond the common consumer.
It is no wonder the food industry has a public relations problem. Even though there are defined standards for determining when a dose becomes poisonous, it is difficult for consumers to understand them or how they operate.

Hence, it looks like an intrigue among a number of distrusted institutions to make a buck at the consumer’s expense.

The dose may make the poison, but the consumer with their purchasing dollar can decide whether he or she is willing to accept a certain dose as poisonous or not. And the law and science do not make it easy to communicate to consumers why one dose should be acceptable and another should not.

Monday, July 25, 2016

The GMO labeling bill – what does it really mean?

By Michaela Oldfield

As we were wrapping the final day of the MSU Food Law Current Issues seminar, Congress finally passed a GMO labeling bill,[1] S. 764, 114th Cong. (2016).  I would be remiss if I didn’t delve into it a little bit.

The bill and labeling issues in general were a hot topic at the seminar. Rather than delve into the political controversies, I am going to draw on the seminar discussion to explore the legal issues that I see in specific provisions in the bill.

For alternative interpretations of the bill, you can read Civil Eats’ story on what it means for consumers (and especially those of the “no-GMOs” variety) and Politico’s coverage of the USDA-FDA battle over interpretation of the bill and analysis of the winners and losers.

Of course, the big news and the reason the bill has just been enacted is that it preempts state labeling requirements. It states:

‘‘FEDERAL PREEMPTION - No State or a political subdivision of a State may directly or indirectly establish under any authority or continue in effect as to any food or seed in interstate commerce any requirement relating to the labeling of whether a food (including food served in a restaurant or similar establishment) or seed is genetically engineered (which shall include such other similar terms as determined by the Secretary of Agriculture) or was developed or produced using genetic engineering, including any requirement for claims that a food or seed is or contains an ingredient that was developed or produced using genetic engineering.” S. 764, 114th Cong. § 295(b) (2016).
You may remember from my discussion of the GMA’s suit over Vermont’s GMO labeling bill litigation that there are three basic types of preemption: express, field and conflict. This is an express preemption.

Before anyone gets overly excited though, the bill has an exception to this preemption. S. 764, 114th Cong. § 296 (2016) states ‘‘Nothing in this subtitle, subtitle E, or any regulation, rule, or requirement promulgated in accordance with this subtitle or subtitle E shall be construed to preempt any remedy created by a State or Federal statutory or common law right.” So the law doesn’t preempt causes of actions a citizen might bring under state consumer protection laws.  That means claims alleging consumer misrepresentation for things like “100% natural” and “all natural” and other terms where consumers object to the use of GMOs in a product are not preempted.

 The bill, of course, defines what bioengineering is with respect to food. It is food.

“(A) that contains genetic material that has been modified through in vitro recombinant deoxyribonucleic acid (DNA) techniques; and

(B) for which the modification could not otherwise be obtained through conventional breeding or found in nature.” S. 764, 114th Cong. § 291 (2016).

This has two regulatory issues. First, this requires trying to prove a negative – that something could not otherwise be obtained through breeding or found in nature. This is a difficult, if not nearly impossible, task. USDA can try to set a standard for establishing the negative, but I would bet this will end up in litigation.

The definition also means there are a lot of foods that will likely not be required to have a GE label because they do not contain any GE DNA, even though a GE was used in the production of the product. For instance, cheeses often contain rennin that is produced by GE fungi, citric acid is produced from a GE mold, and any number of products from GE crops may be so highly refined that there is no genetic material left (more on this last point later).

Another use of GE crops in the food system is as feed for animals. The law explicitly addresses this by prohibiting animal-derived foods from being a GE food just because an animal consumed GE feed. S. 764, 114th Cong. §. 293(b)(2)(A) (2016).

Interestingly, authority for establishing the mandatory labeling requirement is assigned to the USDA. S. 764, 114th Cong. § 291(3) (2016) states that "Secretary" means the secretary of agriculture, and S. 764, 114th Cong. § 293(1) (2016) states that the Secretary shall ‘‘establish a national mandatory bioengineered food disclosure standard with respect to any bioengineered food and any food that may be bioengineered.”

However, under Sec 292, the law will only apply to foods that are regulated under the Food, Drug and Cosmetic Act (which, you should recall, are regulated by the FDA under 21 U.S.C. § 321(d)). If it’s a product subject to the Federal Meat Inspection Act (21 U.S.C. 601 et seq.), the Poultry Products Inspection Act (21 U.S.C. 451 et seq.), or the Egg Products Inspection Act (21 U.S.C. 1031 et seq.), the GE labeling bill will apply if “the most predominant ingredient of the food would independently be subject to the labeling requirements” of the FDCA. S. 764, 114th Cong. § 292(c)(2) (2016). If the product is broth, stock, water, or a similar solution, then the GE labeling will only apply if the second most prominent ingredient would be required to be labeled under the FDCA. S. 764, 114th Cong. § 292(c)(2) (2016).

So read here: this only applies to FDCA regulated foods, or meat, poultry, and egg products if the first ingredient on its own would be subject to FDCA labeling requirements or liquids if the second ingredient on its own would be subject to FDCA labeling requirements. Yet now USDA is going to be responsible for developing and enforcing labeling regulations. If you’ve taken U.S. food law, you will realize this adds regulatory confusion to an already chaotically organized food regulatory regime in the U.S.

For highly refined products, there’s an additional issue that the law allocates to USDA for resolving. ‘‘A regulation promulgated by the Secretary in carrying out this subtitle shall …‘(B) determine the amounts of a bioengineered substance that may be present in food, as appropriate, in order for the food to be a bioengineered food.” S. 764, 114th Cong. § 293(b)(2) (2016). So basically, USDA will have to answer the question “What is the threshold level that will trigger the requirement that the food must be labeled?”

This threshold issue also applies for labeling of grains that have been inadvertently commingled with a GE grain. So, for instance, corn, soy, wheat, and oats are often grown in rotations and handled through the same supply chain. This is so common that USDA commodity grading standards have established allowable thresholds for the presence of other grains in the commodities, 7 C.F.R. §§ 810.101 et seq

Issues with comingling and cross contact have been coming up more with allergens and gluten-free labels. For instance, oats are a non-gluten containing grain that should be able to be labeled gluten-free under FDA’s labeling rules. However, they sometimes contain gluten due to cross-contact in the supply chain. FDA addressed this by establishing a threshold for gluten-free labeling claims that allowed up to 20 ppm gluten in products and ingredients under certain conditions. 21 C.F.R. § 101.91.

How much cross-contact will a company have to prevent if it's not going to label its product as having a GE ingredient? The USDA’s answer to the threshold question will help companies navigate this.

Just because a food isn’t required to have an affirmative GE declaration does not mean that it can be labeled as non-GE. For GE ingredients, the bill states “food may not be considered to be  ‘not bioengineered’, ‘non-GMO’, or any other similar claim  describing the absence of bioengineering in the food solely because the food is not required to bear a disclosure that  the food is bioengineered under this subtitle.” S. 764, 114th Cong. § 294(c) (2016). Going back to an earlier point about preemption, even if a company complies with this prohibition, it doesn’t mean consumers may not bring some other misleading labeling claim under state law. If a food contains GE ingredients below the mandatory labeling threshold, but uses a marketing term that a consumer interprets to mean it is GE-free, this would potentially expose the company to litigation liabilities.

The bill does have a little savings clause for organic foods. Being certified as Organic under the Organic Foods Production Act of 1990 (7 U.S.C. 6501 et seq.), will be considered sufficient to make a claim such as “non-GMO” or “GMO-free” claim. S. 764, 114th Cong. § 2 (2016).
If you’ve followed the news coverage, you should also know that the bill is set up to allow manufacturers to label their product with text or a digital link (i.e. QR code), with some additional potential labeling options for particularly small packages and small companies. S. 764, 114th Cong. § 293(d) (2016).

In terms of enforcement, that too will fall to the USDA. The bill actually amends the Agricultural Marketing Act of 1946 (AMA) (7 U.S.C. 1621 et seq.). S. 764, 114th Cong. § 1 (2016). If you’re not familiar with this law, you should go explore it a little. The law is about giving the USDA powers to develop and manage agricultural markets. Here’s an excerpt from the congressional declaration of purpose, 7 U.S.C. § 1621:

... it is the intent of Congress to provide for (1) continuous research to improve the marketing, handling, storage, processing, transportation, and distribution of agricultural products; (2) cooperation among Federal and State agencies, producers, industry organizations, and others in the development and effectuation of research and marketing programs to improve the distribution processes; (3) an integrated administration of all laws enacted by Congress to aid the distribution of agricultural products through research, market aids and services, and regulatory activities, to the end that marketing methods and facilities may be improved, that distribution costs may be reduced and the price spread between the producer and consumer may be narrowed, that dietary and nutritional standards may be improved, that new and wider markets for American agricultural products may be developed, both in the United States and in other countries, with a view to making it possible for the full production of American farms to be disposed of usefully, economically, profitably, and in an orderly manner.
The GE bill makes failing to make a required disclosure a prohibited act and requires companies to maintain and make available to the USDA any records that the USDA decides to require. S. 764, 114th Cong. §§ 293(g)(1)-(3) (2016). USDA’s enforcement authority of the labeling is limited to inspection of the records and publicizing a summary of the records audit (after a notice and hearing with the person who was the subject of the record). S. 764, 114th Cong. § 293(g)(3) (2016). The bill explicitly prohibits the USDA from requiring a recall of a mislabeled product by stating “The Secretary shall have no authority to recall any food subject to this subtitle on the basis of whether the food bears a disclosure that the food is bioengineered.” S. 764, 114th Cong. § 293(g)(4) (2016).

Normally this would seem like weak enforcement authority that undermines the potential effectiveness of a bill. However, the non-preemption of state causes of action for consumer deception means citizen suits through civil litigation will augment the USDA’s enforcement authority. In the same way that bloggers are testing gluten-free products and publicizing if they contain gluten (even when within the 20 ppm allowed by the FDA), citizens can test, publicize, and litigate any product that fails to accurately disclose the presence of GE ingredients in the food.

So that’s the fun new permutations of labeling law that we’ll have in the U.S.

Here’s some key takeaways for our food law student readers:


So remember, understanding this law (like all laws) will require analysis of the codified statute, the regulations, and any case law, as well as synthesizing interpretations of overlapping and related statutes, regulations, and case law.



[1] For any of you working on your legal research skills, here’s a fun issue you should note. The bioengineered labeling bill is titled “A bill to reauthorize and amend the National Sea Grant College Program Act, and for other purposes”. Say what? Congress has a little switcheroo maneuver that they pull sometimes, where they completely substitute the text of one bill for another. This can make it difficult to find the original text of a bill if you’re only searching by what you think the name is. Sometimes you’re better off searching by the bill’s number or sponsors’ names. 

Blog Archive

Twitter Updates

    follow me on Twitter