Current Status of Food Labels *

Edward L. Korwek

Hogan & Hartson
555 13th Street NW
Washington, DC 20004

[* edited from a transcript of the author’s oral presentation]

[published February 1994]

I would like to discuss the central question that has been asked for several years now: should genetically engineered foods be labeled? And I will have an answer for you (believe it or not) after a mere half-hour presentation.

My goal here is to try to give you a sense of the difficulties associated with answering this type of generic question. I will do this by means of references to statutory provisions (legal provisions) of the Federal Food, Drug and Cosmetic Act (FFDCA), which is the primary law applicable to this issue. The FFDCA is administered by the U.S. Food and Drug Administration (FDA). I will not focus a great deal on the legal niceties of the law because I know most of you are not lawyers. I think it is important instead to emphasize that the FDA labeling initiative regarding products of biotechnology basically is an effort to implement various sections of this law. The degree to which the FDA does or does not vary from the provisions of the FFDCA is a legal question that is in some people’s minds very difficult to answer. You will see, I hope, after this brief discussion that the question is certainly a complex one for a variety of reasons.

Let me go back a bit in time. In May, 1992, the FDA promulgated a food biotechnology policy statement in the Federal Register that primarily focused on issues of development of safety information for genetically engineered products. I have used the term “genetically engineered", although the stated focus of the policy was on new plant varieties developed through any genetic technology. The reality is that the focus was really on recombinant DNA-type products and the issues associated therewith. I use the term “genetically engineered” generically, although that in itself presents some definition problems that will be addressed later. The focus of the FDA policy proposal issued in May, 1992 was largely on when and how the FDA will require pre-market clearances for various types of food or food ingredients produced using “genetic engineering”. The labeling side of the issue was not really covered in detail except that the agency said that it did not feel that genetically engineered foods necessarily had to be identified in any particular way.

An exception to this generalization was the case in which there were allergenic properties associated with various ingredients of a genetically modified food. These ingredients would have to be identified in some way on the product label. In terms of food label identification, the allergy issue is relatively easy to deal with. The more difficult issues are associated with the question of whether the FDA can require labeling for other than safety considerations associated with genetically modified food or, for that matter, any type of food. In other words, what are the various characteristics of the products that the FDA might require labeling of for reasons other than safety considerations?

The FDA has abundant authority to require labeling of foods for anything that poses a safety concern. The most notable example of the exercise of this authority is the requirement of labels for various synthetic colors such as FD&C (Food, Drug & Cosmetic) yellow-5. Even in this case, however, the label requirement is merely a declaration of the presence of the specific coloring ingredient. It is not a warning that this food contains FD&C yellow-5. It is simply a specific identification on the product label that it contains FD&C yellow-5 as part of the ingredient statement. The significance of this requirement is that colors do not normally have to be specifically declared on food labels. The FD&C yellow-5 example is a notable exception to the general rule. Note clearly the precedent that this label requirement represents. The precedent is one of specific identification on the label of a particular ingredient that may pose health concerns, not necessarily a warning of hazard.

Following up on the May, 1992 notice from the FDA, which really did not address the general question of whether or not foods should be labeled, there was a notice almost a year later in April, 1993 in which the FDA has sought further input on the issue of labeling of genetically engineered foods. I encourage any of you who want to enter into this debate to read this notice and you will see how complex the issues are. The FDA does a very good job of outlining the various issues — how one can approach them, the questions that need to be asked, the questions that need to be answered, etc.

I will give you a preview of my answer to the question of whether or not genetically engineered food should be labeled. The answer is simply: I do not know. It depends on the specific characteristics of the “genetically engineered” food in question. Anyone who tries to answer that question generically does not have a good understanding of the complexity of this area. I have reviewed a number of position papers by various organizations on the topic and, in my opinion, they all have failed miserably. The reason why they have failed, as I will try to demonstrate to you, is because it is difficult to generalize in the labeling area. You have to ask very specific questions about what it is you want to have labeled. But until you identify what the characteristics are that you want to have identified on the label, you can not answer the question “Should genetically engineered foods be labeled?”. It is an impossible question to answer. Anyone who generalizes the issue simply can not come up with a reasonable answer. I will try to show you why.

Consider as an example one of the issues raised in the popular press about genetically engineered foods. This issue is source identification: the need to label the source of genes in a food. A common mistake in the case of genetically engineered foods is not to identify what the questionable characteristics of the food might be that would trigger labeling of the types of genes present. Rather, an argument is made of the necessity to identify merely the source of genes present in a food, regardless of the reason why they are present. We see articles and cartoons that provide fictitious menus in which all biotechnology foods are labeled to identify the source of genes. Examples might include spiced potatoes with wax moth genes, juice from tomatoes with flounder genes, blackened catfish with trout genes, pork chops with human genes, scalloped potatoes with chicken genes, corn bread with firefly genes, dessert rice pudding with pea genes and milk with genetically engineered bovine growth hormone. I would argue that this sort of information often is useless to the consumer because it does not convey any material information about the food’s characteristics. I consider it a misstatement of the real issues associated with labeling of foods derived from the use of modern biotechnology to require the general labeling of sources of genes. Unless the source of genes imparts material information to the consumer about some important characteristic of the food, I do not think there is any need for such source information to appear on a food label.

With that introduction, I would now like to summarize for you some of the statutory provisions of the FFDCA. These provisions are very important because, as I stated above, the reason for FDA initiatives in the biotechnology food labeling area is tied to various sections of the FFDCA. Labeling, in terms of FDA law, is related to the concept of “misbranding”. Misbranding means any sort of mislabeling of food that is somehow misleading or false. Misbranding, although serious when it occurs in relation to foods, is not as serious as adulteration of foods, in terms of enforcement priorities. “Adulteration” is a term of art under the law that basically addresses safety concerns associated with the food. If a food contained a contaminant that might pose safety concerns, that would be adulteration. In general, adulteration is a more serious violation of the FFDCA as an enforcement priority. However, there is an exception to the general adulteration concept involving safety concerns that I have just described to you. Certain quality characteristics of food can also relate to what we call “economic adulteration”. Economic adulteration is generally considered to be more serious than most misbranding violations. In economic adulteration one is trying to substitute something of poor quality for a valuable characteristic or make a food inferior in some way to a comparable product. For example, “butter cookies” is a generic term that suggests that cookies contain butter as a basic characterizing property. An example of economic adulteration is suppliers trying to sell cookies with that sort of name that contain very low levels of butter or no butter at all. If you look at the ingredients statement and determine where butter is placed, it would be very low on the ingredient list in terms of the descending order of predominance labeling required for food products. This example would constitute economic adulteration or quality misrepresentation with respect to food. One thing to keep in mind is that all these provisions in general interrelate, so it is difficult to separate one from the other. If you have a quality misrepresentation then you often have a number of other misrepresentations (misbranding) as well.

I will describe first sections 409 and 409(d) of the FFDCA. These sections relate basically to food additives. The FDA can do practically anything it wants to promulgate labeling requirements for food additives. The conditions for use of a food additive can require statements about the food ingredient being present. They can specify use levels that cannot be exceeded and, to assure the safe use of these ingredients, the FDA can also impose labeling requirements on food additives. The minute something falls in the food additive category, the FDA has a lot of discretion to require labeling for that product or ingredient. But keep in mind, this jurisdiction relates primarily to safety considerations, not to other sorts of characteristics of a food that may be related to the use of the additive, such as quality. This type of consideration would have to be addressed by other provisions of the law.

In the case of genetically engineered food, particularly products created by recombinant DNA techniques, the issue is what is a food additive and what is not? The May 1992 policy statement promulgated by the FDA describes situations where the FDA’s food additive authority may be triggered to require pre-market clearance. If some substance introduced by genetic engineering was judged to be a food additive, the agency could impose almost any labeling requirements that it wanted pertaining to safe use of that food additive. Labeling requirements could be imposed fairly easily if the FDA considered a food additive to be an allergen. Again though, keep in mind, the precedent here is not warning labels for allergens, merely ingredient declaration.

The next two provisions, Section 403(a)(1) and Section 201, are probably the statutory workhorses. The FDA uses these sections regularly to promulgate various types of statements requiring food identification. These sections state that a product is misbranded if anything said about it is false or misleading in any particular. That standard is a very general standard. The law goes on further to say that, in considering whether something is false or misleading in particular, one shall consider the representations made or the failure to reveal material facts in light of such representations or the failure to reveal material facts with respect to consequences that may result from use. This is the critical part of the statute that the FDA has considered in the biotechnology area that is relevant to food labeling: the failure to reveal material facts.

What is a material fact? I do not know and I do not think anyone knows in absolute terms what is a material fact. In fact, as I will suggest to you by the end of this discussion, a material fact is anything that the agency says is a material fact. It is difficult to develop standards for what are material facts. However, developing these standards is the central issue in this area. The question of material facts is related to consequences that may result from use is the issue of safety. Clearly, if there is a safety issue, the agency can require labeling. Of course, what the agency can require labeling of is another issue when there is no safety concern. Even though failure to reveal material facts is the most relevant precedent in this area, in terms of labeling requirement for food, these provisions have not been implemented very regularly. Only in the last few years, with the legislative adoption of the Nutrition Labeling and Education Act, has the agency begun to implement these provisions a little more aggressively. The biotechnology area is going to be another area where it will be interesting to see how these provisions are applied.

There is not a lot of precedent in this area of labeling of material facts. The singular precedent that I recall, which is also mentioned by the FDA in its notice, is food irradiation. This precedent requires labels for retail food stating that the food has been treated with ionizing radiation. If you look at the legal basis for the agency’s requirement for that information you will note that the FDA said that irradiation could affect the organoleptic properties of the food and, therefore, the food should be labeled. This is a very interesting concept. In fact, I represented a major food company at the time that was very interested in food irradiation until that labeling requirement came out. We opposed the provision saying if the labeling standard is based on organoleptic quality, then what would not have to be labeled under that standard because anything a manufacturer does to a food has some organoleptic effect. In general, food ingredients have organoleptic functions, either synergistically or otherwise. We stated that under this standard for labeling, the agency had no authority to require organoleptic labeling of foods. Needless to say, the FDA ignored our advice.

The FDA also said that it felt that if the fact of irradiation was not stated, then the consumer might misperceive or assume that the product had been conventionally processed. That is another very interesting concept behind requiring labeling. This suggests that there is a reason why consumers need to know that a food item is irradiation-processed. I would say that, if there are no safety issues associated with irradiation-processing, then how can one say the food should be labeled because consumers might need to know that it is not conventionally-processed. In fact, there were not any safety issues regarding irradiation since the agency cleared the process for use as a “food additive”. So what is the relevance of information that irradiation processing has been used? In later initiatives, the agency has taken the position that for cultural or religious or other reasons it might be material to know how a food is processed or what the source of the food may be. As a result, we find examples of very ambiguous standards which, if imposed across the board on all food ingredients, would require that almost everything about how a food is prepared is stated on the label.

What I am suggesting to you is that I think the FDA has been very creative in the interpretation of some of the statutory labeling provisions. I do not mean to suggest that there are not characteristics of any food, whether or not it is genetically engineered, that should not be labeled. That is not the point. The point is that whatever standards are adopted, they must be applied consistently across the board. If they cannot be applied in such fashion, then there is no sense in adopting standards that apply to one particular technology or to one particular ingredient or situation because of some a priori notion of what consumers “want to know”.

The rest of the provisions of the law are relatively easy to deal with in comparison to the sections I have just described. Section 403(b) basically says a food is misbranded if it is offered under the name of another food. Earlier, we considered the example of a “butter cookie” which contained little or no butter. Such a product could violate this section if it does not contain enough butter to adequately satisfy the characteristics of a true “butter cookie”. In such a case, it is being offered under the name of another food. This example demonstrates the principle that I alluded to earlier; one can violate one or more provisions of the law simultaneously.

Section 403(c) has always been the scourge of the industry; it basically requires “imitation” labeling for certain products. I cannot think offhand of any food that is labeled “imitation” today. This is because once it must be labeled as “imitation” it is “dead in the water” as a product. No consumer would buy the product without a lot of trepidation or understanding of why it is an “imitation” product. Thus, whenever labeling as “imitation” has been raised by the agency, producers reevaluate and say “oh no, we will reformulate; we will change the composition of the product so that is not an imitation”. “Imitation” is a very difficult term to define either scientifically or legally. Imitation labeling, though it is an interesting concept, has really never played an important role in the marketplace, at least in terms of foods that must be labeled with the term “imitation”.

Section 403(g) is another provision that has been used frequently. It concerns standards of identity for foods. In this instance, it is primarily the economic adulteration provisions that come into play, as well as other sections pertaining to the FDA’s authority to promulgate regulations covering “standardized foods”. The basic concept behind standardized foods is that, because they are economically so important, the FDA specifies a recipe for how they are obtained and/or made. Examples are chocolate and vanilla extract. Therefore, if you want to call a product “chocolate", it must be made in a certain way. One good example is “white chocolate”. White chocolate basically is a misnomer because the standardized recipe or requirement for a product called chocolate is that it must be made in a certain way. Since white chocolate does not contain all the standardized ingredients that “chocolate” has, the name is not correct. Although the agency never challenges use of the name, technically such a product may be adulterated and/or misbranded. Its name “white chocolate” suggests the product meets the standard of identity for chocolate, but it does not.

The legal or labeling principle here is that once you call an ingredient or product by a name that corresponds to a standardized food, you cannot vary how it is made or how it is derived, unless it varies according to the prescribed standard. There are various products of biotechnology that may fall under this principle. For example, vanilla extract could be produced by microbial genetic engineering. If you examine the vanilla standard there is a good chance that you would not be able to call such a genetically engineered product “vanilla extract” because the standardized recipe has a source requirement that is not microbial. Therefore, if the source varies you could not call it “vanilla extract”. Using the standardized name would be misbranding and would violate the law. Genetic engineering will, I think, trigger several such issues under various standardized food requirements. As a result, the importance of this provision 403(g) will become more obvious as time goes along.

Perhaps the most significant provision for biotechnology foods is section 403(i). This provision states that foods shall be identified by their common or usual name and, moreover, that food ingredients shall be designated on the labels of fabricated foods in descending order of predominance by weight and shall be identified by their common or usual name. This “naming” issue is a very significant one in the genetic engineering area because it is argued that by altering food composition through genetic engineering you somehow change the basic nature of the food. For example, several companies have proposed modifying various types of dietary oils so that they have enhanced proportions of unsaturated components because of certain health characteristics that are claimed for foods rich in unsaturated oils. One can argue for soybean oil, or for any number of other oils, that if one increased the amount, beyond a certain level, of polyunsaturates that the oil should no longer be called “soybean oil”. Questions that arise are: 1. what is the accepted standardized profile of polyunsaturated fatty acids in soybean oil and 2. how much can one vary from that profile such that the oil could still be called by the common or usual name soybean oil? The FDA has considered these questions, but the issue has not been resolved. It is a very difficult question that comes down to how much can you vary the basic nature of a product, however that basic nature is defined, before you can no longer call it by the name by which it is traditionally recognized.

Another question that may arise is whether such a compositionally altered product is inferior in some way. It may not be nutritionally inferior, but such oil alterations may still misrepresent the basic nature of the product so that its utility as an ingredient may be altered.

An interesting case is represented by the requirement for labeling products that contain glutamate. The FDA has discussed extensively the issue of requiring glutamate labeling on protein hydrolyses. It refers to this example in the food biotechnology labeling initiative by saying that it required glutamate labeling. In this case the FDA did not require labeling because of safety issues, which is very interesting in the light of the glutamate scares that have been publicized. In fact, the agency said it felt the glutamate content was material information which the public needed to know. This raises the question of why it is material for the public to know that a product contains glutamate. I would argue to you that in the absence of safety concerns or some other important characteristic, glutamate content is immaterial.

The agency also said, with respect to protein hydrolysates, that it wanted source labeling identifying the particular source of the product. One could not identify “hydrolyzed vegetable protein” as an ingredient; rather the FDA required identification of the source of the protein, such as “hydrolyzed soy protein” or whatever. Here the agency took the position that, because the amino acid profile of different hydrolyzed proteins depended upon the source, it would be a misrepresentation to generically label it as “vegetable protein”. The amino acid profile could be different depending on the specific source of the protein and that issue was central to the basic nature of the final food. I would argue that the average consumer does not know or understand amino acid profiles of proteins and, therefore, the vegetable source of the protein is totally irrelevant information in terms of the consumer, unless there is also a safety issue related to the source.

It is interesting to observe over time how the FDA interprets these various provisions of the statute. I do not necessarily suggest the various interpretations are wrong. Rather, I want to emphasize that the question is what is the standard for requiring labeling? Does the standard make sense? Legally, scientifically, from a policy standpoint, does the standard make sense? If a standard is adopted, it needs to be applied consistently across the board. The agency should not single out one case and then somehow not apply the rule to other similar situations.

Finally, let us consider sections 402(b)(1)-(b)(4). These sections are the quality standards. They refer to valuable constituents omitted or substituted for in food. Butter is again a good example. When a food does not contain butter, or when some other type of ingredient is used to replace the butter, that could represent economic adulteration as well as misbranding. Again, as I keep emphasizing, various provisions of the law can be triggered depending on what change is made or if damage or inferiority is trying to be concealed. Again, inferiority is related to the suggestion that a product is better than it really is or is deceptive as to quality or value. To repeat, this is economic adulteration and is generally a more serious enforcement issue than misbranding which usually is simply making misleading statements as to a fact, such as “natural”.

There are many familiar examples of these sorts of issues which have been reported by the press. Strength, purity, quality, and size (jumbo, big) are familiar issues. The “freshness” label has had a lot of publicity. Recall the orange juice case, labeled “fresh” orange juice, when in fact the juice was derived from concentrate or previously frozen orange concentrate or the like. There is a lot of controversy about what constitutes “fresh” these days. “Grade” is another controversial parameter; there are various grades of products (e.g., Grade A). “Naturalness” has always been the subject of a lot of controversy. Again, not an easy concept to define. What is natural, what is not, how much is it processed, does it contain artificial ingredients? As you can see, these labeling areas are very slippery slopes; once one starts down them it is difficult to back up in any sort of reasonable manner because the terms are very hard to define. Statements of origin are another issue. For example, what does the term “Italian dressing” suggest. Most people today would say the term “Italian” does not mean it comes from Italy. But again, that relies upon consumer understanding and experience which may suggest that the consumer would not be mislead. These issues come up regularly about labeling food as representative of a particular “style”. An example is “Italian-style dressing” to suggest a product did not come from Italy.

In conclusion, I think it is useful to go over again some basic labeling concepts. First of all, although all the provisions I have reviewed suggest that the consumer has a limited right-to-know about his or her food, the fact is there is no labeling provision in the FFDCA that says the consumer has an absolute right-to-know everything about his or her food. Consumers state “I have a right to know”. In fact, they do not. The law says that you have a right to know certain things about your food, but the reality is that this issue often turns on interpretation. Invariably one’s “right-to-know” is not an easy issue to address; it is not that clear-cut. There is no absolute consumer “right-to-know,” per se, however. Again, I am not trying to suggest that there are not legitimate topics that we should know about as consumers, or that particular segments of the population should not know about. Rather the critical issue is what are the labeling standards and how do you apply them?

Secondly, food additive status does not mean the additive must be declared as an ingredient. I have heard so many people say, “Well, because something is a food additive it must be declared on the ingredient statement.” No, this is incorrect! In fact, the ingredient labeling provisions are totally unrelated to food additive status. The fact that something is a food additive does not mean it has to be labeled. It does not follow that if some component of a biotechnology product is a food additive, therefore it must be labeled. Food additive status means, legally, that a substance or process requires pre-market clearance from the FDA. The ingredient labeling provisions are another section of the law that basically say that the common or usual name of the ingredient must be declared on the product label in descending order of predominance by weight.

The question then arises as part of ingredient labeling whether, if “genetic engineering” is used and introduces an ingredient, that ingredient must be labeled by its common or usual name. In general, plant breeding per se has never been viewed as introducing “ingredients” that must be labeled on food. Conceptually, arguably, one is adding “ingredients” to the genetically modified food crop. But the ingredient provision of the statute relates specifically to “fabricated” foods: it requires listing of the common or usual name of the ingredients of “fabricated” foods. What does the term “fabricated” suggest? It suggests that the producer is manipulating the food in a manufacturing situation and adding ingredients as part of a final recipe. It does not cover breeding a crop and “adding ingredients.” The distinction is that whole foods produced by plant breeding are not “fabricated foods”. Thus far, according to the agency, plant breeding does not trigger automatically the ingredient listing requirement we have discussed.

Third point: what is the jurisdiction of the FDA, the U.S. Department of Agriculture (USDA), the Bureau of Alcohol, Tobacco and Firearms (ATF), and the Environmental Protection Agency (EPA)? Most of these agencies have concurrent jurisdiction over “foods”. We have only talked about one agency of these four, the FDA. The USDA has its own labeling requirements, and they can differ from those that we have just described. The USDA and the FDA have sometimes been at odds about how foods should be labeled because their statutory authorities are different. These differences are sometimes difficult to reconcile. Meat and poultry products are subject to USDA jurisdiction as well as to FDA oversight. The ATF regulates alcohol products and these are subject to different labeling standards. Again, all the labeling statutes I have discussed do not apply to the EPA. The EPA has its own standards. That is not to say always that the FDA does not have jurisdiction, it does. But there are other agencies that have their own laws too and they operate under different standards.

Finally, labels and labeling: the standards are usually the same from an FDA standpoint. Information provided directly on the immediate container of a food product, that is the so-called “label”. Anything that is used in conjunction with marketing a food is considered labeling. As a result, if you want to make statements about food products in brochures or press releases or other informational sources, the FDA can claim that the article to which that labeling relates misbrands that food. Therefore, the principles that we have talked about apply not only to labels on foods, but to anything that is used to market that food in furtherance of selling it.

The food provisions we discussed do not always apply to advertising. However, the distinction between advertising and labeling is often very difficult to delineate and the truth of the matter is that most everything that is distributed in furtherance of marketing of food products probably is labeling. In some areas, The Federal Trade Commission (FTC) has taken jurisdiction over advertising, but, in general, the FDA tends to focus on immediate containers of foods and the labels attached thereto. Everything else the FTC often has regulated.

To go back to the fundamentals question: should genetically engineered foods be labeled? I have no idea. A very easy answer is “No,” as the question is framed. The reason why the answer is easy is because there are too many scientific, legal, and policy questions involved in the determination to say “Yes.” Anyone who answers such a general question “Yes” usually does not know what they are talking about.