Labelling And Risk: The Case Of Bioengineered Foods

More and more, consumers are basing their food purchases on individual preferences about food content.  For many consumers this means a focus on nutrition or fat.  Others care more about the methods of production, taking into consideration whether a product is kosher or organic.  Some choices are made out of concern for personal health and well-being; others reflect subjective value preferences.  In all cases, information on labels helps consumers select the items they prefer.

Given the choice, many consumers say they would like to know about the “genetic status” of their foods.  The relative novelty of biotechnology, specifically the use of recombinant DNA techniques to modify crops and organisms, has created concern that some “bioengineered” foods are less safe, or may harm environmental quality.  Some have called for a moratorium on biotechnology research, others for greater regulation.  At the very least, critics argue, bioengineered products should be labeled as such.

The call for labeling seems persuasive.  After all, don’t consumers have the right to know what’s in the very food they eat?  But mandatory labeling of products is not the best way to provide consumers with information about products or even potential risks.  A better solution is to let producers respond to consumer pressure in the marketplace for information that consumers want most.



The first biotechnology food products became available in 1990 with the introduction of a bioengineered enzyme used to produce cheese.[1]  The Food and Drug Administration (FDA) approved recombinant bovine Somatotropin, a bioengineered growth hormone used to boost milk production in cows, in 1993.[2]  And the first bioengineered plant, the Calgene Corporation’s FlavrSavr slow-ripening tomato, was introduced in 1994.[3]  Since that time, the FDA has evaluated and approved more than 40 new bioengineered plants.[4]


In 1992, the FDA clarified its policy on the labeling of biotech food products.  The agency’s policy is that foods developed from new plant varieties must be labeled if, and only if, the composition of those foods differs “significantly” from their conventional counterparts.  Such differences may include the introduction of an allergen that is not present in the new plant’s conventional counterpart, or the reduction in nutrients from what would be expected of the new plant’s conventional counterpart.  This minimal requirement is applied to all modified foods, whether developed through traditional plant-breeding methods or more advanced genetic techniques.


Producers are permitted to label bioengineered products voluntarily, and some have test-marketed affirmatively labeled products.  Calgene’s FlavrSavr tomato was voluntarily labeled, and was initially well-received by consumers, many of whom were willing to pay a premium for the improved flavor promised on the labels.[5]  Still, most producers have heretofore refused to label their bioengineered products for fear of increased cost and consumer rejection—especially when the added trait provides benefits mainly to producers, not consumers.  However, once bioengineered food products begin to add benefits for which consumers will be willing to pay a premium—and not just reduce the cost of production—producers are likely to label them voluntarily.[6]

Importantly, FDA policy on the labeling of foods derived from new plant varieties is based upon the “objective characteristics of the food product or its components rather than the fact that new development methods were used.”  This standard is consistent with the general scientific consensus that plants developed with new biotechnologies are not inherently more risky than those developed with conventional techniques.  Over the last few decades, thousands of new plant varieties created with recombinant DNA techniques have been developed, tested, and then marketed without incident. There is no reason to believe that bioengineered food products now on the market have caused any harm to human or animal health.[7]

Nevertheless, many critics of biotechnology have challenged this policy on the labeling of bioengineered foods.  In 1993, when the FDA approved recombinant Bovine Somatotropin (rbST), anti-biotechnology advocates demanded that all dairy products from cows treated with rbST be labeled as “bioengineered” or “genetically modified.”  When the FDA refused, activists turned their efforts to state legislatures and regulators—with variable degrees of success.  More recently, however, in the wake of a successful campaign within the European Union, labeling advocates have renewed efforts to convince federal regulators and Congress.


The Harms of Mandatory Labels


When the introduction of a new plant variety (produced by recombinant DNA methods or otherwise) changes the composition of foods significantly from what would otherwise be expected from the conventional variety, the labeling of those foods to reflect that change is scientifically defensible.  But because recombinant DNA techniques themselves pose no apparent danger to human or animal health, a mandatory label on all bioengineered foods can not be justified.  If consumers want the information, it will be provided in the marketplace.

In 1996, the United States Second Circuit Court of Appeals, in the case of International Dairy Foods Association, et al. v. Amestoy,[8]  noted that a labeling mandate not grounded in a product’s measurable characteristics raises serious constitutional concerns.  The court held that food labeling cannot be mandated simply because some people would like to have the information.  That decision held unconstitutional a Vermont statute and companion regulations requiring that “[i]f rbST has been used in the production of milk or a milk product for retail sale in [Vermont], the retail milk or milk product shall be labeled as such.”[9]


Because the state of Vermont could not demonstrate that its interest in compelling acknowledgment of rbST represented anything more than satisfying consumer curiosity, it could not compel milk producers to include that information on product labels.  In the words of the court,


Absent some indication that this information bears on a reasonable concern for human health or safety or some other sufficiently substantial governmental concern, the manufacturers cannot be compelled to disclose it.  Instead, those consumers interested in such information should exercise the power of their purses by buying products from manufacturers who voluntarily reveal it. (emphasis added)


In other words, to be constitutional, labeling mandates must be based in science and confined to requiring disclosure of information that is relevant to health or nutrition.

Of course, many consumers care about more than the safety and nutritional content of the food they buy.  Other information about food products can be extremely important to consumers due to religious, ethical, and/or other strong personal values.  Such information may include certification of kosher or organic production, or other characteristics pertaining to process attributes.  Insofar as consumers truly care about the techniques used to produce their food, there is an incentive for companies to respond by providing it.  For instance, if many consumers want products free of any bioengineered ingredients, companies that meet this need can gain market share by advertising this fact.  The growth of the organic food industry, despite the lack of formal labeling requirements (as of yet) is a testament to how this process works when consumers demand products made in a particular manner.


Information Overload


One often-neglected problem of mandatory labels is that of “information overload” or the “crowding out” of important information.  Many consumers turn to food labels to get important information about such things as nutrient and fat content, the presence of certain allergens, or even proper storage and preparation information.  Each of these bits of information can have a material effect on their health.  The appearance on a food product label of too much information about materially irrelevant facts—those not pertaining to legitimate issues of health—can make it more difficult for consumers to locate important facts about the foods they consume.

Indeed, the appearance of too much information increases the risk that consumers will pay less attention to individual messages within labels, making it more difficult to transmit information regarding real hazards.[10]  Federal laws mandate that all sorts of information be included with over-the-counter and prescription drugs, but how many people actually take the time to read every page of the tightly printed inserts?  As a Michigan Appeals Court noted in Dunn v. Lederle Laboratories, “[m]aking consumers account mentally for trivia or guard against risks that are not likely to occur imposes a very real societal cost.”


Though there is a danger of requiring too much information, some consumers will have an interest in information about the development and/or processing of foods.  It is important, then, that the FDA abstain from micromanaging the specific content of food product labels. Where consumer demand is strong enough, niche markets will emerge to provide consumers with desired information in a format that they find easy to use. When it comes to information about consumer products, especially information relating to personal values, one size does not fit all—consumer preferences are not homogeneous.  To grant consumers a real choice, the FDA need not mandate that specific information appear on labels; it need only protect the right of producers to make accurate, truthful statements.


Competition and Information


The importance of private, voluntary labeling schemes should not be understated.  Much research indicates that voluntary labeling and advertising transmits useful and important information to consumers in a format they find easy to understand.[11]  Indeed, market competition among producers for providing demanded information can even drive producers to make better products.  Studies of voluntary food health claims, conducted by researchers at the US Federal Trade Commission, found that, when food manufacturers were permitted to mention diet-disease relationships on their package labels, consumers began to reduce their dietary intake of fat, saturated fat, and cholesterol, and to increase fiber intake.  Moreover, the resulting competition among manufacturers to provide this information encouraged manufacturers to make healthier products.[12]


Mandatory-labeling advocates argue that the general resistance by producers of bioengineered foods to label products voluntarily represents a failure of the market to provide information that consumers want.  However, it is not clear that, in the case of bioengineered foods, the market has failed to supply demanded information.  Nor is it clear that a government mandate to affirmatively label bioengineered products will provide better or more useful information.  Many real-world examples show that market forces are fully capable of supplying information about process attributes for food products that consumers truly demand.  Consider two real-world examples: kosher and organic foods.


For religious purposes, many Jewish people purchase foods that have been processed according to the Kashrut, or kosher, dietary rules.  There are a large number of private kosher certification organizations that keep to varying degrees of Kashrut strictness, and some purchasers of kosher products seek out only foods approved by specific certifying organizations.[13]  In this way, kosher certification provides a range of information useful to consumers.  The existence of competing certification organizations provides a level of consumer choice that would be impossible under a government-imposed, one-size-fits-all labeling scheme.


Voluntary labeling within the organic food industry represents an even closer analogue to the process-based information now requested of bioengineered foods.  Many people believe that organically produced foods are, in some way, healthier than conventionally produced foods, and they seek out products that are labeled organic because organic agriculture shuns synthetic chemical pesticides, herbicides, and fertilizers.  Even though there doesn’t appear to be a strong consumer demand for products that are certified organic (as opposed to self-identified),[14]  some 33 private organic-certifying organizations have sprung up in the United States alone, to compliment government and international standard-setting bodies.  And while there has been a concerted voluntary movement within the organic industry to codify standards for “organic” certification,[15]  a move by the US government to create a single, national standard failed to satisfy many in the organic food industry.[16]  The Organic Trade Association lamented that “the proposed federal regulations will actually be weaker than the current decentralized regulation practices.”[17]


The organic example is an important one within the context of this paper, as many organic purists insist that foods produced with bioengineered materials can not be considered organic.[18]  Consequently, to ensure that they are purchasing only foods produced without bioengineering, consumers need only look for an appropriate organic certification.  Furthermore, voluntary labeling and certification of non-bioengineered foods specifically is already being pursued.  In 1999, one of the largest organic-certifying organizations formed a private certification company established to meet “growing demand by consumers, governments and the food industry for rigorous and ethical third-party certification of food and fiber production.”[19]  And members of the organic and “natural” food industry in the United Kingdom formed an organization known as Genetic Food Alert, to register food producers that do not use bioengineered products.[20]  Each of these examples demonstrates that, when consumers truly demand certain types of information, the market usually finds a way of supplying it—even in cases where that demand is not extensive.


One measure of robust consumer demand is willingness to pay.  Because certification and labeling are not free, consumers in these examples demonstrate that they are, in fact, willing to pay for the information they receive.  To the extent that a legitimate consumer demand for information about the genetic status of food products currently exists, it comes primarily from consumers wishing to purchase foods that are not bioengineered.  Requiring producers of bioengineered foods to label their products as such shifts the enormous cost of providing this information from those who are demanding it onto consumers who are not—a position that hardly seems reasonable or equitable.




Recombinant DNA techniques have already been used to develop crop plants with traits that increase yields and that allow farmers to reduce their use of synthetic pesticides and herbicides.[21]  The next generation of products promises to provide even greater benefits to consumers, such as enhanced nutritional value and improved medicines.  But a labeling requirement for these products could impose high costs on producers and consumers, without producing any corresponding benefits.  Ultimately, the monetary costs of a labeling mandate could all but preclude many consumers who are indifferent to genetic status from sharing in the vast potential benefits of the bioengineering revolution.

Leaving labeling to private markets encourages firms to differentiate themselves by providing information about their products that actually matters to consumers.  At the same time, economic pressures encourage producers to present this information in the most direct and accessible manner possible.  The only real role for the federal government in this enterprise is to guard against outright fraud.  Whether the issue is biotechnology or some other food characteristics, the rest should be left to the market.


End Notes


1. James H. Maryanski, “FDA’s Policy for Foods Developed by Biotechnology,” in Genetically Modified Foods: Safety Issues, Engel, Takeoka, and Teranishi, eds., American Chemical Society, Symposium Series No. 605 (1995), pp. 12-22.

2. Center for Veterinary Medicine of the US Food and Drug Administration, “BST Update,” CVM Update, March 21, 1996.

3. Center for Food Safety and Applied Nutrition of the US Food and Drug Administration, Biotechnology of Food, FDA Backgrounder, May 18, 1994.

4. Food and Drug Administration, “Notice: Biotechnology in the Year 2000 and Beyond; Public Meetings,” Federal Register, Docket No. 99N-4282; 57 FR 57470, October 25, 1999.

5. Karen K. Marshall, “What’s in a Label?” AgBioForum, vol. 1, no. 1 (Summer 1998), pp. 35-37.  It should be noted that, while sales of the FlavrSavr tomato eventually faded, this is usually attributed to the use of inferior tomato varieties, rather than rejection of bioengineering.

6. Peter W.B. Phillips and Grant Isaac, “GMO Labeling: Threat or Opportunity?” AgBioForum, vol. 1, no. 1 (Summer 1998), pp. 25-30.

7. David L. Aaron, “Testimony of Ambassador David L. Aaron, Under Secretary for International Trade, US Department of Commerce, before the Senate Committee on Finance,” October 7, 1999; and Mark Silbergeld, “Testimony of Mark Silbergeld, Co-Director of the Washington Office of Consumers Union, Before the Committee On Agriculture, United States Senate,” October 7, 1999. Even Silbergeld, a leading critic of bioengineered foods, admitted in his testimony that “there is no evidence of harm to consumers.”

8. 92 F.3d 67 (2nd Cir. 1996).

9. Vt. Stat. Ann. tit. 6, § 2754(c), overturned by the International Dairy Foods Association decision.

10. See, for example, J. Paul Frantz, et al., “Potential Problems Associated with Overusing Warnings,” Proceedings of the 7th International Conference on Product Safety Research (Washington, DC: European Consumer Safety Association and the US Consumer Product Safety Commission, 1999); Aaron D. Twerski, et al., “The Use and Abuse of Warnings in Product Liability—Design Defect Litigation Comes of Age,” Cornell Law Review, vol. 61, no. 4 (1976), pp. 514-517; and Aaron D. Twerski, “From Risk-Utility to Consumer Expectation: Enhancing the Role of Judicial Screening in Product Liability Litigation,” Hofstra Law Review, vol. 11 (1983), pp. 861-935.

11. See, for example, George Stigler, “The Economics of Information,” Journal of Political Economy, vol. 69 (1961), pp. 213-225; and Pauline M. Ippolito, “Consumer Protection Economics: A Selected Survey,” in Empirical Approaches to Consumer Protection Economics, Pauline M. Ippolito and David T. Scheffman, eds. (Washington, DC: Federal Trade Commission, 1986), pp. 1-33.

12. Pauline M. Ippolito and Alan D. Mathios, Information and Advertising Policy: A Study of Fat and Cholesterol Consumption in the United States, 1977-1990, Bureau of Economics Staff Report (Washington, DC: Federal Trade Commission, September 1996); and Pauline M. Ippolito and Alan D. Mathios, Health Claims in Advertising and Labeling: A Study of the Cereal Market, Bureau of Economics Staff Report (Washington, DC: Federal Trade Commission, 1989).

13. See, for example, Michael I. Krauss, “Loosening the FDA’s Drug Certification Monopoly: Implications for Tort Law and Consumer Welfare,” George Mason Law Review, vol. 4, no. 3 (Spring 1996), pp. 457-483.

14. Carole Sugarman, “Organic?  Industry is Way Ahead of Government,” Washington Post, December 31, 1997.

15. Organic Trade Association, American Organic Standards: Guidelines for the Organic Industry, compiled by James A. Riddle, Emily Brown Rosen, and Lynn S. Coody (Greenfield, Massachusetts: Organic Trade Association, 1999), p. 19.

16. See, for example, Sugarman, “Organic?”; and Anon. “Organic farming: Let’s keep it clean,” The Economist, April 17, 1999, pp. 34-35.

17. Organic Trade Association, “Organic Standards At Risk,” press release, January 27, 1998.

18. Organic Trade Association, American Organic Standards, p. 10.

19. Genetic ID, “New Company Forms to Provide Enhanced Consumer Safeguards for Food,” press release, January 28, 1999.

20. Genetic Food Alert, “GM-Free Registration Scheme for Businesses and Products,” current draft as of January 11, 1999,

21. See, for example, Leonard P. Gianessi and Janet E. Carpenter, Agricultural Technology: Insect Control Benefits (Washington, DC: National Center for Food and Agriculture Policy, 1999); and Economic Research Service of the US Department of Agriculture, Bioengineered Crops for Pest Management, October 27, 1999,