Modifying Mother Nature

By Martin Zelder
November 1999
Fraser Forum

Perhaps the most dreaded risk these days is that hypothesized to surround genetically-modified foods. In the view of the Campaign for Food Safety, "With little or no regulatory restraints, labelling requirements, or scientific protocol, bio-engineers have begun creating hundreds of new GE [genetically-engineered] 'Frankenfoods' and crops, oblivious to human and environmental hazards" (Cummins, 1999). In the face of such impassioned rhetoric, it seems hard to resist pronouncements like those of Greenpeace that "Genetically engineered food must be segregated and labelled" ("Genetically Engineered Food," 1999) or of David Suzuki that there should be a moratorium on such foods ("Suzuki Warns of Frankenstein Foods," 1999).

But capitulating to such calls for regulation ignores both facts and logic regarding genetically-modified (GM) foods. The facts are that the risks purported to attach to GM foods are overstated; the logic is that optimal regulation of GM foods does not involve such draconian strictures as segregation, labelling, or prohibition.

The Facts Of GM Risks

The risks ascribed to GM foods are two-fold: damage to human health, and damage to the environment. Yet the evidence supporting these concerns is minimal. To address this evidence, it is useful to provide a brief description of just what GM foods are.

Genetic engineering involves "the detailed manipulation of individual genes.... offering a precision in manipulating genetic material hitherto unobtainable" (Mantell, Matthews, and McKee, 1985, pp. 5-6). Current advances in genetic engineering constitute gene transfer, whereby genes are infused from an unrelated organism. An example is the transfer of the lucerifase gene, which makes fireflies glow, into plants. Consequently, gene transfer allows mutations to occur which are precluded by natural selection or by traditional cross-breeding; both of those forms of genetic alteration depend on extensive enough genetic similarity to allow breeding. Thus, gene transfer permits a wider range of gene exchange than that historically possible in nature.

The critical question is whether this expanded set of possible modifications implies expanded risks. One prominently articulated concern centres around the safety of consuming GM foods. The essence of this qualm has been expressed by Gillian Hadfield, a law professor at the University of Toronto: "if you have no idea whether the tomato you are buying is conventional or genetically engineered, how can you link any adverse reaction you have (or don't have) to its source?" Consequently, according to Hadfield, "We need to label biotech products to preserve our ability to make the unknown known" (Hadfield, 1999).

Hadfield's characterization of the risk inherent in GM foods is exaggerated in two respects. First, only two adverse health reactions from ingesting GM foods have ever been recorded, a point acknowledged by another proponent of labelling, Michael Antoniou, a prominent British biochemist (Antoniou, 1999). Indeed, even safety-conscious Consumer Reports acknowledges that "There is no evidence that genetically engineered food now on the market is unsafe to eat" ("Seeds of Change," 1999, p. 46).

The second exaggeration of risk, that GM foods are different, is more fundamental, but equally untrue: "there is no inherent difference between a bioengineered food and a non-bioengineered one" (Fumento, 1999). What Fumento refers to is the concept of "substantial equivalence," the standard by which Canada and the United States compare GM and traditional foods. Substantial equivalence, developed by the OECD and the World Health Organization, means that a GM tomato (for example) and a conventional tomato are chemically identical if they are "compositionally and nutritionally similar within the limits of the normal biological variation in edible varieties of the plant" (Jonas, 1996, p. 263).

A second purported risk from GM foods is hypothesized to lie in the environmental consequences of growing them. This environmental risk falls into four categories: GM crops may become or create weeds, they may harm other organisms, they may create resistance among pests to genetically-transferred toxins, and they may create new plant viruses (Rissler and Mellon, 1996). The general economic problem that each of these illustrates is one of externalities - costs potentially imposed by one party on another.

While it beyond the scope of this short overview to assess the literature on the seriousness of these environmental threats, some reputable ecologists caution against overreaction to them. According to one view, the probability of harm from any GM organism is the product of six underlying probabilities: the probability the organism will be released, will survive, will multiply, will be transported to an effective location, contains genetically-transferable information, and will be harmful (Alexander, 1985, p. 63). Consequently, if each of these component probabilities is 1 percent, the probability of harm is one in a trillion! Even the authors of a recent book recommending substantially enhanced regulation for environmental reasons admit that "The long-term impacts of transgenic plants in the environment are unpredictable and difficult, if not impossible, to predict at the current level of scientific understanding" (Rissler and Mellon, 1996, p. 125).

The Logic Of Optimal GM Regulation

In the face of these hypothesized risks to human health and to the environment, a number of commentators have urged expansive regulation, even including a moratorium on GM foods (Reaney, 1999; "Suzuki warns of 'Frankenstein foods,'" 1999). Many of these prescriptions, however, fail to provide efficient mechanisms for deterring risk. Indeed, instead of banning or imposing labelling, more parsimonious solutions are preferable.

Regulating Human Health Risks

As noted, the possibility of medical risk from consuming GM foods has elicited support for mandatory labelling (currently in place in the European Union, New Zealand, and Australia) or even outright bans or moratoriums (currently effective in the EU). These more extreme measures are inappropriate, however, both in the countries where they exist, and in Canada.

A less extreme mechanism can be found in products liability law. Products liability law is founded on the understanding that consumption of products of all kinds can lead to direct harm to the individual consumer. Because of this, purchase and use of products inherently involves risk of such harm, risk which is allocated by law between the consumer and the producer. Economic analysis of products liability law consequently indicates that the liability for harm should be allocated so as to minimize the cost of avoiding harm.1 Thus, when it comes to foods, consumers have a reasonable legal expectation that consumption will not be imminently harmful, and the basis to recover damages if it is. Clearly, liability law, if sensibly extended to cover GM foods, will efficiently deter sellers from purveying harmful foods.

This efficient risk-deterrence does not depend on mandatory labelling. Given that many consumers seem to prefer to avoid GM foods, firms have begun providing explicitly non-GM foods (The Economist, 1999). The only centralizing action required is to provide a definition of "non-GM" as less than, say, 5 or less than 1 percent GM, the typical issue currently encountered in classification of conventional foods. Of course, these new non-GM foods will come at a premium; it is estimated that separating GM from non-GM crops (currently co-mingled) would double the price of soybeans and corn (The Economist, 1999, p. 21). Nevertheless, firms which specialized in non-GM products would presumably be able to lower the cost over time.

Regulating Environmental Risks

In some quarters, these proposed risks have been referred to as 'genetic pollution.' This rubric, while somewhat inflammatory, is also appropriate in guiding policy responses. Rather than banning GM crops due to their risks of negative externalities for other farmers and the ecosystem, the lessons of optimal pollution control should be applied. Specifically, the discrete harms potentially generated by GM crops can be efficiently discouraged by standard pollution control devices: clear assignment of property rights, or, if necessary, fines corresponding to damages. Such mechanisms have already been tested in the UK, where a February judgement assessed a $25,000 fine against Monsanto for failing to maintain an adequate barrier between its GM experiment and adjacent natural fields (Dyer, 1999).

Overcoming The Rhetoric

It is easy to raise apocalyptic concerns regarding GM foods. Nobel laureate geneticist Fran¨ois Jacob discerns the source of these fears in "depictions of the Last Judgment rendered over the course of the centuries. .. [in which] the most horrible monsters, those charged with torturing sinners, are precisely unnatural hybrids: revolting mixtures of fish and dog, rat and insect, man and bird." Consequently, "genetic engineering has ... incited passion and hostility.... Not so much because of its dangers, which have been debated and are not worse than those scientists overcame long ago in experiments on bacteria and pathogenic viruses, but simply because the idea ... of what we call 'genetic manipulation' ... seems to us to border on the supernatural" (Jacob, 1999). These primordial fears are, consequently, exaggerated. GM foods offer unprecedented and enormous benefits in terms of expanded food supplies and concomitant health benefits. Squandering that potential with overzealous regulation is truly something to fear.

Martin Zelder is Director of Health Policy Research at The Fraser Institute. He received his Ph.D. in Economics from the University of Chicago.


1.That it has not always done so is well-documented (Huber, 1988).


Alexander, Martin (1985), "Ecological Consequences: Reducing the Uncertainties," Issues in Science and Technology 1:57-67.

Antoniou, Michael (1996), "Genetic Pollution," Nutritional Therapy Today 6:8-11.

Cummins, Ronnie (1999), "Hazards of Genetically Engineered Foods and Crops: Why We Need a Global Moratorium," Campaign for Food Safety (digital document:

Dwyer, Gwynne (1999), "Frankenstein Foods," The Globe and Mail, February 20, 1999.

The Economist (1999), "Food for Thought: Genetically Modified Food," June 19,1999.

Fumento, Michael (1999), "Consumer Reports Biotech Bashing Falls Flat," (digital document:

"Genetically Engineered Food" (1999), Greenpeace International (digital document: (

Hadfield, Gillian K. (1999), "We Need a Label to Identify Genetically Altered Foods," The Globe and Mail, May 10, 1999.

Huber, Peter W. (1988), Liability: The Legal Revolution and its Consequences, New York: Basic Books.

Jacob, Fran¨ois (1999), Of Flies, Mice and Men, Cambridge: Harvard University Press.

Jonas, David (1996), "Safety, Regulation and Innovation in the Food Sector," Current Opinion in Biotechnology, 7:262-264.

Mantell, S.H., Matthews, J.A., and McKee, R.A. (1985), Principles of Plant Biotechnology: An Introduction to Genetic Engineering in Plants, Oxford: Blackwell Scientific Publications.

Reaney, Patricia (1999), "BMA Seeks Moratorium on Genetically Modified Crops," The Globe and Mail, May 18, 1999.

Rissler, Jane, and Mellon, Margaret (1996), The Ecological Risks of Engineered Crops, Cambridge: The MIT Press.

"Seeds of Change" (1999), Consumer Reports, September 1999, pp. 41-46.

"Suzuki Warns of 'Frankenstein Foods" (1999), Vancouver Sun, October 18, 1999.

Return to the home page of Better Foods