Home Page Link AgBioWorld Home Page
About AgBioWorld Donations Ag-Biotech News Declaration Supporting Agricultural Biotechnology Ag-biotech Info Experts on Agricultural Biotechnology Contact Links Subscribe to AgBioView Home Page

AgBioView Archives

A daily collection of news and commentaries on

Subscribe AgBioView Subscribe

Search AgBioWorld Search

AgBioView Archives





October 24, 2000


Uninformation and the choice paradox: GM Food Labeling


Uninformation and the choice paradox
Alan McHughen

Nature Biotechnology
October 2000 Volume 18 Number 10 pp 1018 - 1019

Alan McHughen is a professor and senior research scientist at the
University of Saskatchewan, Saskatoon, SK S7N 5A8, Canada (e-mail:

One of the few unifying features of the genetic food fight in industrial
nations is that a majority of consumers appear to support mandatory
labeling for products of genetic modification (GM) technology. Proponents
of mandatory and indiscriminate GM food labeling congregate under the
banner of "informed choice." They argue that if DNA is introduced into
foods using recombinant DNA technologies, people ought to be able to know
about it. Unfortunately, the problem with mandatory GM labels is that they
inform no one and they diminish consumer choice.

Let us examine how this paradox arises. Following the European model, a
food label might proclaim: "This product contains genetically modified
organisms (GMOs)," or even, less assuredly, "This product may contain
ingredients from genetically modified organisms."

What we need to look at is the manner in which discriminating consumers
might be informed by such statements. For instance, some people want
labels because they consider DNA itself a food contaminant. In a recent
survey, only 40% of respondents in the UK recognized that non-GM tomatoes
contain genes. The suggested label formulations do not say much about DNA,
and they certainly do not unpick the basic biological misconceptions of
this group.

What about vegetarians? They surely are entitled to ask whether a
particular food contains an animal gene or protein. Leaving aside the
scientific niceties of species interrelationships or the technicalities of
gene homology across kingdoms, we can still conclude that the labels would
not be useful to vegetarians.

Other label-perusers might include the technologically curious who seek
out and selectively purchase the ultramodern GM materials. The labels
might help them, but only if they are truthful, which—as we will see—may
not be the case.

The most vociferous groups want labels to avoid the dreaded "Frankenfoods"
entirely. The "anti-GM" group can be divided into many factions.
Process-based opponents are averse to GM technology, period. These
consumers often base their opposition on ethical or religious beliefs,
saying, like HRH The Prince of Wales, that we humans have no business
tinkering in God's domain, regardless of how useful the end result.

Another group, probably encompassing the majority of consumers, is
concerned about specific products of the technology. They might accept GM
insulin or nutritionally enhanced GM rice, but don't want
pesticide-resistant soybeans. Indiscriminate GM labels will be of no use
to this group, because there's no distinction between the "acceptable"
rice and the "unacceptable" soybean. Without any explicit distinction on a
label, there can be no "informed choice" to satisfy these discriminating
shoppers. Similarly, some differentiating consumers might wish to avoid
eating GM DNA or protein, but might accept a refined GM food product
devoid of DNA or novel protein. Lecithin, for example, a common food
additive made from soybeans, contains no DNA or protein. Lecithin from GM
soybeans is identical to lecithin from non-GM beans. Some consumers might
look to their food labels to steer them away from foods containing
lecithin from GM soybeans. The mandatory indiscriminate labels proposed
would not help them at all.

The basic difficulty here is that the proposals for GM labels cannot take
into account the range of purposes for which people might use them.
Contrast that with, for instance, labeling information that indicates the
presence of ingredients derived from peanuts or other nuts. The reason for
the label there is to enable people who may suffer nut allergies to avoid
foods that could cause them severe and possibly fatal anaphylactic
reactions. Both the label's audience and the reason for their concern are
abundantly clear.

But let us continue to consider the issue of GM labeling. To what kinds of
products could they apply? Let's argue (for the hell of it and in direct
contradiction of all the scientific evidence) that we can find a sensible
reason to put a "contains GM" label on cross-species products. We might
then label tomatoes modified with a bean gene, or beans modified with a
tomato gene. But would we label vegetable soup containing the two types of
GMOs? Before processing, both the tomatoes and the beans are detectably GM
products. In the soup, however, the genes and proteins of both tomato and
beans are blended together. If we labeled the soup, we would be doing so
based solely on the process or method of crop breeding (an event that may
have occurred many years before) and not on the composition of the final
bouillon. GM labeling in this case would be entirely misleading and would
contradict safety priorities, which should be based on the composition of
the product.

There are, therefore, lots of reasons why GM labels are uninformative, and
some reasons why they are downright misleading. But where is the paradox?
How could it be, with strict regulations and penalties, that an unlabeled
food product could carry GM ingredients? And how, with mandatory and
indiscriminate labeling of all GM foods, can we avoid GM foods by choosing
products labeled "Contains GM ingredients"?

"GM-free" might not be
The paradox rests on the fact that negative labeling can never be
justified. Those consumers who wish to avoid GM food—for whatever
reason—constitute a legitimate market for non-GM products. Negative labels
such as "GM-free!" or "This product contains no GM ingredients" could help
such people identify desired foods. However, the retailers or
manufacturers printing such labels could never prove their claims. In
theory as well as in practice, negatives can't be proven.

In many jurisdictions, the onus is on the "labeler" to prove the claim
before the label is allowed. Even without the legal aspects, we need to
recognize the difficulty in assigning such labels. Vegetable oils, for
example, may contain no detectable DNA or protein—GM or otherwise. The
claim, "contains no GM ingredients," tells the purchaser nothing: if no
DNA or protein, GM or otherwise, is detectable, how does anyone know for
sure it didn't come from GM plants? Without a reliable means of detection,
there's no means of enforcement. Shady oil processors could quickly debase
the system by printing "contains no GM ingredients" on bottles of refined
GM maize oil, content in the knowledge that regulators have no way to
detect the origin of the product.

That hasn't stopped several groups demanding mandatory labels for such
undetectable products. For the labels to serve any purpose, officials
would need to find a way of enforcing such a regulation. In most
democracies, even an accused corporation is innocent until proven guilty,
and legal proof of guilt would require evidence. Prosecuting authorities
would have to develop ways of detecting an undetectable GMO. Legal
technicalities dating back to the Magna Carta will be likely to interfere
with effective prosecution. In any case, as long as there are
GM-containing products, occasional mixing errors, intentional or
otherwise, will ensure that no product can legitimately claim to be devoid
of GM material.

The authorities cannot seek refuge in negative labels. But maybe that does
not matter. After all, if a food is not labeled as "containing GM
ingredients," then surely it must be GM-free? Unfortunately, things are
not that simple. Current GM labeling regulations stipulate that foods
containing GMOs, or detectable DNA and protein from GMOs, must be labeled.
Products without labels are not necessarily GM free.

In the real world—the one we all share—there are few absolutes. In
addition to the various exemptions from the labeling requirements (in the
UK, for example, certain glass jars and single-serve packets), regulations
dictate tolerance limits for every commodity. Every food carries
contaminants, usually innocuous things like microscopic amounts of dirt,
insect parts, bacterial spores, or fungal toxins. But regulations allow
food processors a tolerance. As long as the contaminants are less than the
regulated limit, there's no need to declare them. Organic producers, for
instance, rather generously are allowed to include up to 5% nonorganic
"contaminants" without risking their organic label. The same principle
applies to GM ingredients. The European Union (EU) has approved a
tolerance of 1% for GM ingredients. Thus a food carrying just under 1% of
GM ingredients does not need a "GM" label.

GM ingredients might slip into other foods under conventional tolerances
for impurities. Say, for example, regulations allow 2% impurities in a
grade of conventional vegetable oil. Suppose now we have a batch of
conventional rapeseed oil approved at 98% pure. It would avoid triggering
the mandatory GM label. Can the remaining 2% be GM rapeseed oil without
requiring a label? If "yes," doesn't that defeat the purpose of the
indiscriminate GM labeling? If "no," doesn't that defeat the purpose of
the 2% impurity allowance for this grade of oil? What if we don't know
exactly what the impurity is? Does the approved batch get a GM label on
the chance that it might be GM material? As long as "non-GM" products sell
at a premium, unscrupulous processors will be tempted to carefully blend
oil from GM varieties to just below the threshold level in a
"conventional," non-GM batch, and sell it as "non-GM" oil.

The real world also must deal with costs. The cost of labeling is far more
than just the ink and sticker. The expense comes in administering the
segregated and monitored growing, transport, processing, testing,
shipping, and marketing all of a particular product, say, tomatoes, and
not just the GM ones. Each step adds an additional expense burden borne
ultimately by the consumer. The administrative expense will be that much
greater for multiple-ingredient processed foods. "GM-free" will be
especially expensive, in that the burden to ensure purity within the
tolerance limit at every step from growing, transport, processing
shipping, and marketing is particularly onerous. The premiums commanded by
organic commodities may give a good guide to the magnitude of the cost.

Thus consumers wishing to avoid GM entirely will pay more for the
privilege. But they're being misled and misinformed: they are still
consuming at least small amounts of GM ingredients. Neither explicit nor
implicit claims ("GM-free" labels, or the absence of "GM" labels) provide
these people with the information they require.

Food labeled "contains GM" might not
We have already seen that the information value of GM labels is threatened
by unscrupulous producers. Oddly, it is threatened as much if not more by
impeccably scrupulous manufacturers or retailers, too. Regulations come
with penalties that apply when they are broken. In the UK, for instance,
food producers and retailers who fail to label GM foods face fines of up
to £5,000 (US $7,500) or incarceration. The scale of these official
rebukes is, of course, next to nothing compared to the retribution that
the market could exact on wrongdoers exposed in the media. In contrast,
there is no penalty for putting labels saying "may contain GM" on non-GM
foods. Many foods, especially processed or multi-ingredient foods, cannot
readily be guaranteed GM-free. Food processors and vendors are not risk
takers. If they are uncertain, they'll label the product to avoid risking
prosecution or public reprisals, even at the risk of losing some sales.
Thus, "may contain GM" will become the default status. Who is going to
check that the food actually contains GM ingredients? No one. If there's
no penalty for this kind of false labeling, there will be no enforcement.

Under EU regulations, in order for a tomato sauce to avoid the GM label,
it must have documentation—a paper trail—to show it contains no GM
products. This additional burden affects everybody, but whom does the
paper trail hit the hardest? Certainly not the big corporations making
branded products; they are used to the intricacies of regulatory
compliance. Possibly not the large up-market retailers either, because
their customers have proved their willingness to pay inflated prices. It
will, however, affect the businesses of "generic" low-cost manufacturers.
It will also hit small producers supplying produce to local markets.

Where mandatory indiscriminant GM labeling is required, low-margin
businesses will institute labels proclaiming "may contain GM." In all
probability—though without certainty—the produce that arrives at local
markets does not contain GM ingredients. Yet, as an insurance policy, it
is among the most likely to be labeled. And consumers still will have no
idea whether the product contains GMOs.

Mandatory GM labels satisfy no one
Mandatory GM labeling regulations satisfy no one. Those wishing to avoid
all GM products, including "derived-from" foods, are annoyed at the
exemptions for products lacking DNA or protein. They are also cheated by
the tolerances allowing some GM content without triggering the required GM
label. Those wishing to avoid only certain GM products (such as
vegetarians) must eschew all such products, because the mandatory label
does not specify the source of the novel DNA. The trendsetters actively
seeking out GM foods might not, in spite of the GM label, be eating any.
Ordinary consumers who don't have strong feelings one way or another are
paying higher prices for the labeling bureaucracy, whether they buy foods
labeled GM or not. And the poor, who must buy at the bottom of the market
regardless of their personal opinions, pay a disproportionately higher
share of the increased costs to the benefit of no one, especially
themselves. No matter what your position, GM labels fail to provide their
intended raison d'etre—informed choice.