1

Agricultural BiotechnologyWhy Do Regulations in the European Union,

the United States, and Japan Differ?

Thomas Bernauer

Working Paper 5-2002

Center for International Studies (CIS)

ETH Zentrum, SEISeilergraben 498092 ZurichSwitzerlandTel +41-1-632-6774Fax +41-1-632-1946www.ib.ethz.ch

2

Introduction

The biotechnology policies of OECD countries are very similar in many ways. Most countriespromote their domestic biotech industry, for example through government funded researchand subsidies or tax breaks for start-ups. Regulation of the biotech sector is risk-based – ittargets what scientific risk-assessments identify as the highest risks to humans and theenvironment. And most biotech inventions are protected by patents or other forms of propertyrights protection.These similarities notwithstanding, European countries, the United States, and Japan havereacted very differently to the increasing use of biotechnology in agriculture. Europeancountries have imposed severe restrictions on this new technology, particularly in terms ofapproval and labeling of genetically modified organisms (GMOs). The United Statesmaintains a far more permissive approval policy and does not require labeling of GM-foods.Similar to the United States, Japan has approved many GMOs. But in contrast to the UnitedStates Japan has in early 2001 made labeling of GM-foods mandatory. In this paper, I explainthese differences in GMO regulations primarily in terms of variation in the collective actioncapacity of environmental and consumer as well as producer interest groups. I also assess thetrade implications of regulatory differences.

Controversy over agricultural biotechnology

Since the mid-1990s, crops, foods, and food ingredients developed through recombinant DNAtechnologies1 (commonly referred to as GM-seeds, GM-crops, or GM-foods2) have enteredthe markets of the United States, Japan, and several other industrialized countries on amassive scale.3 Controversies over whether and how to regulate this new technology haveconcentrated on the approval of GM-crops and GM-foods, and on the labeling of suchproducts.4

Many consumer and environmental groups have voiced strong concerns over the immediateand long-term effects of GMOs on human health. They have highlighted environmental risks,such as the reduction of biological diversity, out-crossing of superbugs, gene leakage, andagricultural sustainability of GM-crops. They have, moreover, raised broader social, ethical,religious, and economic issues associated with biotechnology.5 Most European consumershave opposed agricultural biotechnology ever since (in 1996) GMOs entered their market. Asa minimum, they have requested extensive labeling of GM-products, appealing to consumersovereignty and the right to chose. Japanese and US environmental and consumer groups havebecome more critical of GMOs only since the late 1990s.Among the strongest proponents of agricultural biotechnology have been export-orientedfarmers in the United States and agbiotech firms. Their economic stakes are substantial.6Worldwide, transgenic crops increased from 1.7 million hectares in 1996 to 44.2 millionhectares in 2000. In 2000, only four countries grew 99 percent of these crops: the UnitedStates (30.3m hectares), Argentina (10m hectares), Canada (3m hectares), and China (0.5mhectares).7 Agbiotech firms have invested heavily in research and development of transgeniccrops8 and have been determined to expand their 2.5 billion USD p.a. seed market for GM-soybeans, corn, cotton, and canola to cash in on these investments.9

3

Empirical and theoretical puzzles

Besides the evident political and economic importance of the issue, the comparison of GMOregulations in the EU, the United States, and Japan raises several empirical and theoreticalpuzzles.Particularly in the 1960s and 1970s, US environmental, health, and safety regulations havebeen strongly influenced by public risk perceptions and the precautionary principle. Examplesinclude the 1958 Delaney clause to the Food, Drug and Cosmetic Act, the Alar ban, and the1970 Clean Air Amendments, the Clean Water Act, the Endangered Species Act, the NationalEnvironmental Policy Act, and Superfund legislation.10 European consumers and regulators,in contrast, accepted many food products without much public debate, for example cheesemade from unpasteurized milk and countless food additives. Many of the environmental andpublic health concerns that moved high up on the public agenda in the United States and ledto stringent regulations, e.g. those over mercury, lead, asbestos, or cyclamates, have onlyrarely and/or with considerable delay made it into the headlines in Europe.In regard to GMOs, we observe the opposite tendency: stringent European regulations forGMOs are accompanied by widespread public concern over this new technology, whereasconsumer acceptance of GMOs is much higher in the United States and GMO regulations aremuch more permissive. The United States has, on average, adopted the strictest food labelingregulations in the world. Not so with regard to GMOs. Given that biotechnology may well bethe most rapidly commercialized food-production technology in the history of US agriculture,the absence of mandatory labeling is prima facie surprising. The EU has, on average andcompared to the United States, adopted less stringent food labeling legislation, but hasenacted much stricter labeling standards for GM-foods. How can we account for theseseemingly contradictory patterns of regulatory policy?The world’s agrochemical sector is highly concentrated, with only three firms accounting foralmost 50 percent of the global market (Syngenta, Aventis, Monsanto). GM-seeds, which in2000 made up around 10 percent of the world’s commercial seed market, are produced almostexclusively by these firms.11 Concentration is even more pronounced in the agbiotech market,where Monsanto (owned by Pharmacia & Upjohn since January 2000) has an 80 percentmarket share (Figure 1). In 1999, Monsanto’s seed traits accounted for 85 percent of the totalworld area covered with GM-crops.12

Figure 1: World Agbiotech Market Shares

Monsanto80%

Aventis7%

Syngenta5%

BASF5%

DuPont3%

Source: Wood Mackenzie, www.woodmac.com.The most widely accepted political economy explanation of regulation holds that concentratedinterests dominate over diffused interests, and that regulatory outcomes thus reflect the

4

preferences of concentrated interests.13 Yet in the EU, the concentrated and financiallypowerful agrochemical and agbiotech sector has been overwhelmed by a very heterogeneouscoalition of consumers, farmers, food processors, wholesalers, and retailers. A similarsituation has arisen in Japan, albeit with some delay. This development is particulary strikingbecause national and supranational regulators in Europe and Japan are widely assumed tomaintain close ties with the industries they regulate.14 In the United States, permissive GMOapproval regulations and the absence of mandatory labeling, despite substantial public supportfor stricter regulations15, are in conformity with producer preferences. How to explain thedeviation of the EU’s and Japan’s GMO policies from the outcome predicted by theconventional explanatory model?Finally, in the United States, environmental, health, and safety issues have often been highlypoliticized and controversial.16 Associated regulatory processes and their outcomes have beeninfluenced significantly by environmental and consumer interest groups, which enjoysubstantial access to information and government decision-making processes. In comparison,many observers note that regulatory processes in the EU and EU member countries are oftenshielded from NGOs, and that environmental and consumer groups have found it much harderto organize at the EU level than producer groups.17 Why have NGOs in Europe concentratedmuch more on GMOs than their American counterparts? Why have they been relativelysuccessful in achieving their prefered regulatory outcome?The following part of the paper outlines the theoretical argument that guides the comparisonof GMO regulations in the EU, the United States, and Japan. I then present the results of casestudies on GMO regulations in the three political units. In each case study I first describeregulatory outcomes. I then explain these outcomes by focusing on the collective actioncapacity of consumer and environmental as well as different types of producers groups.Finally, I compare the results of the three case studies and assess the implications ofregulatory diversity for international trade.

Theoretical Argument

The conventional political economy explanation of regulation holds that producer (industry)interests tend to dominate over consumer interests, and that regulations more often than notreflect the preferences of producers.18 Whenever consumers benefit from regulation (or theabsence of regulation, or deregulation), these benefits are usually coincidental.19

Our analysis of cross-national variation in GMO regulations focuses on explaining deviationsfrom the regulatory outcome predicted by the conventional model. Such deviation is mostpronounced in the European Union, followed by Japan. In contrast, regulatory outcomes inthe United States reflect the preferences of producers (permissive approval practices, nomandatory labeling). To explain deviations from the conventional model, I concentrate on thecollective action capacity of environmental/consumer and different types of producerinterests. The explanation thus consists of two elements:Collective action capacity of environmental and consumer groups. Due to their large andheterogeneous membership such groups are, in the conventional model, assumed to facedebilitating collective action problems. They find it harder than producers to mobilize supportand financial resources, both of which are necessary to effectively exert political pressure(e.g. lobbying of regulators) and market pressure (e.g. consumer boycotts). I argue that, dueto negative consumer perceptions of GMOs, regulatory failures in other food issues, and

5

institutional properties of the regulatory process, the collective action capacity of Europeanenvironmental and consumer groups opposed to GMOs has been higher than the capacity oftheir US and Japanese counterparts. As a result, European anti-GMO groups have been moresuccessful in influencing GMO regulations.Collective action capacity of producers. I show that the collective action capacity of pro-GMO producers has varied substantially across the EU, the United States, and Japan. Idistinguish upstream producers (agbiotech firms), farmers, and downstream producers(wholesalers, retailers, food processors) to show that differences in industrial structure and theextent of NGO opposition to GMOs account for variation in the collective action capacity ofpro-GMO producers. In the extreme case, preferences of different producer groups maydiverge so much that some producers join forces with anti-GMO environmental and consumergroups. In such cases, I expect stricter GMO approval and labeling regulations. This scenariohas largely materialized in the European Union, and to a smaller extent also in Japan. In theUnited States, cohesive and well-organized producer interests, comprising upstreamproducers, farmers, and downstream producers emerged very early in the regulatory processand have to date not encountered substantial opposition from anti-GMO environmental andconsumer groups.This explanatory framework and its application to agbiotech regulation contribute totheorizing on comparative social regulation in four ways. First, they go beyond the mostpopular explanation, which views regulatory differences as a direct result of variation inpublic attitudes.20 Second, they depart from explanations that concentrate on “unique” factors,such as political culture and regulatory style.21 Third, they focus on the analysis of causalhypotheses rather than the description of policy networks.22 Fourth, they provide anexplanation for empirical deviations from the conventional economic theory of regulation.23

Environmental and Consumer Group Influence

In this model of interest group politics I assume that variation in the stringency of regulationresults primarily from variation in the collective action capacity of consumer/environmentaland producer interest groups.24 In this subsection I concentrate on when and whyenvironmental or consumer groups are able to campaign effectively for more stringentregulations. The next subsection focuses on producer interests.The starting point of a theoretical answer to why Stigler’s (1971) producer dominanceargument appears to be inconsistent with the empirical evidence25 can be derived fromOlson’s ‘Logic of Collective Action’ (1965).26 Olson hypothesized that large (latent) groupsare difficult to mobilize. Without providing a specific answer he claimed that such groupscould only be mobilized through selective incentives of an economic or social type.27

I pick up at this point and assume that environmental and consumer interest groups are awareof their collective action problem and thus focus largely on issues that allow for maximummobilization of membership and financial resources. Issues that provoke public outrage aretop candidates. Public outrage is the fear or anger a risk induces in a relatively large part of acountry’s population. Its extent depends on specific characteristics of a given risk. Publicoutrage is stronger the more the risk in question is perceived to be involuntary, uncontrollable,or invisible, has a delayed or catastrophic effect, is memorable, very uncertain, poorlyunderstood, unfamiliar, unfairly distributed, and a technological hazard. Public outrage overrisks with these properties tends to be even stronger in cases where less risky alternativesexist.28

Public outrage increases environmental/consumer groups’ capacity to influence the regulatoryprocess through politics and the market. First, by redistributing political influence from

6

industry to environmental and consumer groups public outrage can contribute directly toregulatory outcomes that reflect consumer/environmental instead of industry preferences. Theunderlying logic is the following. Consumer and environmental organizations are pressuregroups that offer a public good, i.e. consumer or environmental protection. Public choicetheory tells us that the production of public goods is plagued by a free-rider problem: anyonecan benefit from a public good promoted by a pressure group without joining or in other waysactively supporting the group. In addition, consumer or environmental groups involve largenumbers of people, which increases the free-rider problem and the heterogeneity ofpreferences. Free-riding and heterogeneity of preferences reduce the organizational ability ofinterest groups. If organizational ability is weak, influence on regulatory processes will besmall.Focusing on issues with a high public outrage potential enables environmental/consumergroups to activate existing members, receive attention by the media, attract more members,mobilize non-members (latent supporters of a cause), and raise more funds, all of whichincreases political influence in the regulatory process.29 Once environmental/consumer groupshave overcome their collective action problem, their liability of large numbers turns into anasset: by virtue of their large membership, they can exert substantial political influence onvote-maximizing politicians and budget- or political support-maximizing regulators –consumers are also voters!The second effect of public outrage operates through the market. If public outrage is strongenough to significantly raise environmental or consumer groups’ collective action capacitythey can take more effective action against producers. They can, for example, boycott specificproducts, firms, or even entire industries.30 Or, short of boycotts, they can launch publiccampaigns that tarnish the image of firms, industries, and products. Through such action,environmental/consumer groups can bring about changes in producers’ preferences andbehavior in the market place, with evident implications for regulatory processes.Public outrage can vary not only across issues, but also across countries and over time.31 Thecase studies below show that variation in public outrage across the EU, the United States, andJapan accounts in part for variation in the collective action capacity of the respectiveenvironmental/consumer groups and their influence on the regulatory process.

Producer Influence

The conventional political economy model of regulation holds that producers are more likelyto support or tolerate stricter environmental or consumer protection standards if they canbenefit economically from the according regulation. Stricter regulation can yield several typesof benefits to producers.First, protectionist rents. Environmental or consumer protection will tend towards greaterstringency particularly in areas where regulation can be designed to shield import-competingdomestic firms from foreign competition. The assumption here is that producer demands forimport-restricting regulation are likely to attract more political support if justified in terms ofprotecting public health, rather than in terms of protecting domestic firms. The latter is moredifficult to justify because it transfers wealth from domestic consumers to domesticproducers.32 Situations in which import-competing firms lobby for more stringentenvironmental or consumer regulations that have the “side-effect” of limiting market accessby foreign producers can, in some cases, coincide with similar demands by environmental orconsumer groups. Such “strange bedfellow” coalitions tend to drive environmental andconsumer protection standards up.33

7

Second, domestic economic rents. Similar to the argument on protectionist rents, thisargument assumes that firms’ interests are shaped by industrial structure and that firms seek toimprove their competitive position through regulation. However, it differs from the aboveargument in several ways. In contrast to the classical regulatory capture argument34, it doesnot assume that there is a single industry with homogeneous interests within a given country.Individual firms or groups of firms within a specific industry may lobby for stricter or laxer,and for harmonized or particularistic regulations, depending on industrial structure andcompetitive position. For example, some large firms may lobby for stricter environmental orconsumer regulations that would be too costly for smaller firms to implement, while smallerfirms within the same industry and the same country oppose them35. In other words: whenexplaining regulatory outcomes in terms of producer interests we also need to take intoaccount different producer interests within the same issue area and the same country.Third, as proposed in the previous subsection, public outrage and associated campaigns byenvironmental and consumer groups can influence very directly interests on the producer side.Depending again on industrial structure and competitive position, individual firms or groupsof firms within a given industry may decide to give in to such pressure and support or toleratestricter environmental or consumer regulations, or they may not. For example, someproducers in a given industry may be more sensitive to consumer pressure than otherproducers, for example because they have a valuable brand to protect. Another example: inthe food industry, producers of seeds are less directly affected by consumer boycotts thanretailers or food processors. The latter are thus more likely to adjust to environmental andconsumer group demands and support stricter regulations.In summary, I expect stricter regulation in cases of public outrage and substantial political andmarket pressure by environmental and consumer groups on the one hand, and heterogeneousproducer interests on the other hand.36

Finally, when explaining regulatory outcomes in terms of interest group politics, institutionalstructures and the interests of regulators have to be taken into account. Institutional structuresmay vary in the extent to which they provide access to particular interest groups.37 They mayoffer priviledged access to some interest groups at the expense of others, thus affectingpolitical influence and ultimately regulatory outcomes.38 Moreover, regulatory agenciesfrequently have preferences of their own, as well as some room of maneuvre within whichthey can pursue their prefered policies.39 Consequently, the distribution of authority amongregulatory agencies in regard to a specific issue may have an impact on regulatory outcomes.

GMO Policy of the European Union

In this case study I first describe regulatory outcomes in the European Union. I then explainthe emergence of strict approval and labeling regulations in terms of the collective actioncapacity of consumer and producer interests.

Regulations

At the outset of the regulatory process in the mid-1980s, authorities in the EU, the UnitedStates, and Japan were divided over whether to restrict or promote biotechnology inagriculture, and over whether to regulate genetically modified organisms (GMOs)predominantly in terms of products or production processes.40 Regulating products implies

8

that there is nothing inherently unique in employing genetic modification techniques inagriculture, and that the resulting products are essentially the same as products obtainedthrough conventional agricultural methods. Regulating production processes, on the otherhand, implies that genetic modification poses unique risks and must, as such, be regulated.The latter approach has prevailed in the EU: the EU has adopted a variety of regulationsspecifically for GMOs, emphasizing the precautionary principle. As shown in the two othercase studies, the United States and Japan embraced the product-oriented approach.In 1990, the European Council adopted Directive 90/220 on the deliberate release ofgenetically modified organisms into the environment.41 While this directive was primarilydesigned to address environmental aspects of biotechnology42, it was, until the Novel FoodsRegulation was adopted in 1997 (see below), interpreted broadly to apply also to the approvalof GM-food products.43 Complicated decision-making procedures44 implied, however, thatimports of GM-seeds, field trials, commercial planting of GM-crops45, and commercializationof GM food and feed products were, de facto, subject to approval by each EU member state.46

The limitations of the 1990 directive became apparent when the Competent Authorities of theEU47 approved a British firm’s GM-canola (1995), a variety of Bt-corn produced by Novartis(1996), and Monsanto’s Roundup Ready soybeans (1996).48 Disputes among EU membercountries over the conditions for approval, stern NGO opposition, and changing views amongdownstream producers led to delays in the approval process, and to uneven implementation ofEU decisions. In particular, some EU countries imposed unilateral restrictions or bans on GM-products that had been approved by the EU.49

By 1997, the EU’s GMO approval policy was in disarray. Since April 1998, there has been ade facto moratorium on new approvals.50 The strongest supporters of this moratorium areDenmark, France, Greece, Italy, and Luxembourg. These countries have taken the positionthat they will not approve any new GMOs until strict labeling and tracing regulations areestablished. Belgium, Sweden, Austria, Portugal, the Netherlands, Finland, and Germany haveadopted a somewhat more moderate position. They have emphasized the precautionaryprinciple and have stated that no approvals of new GMOs should take place unless companiescan demonstrate that GMOs for which approval is sought do not have adverse effects for theenvironment or human health. Only the United Kingdom, the Netherlands, and the EUCommission have opposed a formal moratorium.51

Facing disarray in the approval process, strong opposition against a lifting of the de factomoratorium on further approvals, and the fact that several GM-products had, by 1997, alreadybeen approved by the EU for commercialization, the European Union engaged in furtherefforts to harmonize approval and labeling requirements for GMOs.In 1997 the European Parliament and the Council adopted Regulation 258/97, the NovelFoods Regulation. It applies to foods hitherto unknown in Europe, including also GM-foods,and installs approval procedures in which importers or producers have to show that the foodin question is not detrimental to human health. The regulation requires labeling if a geneticmodification can be detected with standard novel DNA or novel protein detection methods.No label is required for products that, in terms of their nutritional or allergenical properties,are „comparable“ to conventional products. Virtually all enzymes, vitamines, flavorings andother food additives were thus exempt from the Novel Foods Regulation as adopted in 1997.These exemptions were reduced only later on.52

The 1997 Novel Foods Regulation marks a sea change in the EU’s food policy. Thus far, allfoods had been allowed to enter the market without specific approval or even notification aslong as they were not incompatible with general or specific regulations (abuse principle). Thesafety of food products was, by assumption, derived from experience, and not from specificscientific risk assessment. In contrast, the new regulation imposed legal conditions not onlyon specific substances, but on entire foods. Before 1997, such conditions (in terms ofthresholds and risk assessment) had only applied to food additives. Since 1997, GM-foods

9

(and other „novel foods“) require individual approval before they can be brought on the EUmarket. Any novel food that is not explicitly approved is prohibited (prohibition principle).53

Regulation 258/97 as such was largely inoperable. It did not contain specifics onimplementation and it was left to individual member states to define thresholds, testingmethods, products subject to testing, and the content of labels. The regulation did not apply togranted or pending approvals.54 Bt-corn and Roundup Ready soybeans, which had beenapproved by the EU in 1996, were thus not covered by 258/97. As noted before, the latterRegulation also excluded enzymes, vitamines, flavorings, and other food additives (e.g.lecithin produced from GM soy). These gaps motivated some EU countries to introduce theirown labeling regulations, a development which risked inhibiting the free flow of goods in theCommunity.55

To stop this trend, the Commission and the Council issued several additional regulations.Regulations 1813/97 and 1139/98 require the labeling of food products containing GM-soybeans and GM-corn, but did not close the other gaps.56 In January 2000, the Commissiondefined a standard (49/2000) by which labeling is required if the product is at least 1 percentgenetically modified. For such products, the label „genetically modified“ is mandatory.Regulation 50/2000 extends the labeling requirement to food ingredients containing GM-additives and flavorings if GMOs can be detected.In February 2001, the EU Council and the European Parliament reached a compromise onrevisions to Directive 220/90. These revisions include time-limits on approval (maximum 10years), explicit procedures and schedules for each stage of the approval process, publicregisters detailing the locations of GMOs released for trial or commercialization, stricterguidelines for risk assessment, the gradual elimination of antibiotic resistance markers inGMOs, and calls for new and stricter legislation on traceability, liability, and labeling. Theseregulatory reforms, a large part of which are still at the stage of proposals and plans, are tocover the entire production chain of all foods and feeds produced with biotechnology, even ifno GMOs are detectable in the final product. Current proposals by the EU Commission for anew GM-food law, covering approval and labeling, suggest that, if anything, EU regulationsare bound to become stricter.57

Whether these regulatory reforms will weaken the blocking minority supporting themoratorium on approvals remains uncertain. Six countries – France, Italy, Denmark,Luxembourg, Austria, and Greece have stated that they will veto any approval applicationsuntil stricter labeling, tracing, and liability rules are in place. This could take several years.The currently applicable approval procedures hold that, if one or more member states opposean application, it is decided by qualified majority, which puts the six countries into a vetoposition.58

As a result of the de facto moratorium on new approvals (under Directive 220/90) and GM-foods (under the Novel Foods Regulation), the EU had, as of April 2001, approved a muchsmaller number of GM-products than the United States or Japan (see Table 1). 59

Table 1: GM-crops approved by the EU (as of April 2001)Product Year of approval Approval of plant Approval for food and/or

feedcornAgrEvo T25 (Aventis) 1998 Yes: planting, import,

processingYes: food and feed

Bt 176 corn (Ciba Geigy,today Novartis/Syngenta)

1997 Yes: planting, import,processing

Yes: food and feed

MON 810 (Monsanto) 1998 Yes: planting, import,processing

Yes: food and feed

MON 809 (Monsanto) 1998 No Yes: only for foodadditives

10

Bt 11 (Northrup King,now Novartis)

1998 (limited approval) Yes: import, processing,but not planting

Yes: only for foodadditives

soybeansRR soybeans (Monsanto,today Pharmacia)

1996 Yes: import, processing,but not planting

Yes: food and feed

canola (rapeseed)PGS MS1xRF 1 (Aventis) 1996 Yes: only seed production NoPGS MS1xRF 1, MS1xRF2 (two applications) (PlantGenetic Systems N.V.)

1997 Yes: planting, import,processing

Yes: refined oil

AgrEvo Topas 19/2(Aventis)

1998 Yes: only for import andprocessing, not forplanting

Yes: refined oil

AgrEvo Falcon GS 40/90(Aventis)

1999 No Yes: refined oil

Note: Other approved GM-products include: Bejo Zaden (RM 3-3; RM 3-4; RM 3-6) Chicorée/Radicchio forseed production, Seita/F tobacco, and several varieties of flowers. Zeneca’s TGT7-F tomato appeared to be closeto approval in 2001, but only for processed food. Plants were approved according to Directive 220/90. Forplanting, individual EU country permits are required in addition. Foods were approved according to the NovelFoods Regulation – usually through notification. Source: www.transgen.de.

As shown in Figure 2, also the number of field trials testifies to the much greater regulatoryrestrictions on GMOs in the European Union.

Figure 2: GMO field trials in 10 EU countries, the USA, and Japan, 1986 – 1998

0

500

1000

1500

2000

2500

EU10 USA Japan

num

ber

of G

MO

fie

ld t

rials

, 19

86 -

199

8

Source: OECD database

These figures for the EU mask substantial differences across EU member states. EU-wideregulations have, in essence, established floor-standards. No EU countries have permittedGMOs not approved by the EU. Neither have they permitted unlabeled GMOs subject tolabeling by EU regulations. However, some member states have moved beyond EU rules andhave adopted more restrictive approval and labeling policies. Some countries have evenbanned individual GM-products approved by the EU. The Commission and the more GMO-friendly EU countries have, so far, tolerated such deviations towards stricter standards.60

In summary, the European Union has moved from a situation of no regulation of agriculturalbiotechnology in the mid-1980s to very stringent regulations on the approval of GM-cropsand GM-foods, and to increasingly stringent and harmonized labeling requirements. As aresult, very few GM-crops and GM-foods have been approved for commercialization, thenumber of field trials has remained low, and virtually no GM-crops are commercially grown

11

in the EU. Moreover, despite (some would say, because of) strict labeling laws there areextremely few labeled GM-products on the European market, suggesting that (for reasonsexplained below) consumer demand for GM-products is almost non-existant.

Explanation

While some analysts had already in the early- to mid-1990s warned of growing opposition toagricultural biotechnology61, no one had anticipated the coming upheaval that would virtuallyshut down the EU’s market for GM-crops and GM-foods. Ex post, the strong trend towardsstricter GMO regulations in the EU can be explained in terms of an increasingly broad andinfluential anti-GMO coalition. This coalition includes environmental and consumer groups,most farmers, many downstream producers, and important parts of the EU’s regulatoryinstitutions.

Environmental and consumer interests

The theoretical argument holds that environmental/consumer group exert influence throughthe market, through the policy-making process, or both. Environmental/consumer groups caninfluence public opinion in a way that affects consumption behavior and thus ultimately alsoproducers, e.g. by launching campaigns or organizing boycotts against specific producers orindustries. Their influence can also operate through direct lobbying of policy-makers andvoter support for specific policy-makers and their policies (consumers are also voters!).In this subsection I show that European environmental/consumer groups have focused verystrongly on biotechnology issues ever since the mid-1980s, when the regulatory process in theEU began. Benefiting from latent public scepticism about technological innovation inagriculture, various food scandals particularly the 1990s, and low public trust in regulatoryagencies, they have been able to mobilize their membership and the wider public againstGMOs, and to launch major anti-GMO campaigns against producers.62 There is solid evidencethat these NGO activities have divided producers and have thus reduced the latters’ collectiveaction capacity, and that they have also influenced the decisions of regulators.Survey results show that a large part of the European public is deeply ambivalent aboutbiotechnology and GM-foods in particular (Table 2).

Table 2: Consumer acceptance of GM-foods in the EU, 1999Mostly agree Mostly

disagreeDo not know Maximum

Would buy GM-fruitif it tasted better

22% 66% 11% Mostly agree:Netherlands, 30%Mostly disagree:Greece, 88%

Would pay more fornon-GM food

53% 36% 11% Mostly agree:Greece, 83%Mostly disagree:UK, 51%

Willing to buycooking oilcontaining a littleGM-soy

22% 62% 16% Mostly agree:Netherlands, 37%Mostly disagree:Greece, 88%

Willing to eat eggsfrom hens fed onGM corn

19% 66% 15% Mostly agree:netherlands, 33%

12

GM corn Mostly disagree:Greece, 88%

Source: Eurobarometer 1999.

Attitudes appear to be driven by moral considerations, fears about unforeseen health andenvironmental risks, and the perception of GM-products as „unnatural“.63 Most Europeanconsumers view the GMO issue in health and safety terms rather than in terms of progress inscience and technology, and general views on GM-foods are predominantly negative (Figure3). On average, Europeans perceive somewhat greater risks in GM-foods than USrespondents.

Figure 3: Public perceptions of GM-foods, 1999

0

10

20

30

40

50

60

70

80

90

100

Awareness Negative view onGM-foods

Would not buyGM-food

No advantage inGM-foods

GM-foods posehealth and safety

problems

%

of

resp

onde

nts

Germany France United Kingdom USA Japan

Source: Angus Reid World Poll, 1999.64

When asked whether they would buy GM-foods European responses have been significantlymore negative than responses in North America (Figure 3).65 Comparisons of survey dataacross time show that consumer acceptance of biotechnology in the EU was lower than in theUnited States even before 199666, and declined substantially in the second half of the 1990s.67

The available data support the conventional wisdom among social scientists that consumersperceive risks very differently from scientists.68 Risk perceptions obviously vary acrosscountries, as shown in Table 3, whereas the large majority of biotechnology experts viewGM-foods as safe.

Table 3: Support and Opposition to GM crops and GM foodGM crops GM food

Greece -1 -2Norway -2 -2Austria -2 -2Luxembourg -2 -2Sweden -1 -2Denmark -1 -2France -1 -1Ireland -1 -1UK -1 -1Belgium 1 -1Germany 1 -1Italy 1 -1

13

Netherlands 1 -1Finland 1 1Portugal 2 -1Spain 2 1Note: Countries are ranked from most negative (-2) to most positive (+2). Source: Nature Biotechnology, Vol.18,Sept.2000:938, Eurobarometer, November 1999 survey.

Analysts have so far provided only sketchy answers to why public acceptance ofbiotechnology, and GM-foods in particular, varies across countries (and also time).Explanations range from arguments about food culture, eating habits, and agriculturaltraditions to arguments about environmental and consumer group influence on public option.69

Many of these explanations are neither theoretically nor empirically plausible70, others aremore convincing. One of the more compelling findings is that public perceptions ofauthorities regulating and actors using the technology in question can account for deviationsbetween scientific and public risk perceptions.71 Survey results indicate that people who donot trust that biotechnology is adequately regulated are more concerned about possible risksand perceive smaller benefits of the technology’s products. They also show that Europeanstrust regulatory agencies far less than North Americans.72 Similarly, while US consumers trustscientific associations and the FDA as sources of information on food safety issues,Europeans express far more trust in consumer and environmental groups.73

Low levels of public trust by Europeans in their regulatory authorities have been attributed toa variety of influences, most notably general features of the EU’s regulatory policy-making,specific food safety crises, and the extent of NGO activity and public debate.The EU does not have a scientific apparatus of its own. The Commission works with a widerange of scientific committees, whose members are government officials, university scientists,and scientists from the private sector. This socalled „comitology“ system has been widelycriticized as intransparent and open to political influence. Many analysts believe that thissystem is ultimately responsible for the severe mistakes in handling the BSE crisis andsubsequent food scandals.74 Public distrust in the EU’s and member states’ regulatory capacityreceived an additional boost by corruption scandals and other management failures in the EUCommission.75 The EU has engaged in a variety of reforms. These are reflected in the 1997New Approach to Consumer Health and Food Safety, the 1997 Green Paper on the GeneralPrinciples of Food Law in the EU, and the Commission’s 2000 White Paper on Food Safety.The aim of these reforms is to create a more transparent and science-based approach to risk-assessment, risk-management, and risk-communication. The proposed European FoodAuthority, once established, will have no regulatory powers, but will provide scientificadvice, collect and analyze information, and communicate with the public on issues of risk.76

Whether this new system will engender greater trust in the EU’s food regulation system,similar to the FDA in the United States, remains unclear.It is widely assumed, however, that the BSE crisis77 and other public health and safetyscandals (e.g. the dioxin scandal in Belgium in 1999, and HIV contaminated blood in Franceand elsewhere) have been primarily responsible for dramatic declines, in the second half ofthe 1990s, of public trust in regulatory authorities and the scientific expertise on which theyhave relied.78 These crises have also increased the receptiveness of the media to public healthand environmental issues and have contributed to more negative press coverage of GMOs,with obvious consequences of public perceptions.79

Campaigns by environmental/consumer groups appear to have contributed to increasingpublic awareness of biotechnology issues. They also seem to have reinforced the „right toknow“ justification for labeling, which tends to thrive in situations of widespread publicconcerns over food safety and low trust in regulators.80 The same applies to negative publicperceptions of GM-foods. However, empirical demonstration of these effects, which areconventional wisdom in industry circles, is very difficult. The fact that European

14

environmental/consumer groups have focused on GMO issues much earlier and muchstronger than their North American counterparts, and that surveys data show strongtransatlantic differences in the perception of GMOs at the same time, does not demonstrate acausal relationship.81

While it is difficult to pin down the influence of European environmental/consumer groups onpublic perceptions of GM-foods, it is more evident that public adversion against GM-foodsand public health and safety crises in Europe in the second half of the 1990s have helpedNGOs in mobilizing their members and other (latent) supporters of their cause. NGOs havethus been able to shape regulatory policies and their outcome through market strategies(pressure on producers) and lobbying of policy-makers. The latter is discussed in thissubsection. Market pressure is examined in the subsection on producers.Early interest in biotechnology issues notwithstanding, GMOs rose to the top ofenvironmental/consumer groups’ agendas only in the mid-1990s.82 NGO campaigns weresparked particularly by the EU’s authorization of Roundup Ready soybeans in May 1996 andBt-corn in December 1996. The EU’s regulatory authorities evidently misjudged the extent ofopposition when they approved these two GM-crops without having put in place an effectivelabeling requirement and post-marketing controls (e.g. tracing and liability regulations). Thesame applies to upstream and downstream producers.Greenpeace and Friends of the Earth Europe (FOE), the largest environmental interest groupsin Europe, made GMOs one of their top priorities and have ever since invested heavily incampaigns. The same applies to the European Bureau of Consumers’ Unions (BEUC), aBrussels-based federation of independent national consumer organizations from all EU andother European countries. While BEUC has lobbied mainly for stricter labeling of GM-foods,Greenpeace and FOE have taken stronger anti-GMO positions. All three groups, as well asmany other environmental/consumer groups in Europe83 have persistently and sharplyexposed and criticized inadequacies in the EU’s GMO regulations and their implementation.84

NGOs were not at the negotiating table when GMO regulations were designed and adopted.But one can spot their influence at several key points in the regulatory process. In the secondhalf of the 1980s, the EU’s Directorate General for Environment, Consumer Protection, andNuclear Safety (DG XI) emerged as the leading agency for the regulation of GMOs. DG XI,whose „constituency“ includes environmental/consumer NGOs, was mainly responsible forestablishing the process-orientation of the EU’s GMO policy, which is reflected in Directive90/220. DG XI’s role as „chef de file“ for that directive and also for the Novel FoodsRegulation adopted in 1997 opened an important door for environmental/consumer NGOinfluence on the EU’s regulatory policy, not least because most regulatory activity is initiatedby the Commission.85 Historical accounts of the regulatory process86 have shown that theleading role of DG XI and the influence of NGOs were in large measure responsible formoving the EU’s GMO policy from a system of monitored self-regulation towards process-oriented regulations grounded in the precautionary principle and following the model ofDanish and German policies.87

The opportunities for environmental/consumer groups to influence the EU’s regulatory policyincreased further when the 1986 Single European Act, the 1993 Maastricht treaty, and the1997 Amsterdam Agreement strengthened the European Parliament (EP). The EP, eager toascertain de facto its formally expanding authority, has been particularly open to NGOinfluence.88 Motivated in large part by NGO lobbying, the EP has repeatedly and successfullychallenged proposals by the European Commission and the Council of Ministers that wouldhave led to a relaxation of approval and labeling regulations.89 Similarly, the NGO-led publicoutcry over the Commission’s approval of GM-soy and GM-corn in 1996 and its initialdecision not to require labeling of these products strengthened the position of DG XI withinthe Commission.90 The decision not to require labeling was consistent with the Novel FoodsRegulation, as it was then under discussion.91 Strong opposition by environmental/consumer

15

groups combined with resistance on the part of downstream producers (see below) enabledDG XI to make a technical adaptation to Directive 90/220 in June 1997. This adaptation ineffect overturned the Novel Foods Regulation and required mandatory labeling for all GM-foods.NGO campaigns have influenced the regulatory process also in more indirect but importantways by contributing to a „ratcheting up“ effect. In 1986-1998, for example, they motivatedsome EU countries to introduce mandatory labeling for GM-soy and GM-corn in the absenceof EU-wide legislation.92 They have also contributed to preventing the marketing of EUapproved GMOs in several EU countries.93 Such differences in regulatory responses byindividual EU countries have been perceived by the Commission and other EU bodies asconstraints on the free movement of goods in the EU’s internal market. Like in other EUpolicy areas, EU-wide harmonization of regulations has been the obvious response to thisproblem. Member countries with more stringent regulations have, not least because of strongNGO and general domestic consumer opposition, not been willing to relax their policies. Evenin cases where such regulations are incompatible with EU-legislation, the Commission hasbeen very reluctant to challenge them, for example throught the European Court of Justice.The only feasible solution has thus been harmonization at levels of stringency supported bymore GMO-adverse countries.

Producers

Persistently negative public perceptions of GMOs, concerted anti-GMO campaigns by NGOs,and growing regulatory restrictions on GMOs imposed by the EU and individual memberstates have driven most farm interest groups and downstream producers into supporting morestringent GMO regulations. Agbiotech firms have found it impossible to stop this trend.Upstream Producers. Concerned about growing cross-national differences in regulation, theagrochemical, pharmaceutical, and food industries in the EU argued already in the mid-1980sthat there was a need for EU-wide regulation of biotechnology.94 Surprisingly though, suchstatements were, in the early phases of the policy process, not accompanied by more concreteinitiatives.Greenwood/Ronit and Patterson provide the following explanations.95 First, European firmswere accustomed to organizing along product but not production process lines. Thus they didinitially not regard themselves as biotech firms with a common interest. Second, the largenumber of medium or small firms that had only recently entered the biotech field, combinedwith substantial economic competition among these firms, made collective action difficult.Third, the industry was concerned that organizing along process lines would invite regulatorseven more to install process-oriented regulations (horizontal regulations in the EU’sterminology). Fourth, expectations differed among small or medium firms and largemultinational firms as to the need for harmonized regulations, and the cost implications ofregulation. Finally, the industry overestimated the influence of DG III, DG VI, and DG XII,which opposed process-oriented (horizontal) regulation of GMOs and had previouslysucceeded in excluding a wide range of products from drafts of Directive 220/90. Only inJune 1989, when agbiotech firms finally realized that the tide was moving against them, didthey create a Senior Advisory Group for Biotechnology (SAGB).96

Most authors thus claim that the unwillingness or inability of upstream producers to organizeearly enough is responsibly for the industry’s inability to prevent the EU’s 1990 decision tofocus on process-oriented GMO regulation. Upstream producers had viewed GMO regulationlargely as a means to facilitate the EU’s internal market. Directive 90/220, however, has putthe EU on a trajectory where GMO-regulations have in effect installed floor standards forenvironmental and consumer protection. As exemplied by the very slow reforms of Directive220/90, the path dependency of the EU’s complex regulatory system has, ever since,

16

prevented a return to the regulatory concept of the early to mid-1980s.97 The latter concepthad focused more strongly on science based and monitored self-regulation. The outcome ofefforts to reform Directive 90/220 (see above) show, however, that attempts by upstreamproducers to obtain a relaxation of approval and labeling rules and return to the status quoante have been entirely unsuccessful.The above explanation focusing on the timing of the industry’s political organization accountsrather well for why upstream producers have not been able to prevent Directive 90/220. Butthe path dependency argument explains only in very vague terms why upstream producershave never been able to reverse the regulatory trajectory initiated in 1990. The theoreticalargument outlined above provides the missing link.For a variety of reasons, the biotech industry has developed much slower in Europe than inthe United States and has concentrated on medical rather than agricultural applications.Another major difference between Europe and the United States is the small number of largefirms in Europe.98 The conventional economic theory of regulation suggests that this marketconcentration in Europe should bolster upstream producers’ collective action capacity andtranslate into successful lobbying for less restrictive GMO regulations. However, one mightalso expect large European biotech firms to lobby for GMO regulations that protect theircompetitive position – notably against US producers. The evidence disconfirms bothpropositions.First, persistently negative public perceptions of GMOs and campaigns by NGOs99 haveneutralized virtually every attempt by European biotech firms and EuropaBio, the EuropeanAssociation of Bioindustries, to lobby for a relaxation of approval regulations and to preventthe introduction of mandatory labeling.100 Activists from Greenpeace and FOE, as well asmany smaller groups, have engaged in a large number of media-effective anti-GMO activities,for example uprooting of GM-crop fields, attempts to prevent the unloading of ships with USGM-crops, and specific campaigns against leading agbiotech firms.101

NGO campaigns against large agbiotech firms have been particularly effective in inflictingreputation costs on upstream producers, ultimately the source of all GMOs. These campaignshave turned large agbiotech firms into a highly visible target onto which the wider public hasbeen able to project GMO-related fears, but also more diffuse fears about globalization,technological innvovation, and US dominance over Europe. With slogans such as “Monsantois trying to force-feed Frankensteinfood to Europeans”, activist groups have portraid GMOsas a risk imposed on Europeans by an American firm. The United States has inadvertendlyhelped such campaigns by trying to pressure the EU into lifting its ban on growth hormonesused in beef production.102 Similar to the growth hormones case, US pressure on the EU torelax its restrictions on GMOs has been widely perceived in Europe as an effort to impose, forparocial commercial reasons and against the will of consumers, a new and risky agriculturaltechnology. Social science research on risk suggests that people are more willing to acceptrisks imposed by nature, but are likely to be more upset when forced to take risksinvoluntarily.103.Monsanto, Novartis, Rhone-Poulenc, AgrEvo/Aventis, Zeneca, and other agbiotech firmshave found it extremely difficult to defend against this onslaught.104 Most European biotechfirms as well as EuropaBio eventually changed their position and have reluctantly supportedmandatory labeling.105 Permissive approval regulations and no mandatory labeling would havebeen their preferred outcome. But in view of strong consumer concerns over GMOs, sternopposition by NGO, and increasingly heterogeneous regulatory responses by individual EUcountries, stricter and harmonized labeling regulations appeared to be the only way ofprotecting agbiotech interests.As to the second proposition, economic theories of regulation suggest that stricter GMOregulations in Europe might reflect protectionist rent-seeking by the European biotechindustry. The available evidence does not support this hypothesis. European agbiotech

17

producers have persistently lobbied for laxer approval and labeling regulations and havespoken in favor of mandatory labeling only once they have had no other options. Moreover, itis difficult to see how European upstream producers could derive economic rents from stricterGMO regulations. They have focused more on medical biotech applications. Even if they hadengaged much more in agricultural biotechnology, it is unclear whether and how the EU couldhave designed GMO regulations benefiting the EU’s agbiotech industry at the expense offoreign competitors. Protectionist regulations would not have helped the EU’s agbiotechindustry with its key problems, namely public resistance against GM-products, what industryviews as overly restrictive regulations, and a shortage of venture capital.Agricultural Producers. Assuming large economies of scale in GM-crop farming manyanalysts predicted fierce opposition by EU farm interests to the introduction of GMOs.106

From this assumption, they have jumped to the conclusion that EU restrictions on GMOs arethe result of an „unholy alliance“ between protectionist farm interests andenvironmental/consumer groups. This argument is not supported by the evidence.Farm interest groups in the EU, most notably the European Farmers’ Coordination, have untilthe late 1990s repeatedly expressed concerns about GM-crops, but have not actively lobbiedregulators to impose restrictions. The absence of substantial opposition by key European farminterest groups was largely due to the fact that there was no direct economic competitionbetween imported GM-crops and European farm produce.Imports of soybeans and corn from the United States have, since the mid-1990s, declined, inlarge part due to regulatory restrictions. The US share in the EU’s corn imports dropped from86 percent in 1995 to below 10 percent in 2000, resulting in economic losses to US producersof more than 200 million USD per year. US soybean exports to the EU fell from around 2.6billion USD to less than one billion. However, imports of these crops from other non-EUcountries as well as total EU imports of soybeans and, to a lesser extent, also corn haveincreased at the same time.107 The EU is largely self-sufficient in corn, whereas EUproduction of soybeans amounts to less than 10 percent of domestic consumption. Becausemost of the corn and soy consumed in the EU is used for animal feed, premiums for importednon-GM-soy and, to a much lesser extent, corn may even have imposed costs on EU farmers.In other words, the EU’s GMO-regulations have, de facto, not improved the competitiveposition of EU farmers vis-a-vis their US counterparts.108

More active support by EU farm interest groups since the late 1990s for more restrictiveGMO regulations stems primarily from general concerns about disturbances in European foodmarkets.109 These problems have resulted from the BSE and other food safety crises110, butalso from more general problems associated with food surpluses and massive restructuringand concentration processes in European agriculture. In this context, support for stringentGMO-regulations, and in some cases even explicit support by farmers for anti-GMOcampaigns by NGOs, must been viewed as part of an overall effort to reduce consumerconcerns and stabilize markets by means of stricterer food safety regulations across theboard.111 Anti-GMO views among EU farmers have also been fueled by Monsanto’s strategyof market entry – notably its purchase of many seed companies and the introduction of the“terminator” gene – and the imposition by the US of trade measures of around 100 millionUSD per year against EU agricultural products in the trade conflict over growth hormones.Increasingly negative views on GMOs among farmers in the EU have been reinforced bycases in several EU countries where fields were accidentally planted with GMO-contaminatedcrops.112 NGOs and farmers have accused governments of being incapable of ascertaining thatimported seeds are GMO-free. More generally, and similar to the BSE crisis, these cases havereinforced the view among consumers and also farmers that governments have lost control infood safety matters. In most of these cases, GMO-contaminated crops were destroyed andfarmers were compensated by the respective government, an implicit acknowledgement oftheir problems in controlling seed imports.

18

Although opposition to GMOs by EU farmers has so far clearly not been motivated byprotectionist rent-seeking, some analysts anticipate that this motivation will become strongeronce other GM-products with direct competitive effects on EU farmers enter the world marketon a larger scale (e.g. wheat, rice, fruit, fish, meat). European farmers thus clearly seem to beset on a trajectory towards greater opposition to GMOs. If the competitiveness issue will infuture dominate over the issue of consumer confidence, stricter labeling regulations will notsolve the problem for EU farmers. In that case, farmers will lobby more strongly for anoutright ban on GM-crops and GM-foods.113

Downstream Producers. Food processors, wholesalers, and retailers in the EU initially tried toresist demands for mandatory labeling. Since 1996, however, they have adjusted swiftly to thedemands of environmental/consumer groups.114 While most downstream producers have, inprinciple, not opposed GM-products, they have given in to market pressure and have accepteddemands for segregation, tracing, and labeling.115 However, their strategy of differentiatingproducts along GMO/non-GMO lines and stopping the anti-GMO trend half way by acceptinglabeling has failed. Under fierce pressure from NGOs labeled GM-foods have graduallydisappeared from the market, whereas thousands of products containing GMOs in forms orquantities not subject to mandatory labeling remain on the EU market. Due to their lowcollective action capacity (very large number of firms with diverse interests) downstreamproducers have not been able to organize effectively and stop the anti-GMO trend at the pointof mandatory labeling. By mid-1999, almost all major food retailers and food-manufacturingcompanies had announced that their products would be GM-free.116

The first to adjust were large food processors worried about their valuable brands. In 1997,Nestlé, a Swiss company that was initially opposed to GM-food labeling, began to label itsGM-products in Europe without being legally required to do so.117 Unilever, a company basedin the Netherlands and the United Kingdom, that used around 7.5 percent of all soybeansimportend into the EU, did the same in 1998.118 At that point in time, Nestlé was the world’slargest food processor, Unilever the third largest.119 Both firms justified labeling withconsumers’ right to know but emphasized that GM-foods did not pose any health risks. Manyother food processors have followed these examples since.Similar to what we observed for upstream producers, changes in the positions of foodprocessors have also been prompted by concerns over increasing differences in the regulatorypolicies of individual EU countries. These differences threatened to reduce the economies ofscale advantage of large food processors. Concerns of this nature have also been expressed bythe Confederation of Food and Drink Industries (CIAA), the most important association of thefood and drink industry in the EU.120 Food processors have thus come to prefer stricter butuniform EU labeling and also approval regulations over permissive EU regulations butincreasing heterogeneity of regulations among EU countries. Because EU-wide harmonizationhas remained slow and incomplete, large food processors have moved ahead on their own. InAugust 2000, for example, Novartis, which in October 2000 spun off its agricultural business(now Syngenta), declared that it would no longer use GM-ingredients in any of its foodproducts worldwide.121

Wholesalers and retailers have followed the same trajectory for the same reasons.122 Underpressure from NGOs and consumers they have embraced mandatory labeling requirementsand, in many cases, went further and withdrew GM-products altogether.123 In response toproblems with EU harmonization efforts they have also issued joint standards to reduceuncertainty among consumers and producers, establish a level playing field, and reducetransaction costs in cross-border trade.124

From 1999 on, a fairly consistent pattern of interaction between NGOs and downstreamproducers emerged. Downstream producers labeled a variety of GM-foods in conformity withdeclarations by individual firms, joint standards issued by producer associations, and finallyEU regulations issued between May 1997 and April 2000. Examples are tomato sauce made

19

from Zeneca tomatoes and Nestlé’s „butterfinger“ (a peanutbutter-cornflakes candy bar).Market analysts note that consumers were apparently not deterred by these labels. However,NGOs then singled out labeled products, such as Nestlé’s „butterfinger“, for massive publiccampaigns. In response, retailers and wholesalers withdrew these products and producersfurther upstream reduced or halted their production of these foods. In other words,downstream producers’ support for labeling produced a „Catch-22“ situation: if they labeled aGM-product, they were attacked by NGOs and were forced to withdraw the product; if theydid not label a GM-product, they risked NGO campaigns plus legal action for violation of EUor national regulations. This pattern of interaction has resulted in a rapidly declining numberof labeled GM-products on the EU market.125 In other words, consumer demand for GM-foodswould probably not have collapsed without mandatory labeling.126

Pressure on downstream producers has grown further due to the increasing amount of testingfor undeclared GMOs by NGOs and public authorities that has gone hand in hand withincreasingly precise and decreasingly costly testing methods.127 Whenever GMOs have beendiscoved in food products, even if below the EU’s threshold, retailers have, in most cases,quickly withdrawn these products, particularly if they had previously declared that they wouldstop selling foods containing GMOs. While the costs of EU regulations had so far fallenprimarily on producers outside the EU (US agricultural exporters, a few biotech firmsproducing GM-seeds), the cost implications of stricter GMO regulations for downstreamproducers in the EU have begun to rise.These costs vary across GM-products. The most severe problems for downstream producersin the EU have resulted from soybean imports. For climatic reasons, the EU produces onlyaround one percent of the world supply of soybeans and imports around 14 million tons ofsoybeans annually, as well as more than 10 million tons of shredded soybeans and other soyproducts. The EU’s own production amounts to a total of 1.6 million tons of soybeans,produced mainly in France and Italy. The largest part of the EU’s imports comes from theUnited States, Argentina, and Brasil, the first two of which are major producers of GM-soybeans (Figure 4).128

Figure 4: World production of soybeans; EU imports of soybeans; 1999

USA50%

Brasi l21%

Argentina9%

China8%

EU1%

Other countries

11% USA37%

Brasil41%

Argentina22%

Source: Transgen, January 25, 2001, www.transgen.de; Töpfer International. 1999.Statistische Informationen zum Getreide- und Futtermittelmarkt. Hamburg. AmericanSoybean Association (www.asa.org).

Downstream producers have encountered much smaller problems with regard to GM-corn.Certification that only the four GM-corn varieties approved by the EU are present129, or that,in the case of EU countries that have banned Bt corn, no Bt-corn is present, is difficult andcostly. The EU’s corn imports from the US have thus decreased sharply.130 However, there is

20

an abundant supply of non-GM corn from within the EU and other countries. Buyers havetherefore not had to pay significant premia for GM-free corn.131

Despite the mentioned problems with soybean imports it is very unlikely that downstreamproducers will reverse their position and seek to re-establish a market for GM-foods in theEU.132 Their principal problem now are enzymes, vitamines, and food additives producedthrough biotechnology, which are contained in thousands of food-products on the EU market,usually in forms and quantitives currently not subject to formal approval and mandatorylabeling.133 If all these products had to be formally approved and labeled under EUregulations, the food supply in the EU would possibly come to a halt. The strategy used bydownstream producers since 1999, to remove GM-foods from the market rather than labelthem, would not work in many of these cases or be too costly. Not surprisingly then,downstream producers who have declared their products GMO-free have tried to restrict theimplementation of this commitment to soy- and corn-ingredients subject to mandatorylabeling under EU regulations. Governments, EU-authorities, and NGOs, for their part, haveso far refrained from coercing downstream producers into accepting approval and labeling ofall GMOs in food products and all products produced with GMOs.

Conclusion

In the mid-1980s, the biotechnology policy of the European Union was very similar to that ofother OECD countries. Since then, the EU has, largely on the basis of process-orientedregulations, established much stricter GMO regulations than the United States and Japan.Starting in 1990, the EU has adopted and implemented a series of increasingly stringentapproval and labeling regulations. As a result, very few GMOs have been approved by the EUfor marketing as food or feed, even fewer for commercial planting. The number of field trialsis far lower in the EU than in the United States. The number of labeled GM-food products onthe market has declined to almost zero, suggesting that the market for GM-foods in Europehas shrunk to (an also declining) market for GM-food-ingredients not subject to labeling.The theoretical framework outlined above helps in accounting for this trend. Benefiting fromgrowing consumer concerns over agricultural biotechnology, food safety problems in the EU,and low public trust in regulatory authorities, European environmental/consumer groups wereable to effectively mobilize their supporters and the wider public. They were thus able tolaunch powerful campaigns against upstream and downstream producers. These campaignshave also transformed the GMO issue into one of „highy saliency“ and (in the eyes of thewider public) „low complexity“ – which usually produces „hearing room“ patterns of policy-making.134 These hearing room patterns, combined with the multi-level nature of EU policy-making, has opened the door for NGO influence. NGOs have thus been able to influencedirectly not only the behavior of producers (market pressure), but also regulatory decision-making, predominantly via the European Parliament and the Commission’s DG XI.Public concerns over GM-foods in combination with NGO campaigns have dividedproducers, thus reducing their collective action capacity in ways not anticipated byconventional economic theories of regulation. Large upstream producers, such as Monsantoand Novartis, became prime targes of NGO action and were never able to put up effectiveresistance.Farmers, concerned primarily about major disturbances in European food markets,increasingly turned against GMOs. Their opposition is likely to grow further if GM-productswith more direct negative effects for their competitive position appear on the EU’s market.

21

Because labeling can not solve their competitiveness problem, European farmers are likely tobecome increasingly strong advocates of a general ban on GMOs.Downstream producers were motivated by public risk adversion and NGO pressure toembrace labeling. This policy change has had unintended effects, exposing GM-foods to evenmore pressure by NGOs and forcing downstream producers to desert the market for labeledGM-foods. The cost implications of this trend for downstream producers vary across GM-products. While substitutes for GM-corn have been easy to find, premiums for non-GM soyhave risen because world markets of soy are supplied by very few countries, most of whichare major producers of GM-soy. Nonetheless, given their low collective action capacity (largenumber of firms with very diverse interests), strong NGO opposition and public riskadversion, and opposition by farmers, downstream producers are unlikely to reverse theirposition and try and re-establish a market for GM-foods.135

The combination of low public acceptance of GM-foods, pressure by NGOs, growingopposition against GMOs among farmers, strong incentives of downstream producers towithdraw from the market for labeled GM-foods, and institutional inertia in EU policy-making make a reversal of the trend against GMOs in the European Union very unlikely. Tothe contrary, mandatory labeling is a solution for those consumer concerns that center oninformation asymmetry, but not a solution for consumer concerns centered on risk shielding.Moreover, mandatory labeling does not solve problems of agricultural excess capacity andinferior competitiveness of European farmers on a global scale. Mandatory labeling may thusbe only an intermediary step on the way to complete closure of the EU’s market to GMOs.

GMO Policy of the United States

Regulations

US regulatory policy on GMOs has been built on the assumption that biotechnology impliesinnovation in how food is produced, but does not lead to substantially different properties offood products themselves, compared to what would be obtained with conventional methods.Following this product-orientation, the US Food and Drug Administration (FDA) and the USDepartment of Agriculture (USDA) have approved most industry requests for field testing andcommercialization of GM-seeds and GM-foods. Around 50 GMOs have been approved forcommercialization. GM-crop acreage in the United States has increased dramatically between1996 and 2001. In 2001, one third of US all corn crops and more than half of all US soybeancrops were produced from GM-seeds. Around 60 percent of all food in US grocery storescontained GMOs.136 The FDA does not require labeling of GM-foods. Proposals by the USEnvironmental Protection Agency (EPA) to establish GMO-related process regulations forenvironmental protection purposes have not been adopted.In the early to mid-1980s different parts of the US government, in particular EPA, USDA, andFDA, struggled for regulatory authority in biotechnology policy. Similar to Europe, they weredived over whether to restrict or promote this new technology, and whether to adopt aproduct- or process-oriented approach to regulation. The White House Office of Science andTechnology Policy (OSTP), USDA, and FDA preferred the promotion of biotechnology and aproduct-approach, whereas EPA advocated the development of new risk assessmentprocedures for GMOs and process-based regulation.137

22

In 1984, responding to increasing criticism, notably by industry, of lacking coordination inbiotech policy, the Executive Office of the US President created a Cabinet Council WorkingGroup with representatives from fifteen agencies. By taking this step, the Reaganadministration prevented a takeover of regulatory authority by the EPA138, more activeinvolvement of the US Congress, and more public scrutiny of the regulatory process ingeneral.139

In 1986, this working group issued a Coordinated Framework for the Regulation ofBiotechnology. To date, all biotech products in the United States have been regulated in thisframework. The latter set up the Biotechnology Science Coordinating Committee (BSCC),assigned the primary responsibility for regulation to EPA, USDA, and FDA, and establishedprinciples for coordination and cooperation among these authorities.140

Transportation, growing (including field testing), and propagation of GM-crops are governedby the Animal and Plant Health Inspection Service (APHIS) of the USDA under the FederalPlant Pest Act. If a GM-plant is not intended for human consumption and is not modified tocontain a pesticide the USDA is the primary agency involved in regulating it. It issues defacto commercial permits stating that the plant concerned is not a pest and needs no furtherregulation. At the same time it carries out an environmental assessment of the plant under theNational Environmental Policy Act.Pest-resistant GM-crops fall under the authority of the EPA under the Federal Insecticide,Fungicide and Rodenticide Act and the Toxic Substances Control Act. They are subject to arather strict testing regime, where producers must submit testing data to the EPA. These dataare publicly available for a 30 day notice and comment period before approval.The FDA has regulatory authority for GM-foods, food additives, processing aids, and biotechmedical products. It also has primary responsibility for labeling, with the exception of meatand poultry products (governed by the USDA). Most analysts note that of the three agenciesresponsible for US biotech policy FDA has had the greatest influence because most biotechproducts on the US market have been health care or food products.141 The FDA’s approvalprocedure is based on voluntary consultations with producers, the submission of summaries oftests, and no publicly available information.The Coordinated Framework firmly established the product-orientation of US biotech policyand denied the necessity of new legislation specifically for GMOs. It locked in the prevailingassumption that commercial biotechnology should not be viewed as posing special risks forhuman health and the environment. This view was supported and legitimized by a 1989 reportof the National Research Council.142 Broader social or political implications of biotechnologywere not addressed.143 In the United States, GM-products are thus regulated on the basis ofpreexisting legislation.Both the USDA and the FDA have, throughout the 1990s, relaxed their GMO approvalpolicies. In 1993, the APHIS introduced a notification (instead of permit) procedure for sixtransgenic crops (corn, soybean, cotton, tomato, potato, and tobacco).144 In 1997, it extendedthis notification policy to the majority of GM-crop varieties used in the United States andgreatly simplified notification procedures for imports, release into the environment, andtransportation of GMOs.145 It has, moreover, removed an increasing number of GM-plantsfrom its oversight.Already in 1990, the FDA approved the first food-ingredient produced through biotechnology,chymosin, an agent that helps milk clot in making cheese. This was followed by rennin, whichis also used to make cheese, and recombinant BST, used to increase milk output in cattle. In1994, in accordance with its 1992 interpretation of the Federal Food, Drug, and CosmeticAct146, FDA installed a simplified approval procedure for GM-foods, starting with Calgene’sFLAVR SAVRTM tomato.147

Under this procedure, applications for approval of GM-foods do not have to undergo acomprehensive scientific review because the production process as such is, in general terms,

23

deemed to pose no specific risk. GM-foods are thus considered to be „generally recognized assafe“ (GRAS) unless, in the judgement of the manufacturer, there is a reason for concern.Foods considered GRAS are not suject to pre-market safety testing.148 In essence, this systemallows biotech firms to decide on their own whether a GM-product is safe. The FDA is onlyconsulted.This strong product-orientation is also reflected in US policy on labeling. GM-foods aresubject to the same labeling requirements that are applied by the FDA to all foods on the USmarket.149 Labeling of GM-foods is only mandatory if a particular GM-food is no longersubstantially equivalent to the corresponding conventional food in terms of composition,nutrition, or safety.150 In 1992, the FDA ruled that if any of the most common allergens (milk,eggs, wheat, fish, crustacea, tree nuts and legumes, especially peanuts and soybeans) areadded to food via biotechnology, the producer must either provide scientific evidence that theallergen is not present in the new food or label the product. Such products are subject toadditional safety testing. The label does not have to indicate that the food was produced withbiotechnology. It must only state that a potential allergen has been added to the food.151

Labeling is also required if a food product is transformed through biotechnology so that itsnutrient content does not conform to what is normally expected.152 The vast majority of GM-foods on the US market thus do not require labeling, and no GM food product is labeled assuch.153 The FDA has full legal and regulatory authority to remove food from the market evenafter approval if it is considered to be unsafe or incorrectly labeled. Its authority is immediateand final.The EPA has favored a more process-oriented approach to GMO regulation than the FDA andthe USDA.154 In late 1994, it proposed a set of regulations for GM-plants designed to resistpests, treating such plants as if they were pesticides. Due to opposition from other parts of theUS government, industry, and scientists, the proposed rules were not adopted.155 In 1999,more intense public debate and NGO involvement in GMO issues was triggered by a researchreport suggesting that a particular variety of Bt-corn caused harm not only to pests (notablythe corn borer), but also Monarch butterfly larvae.156 In response, the EPA has reexaminedwhether to subject certain GM-crops to pesticide regulations. In January 2000, it askedfarmers using Bt-corn to plant buffer zones with conventional corn to protect the Monarchbutterfly.157 Continuing opposition has thus far prevented the establishment of process-oriented GMO regulations beyond such ad hoc measures.In parallel to increased public attention to the environmental consequences of GMOs a moreintense public debate on labeling of GM-foods has developed. Opinion polls suggest thataround 80 to 90 percent of US respondents favor mandatory labeling of GM-foods.158 InNovember 1999, 49 members of Congress sent a letter to the FDA requesting mandatorylabeling on the basis of the Food Drug and Cosmetic Act.159 In March 2000, more than 50 USconsumer groups filed a legal petition against the FDA calling for a moratorium on GMOuntil stricter approval procedures were installed.160 Two bills on labeling of GM-foods, andtwo bills on safety testing were introduced in the US Senate and House in 1999 and 2000.161

Moreover, NGOs have initiated a variety of lawsuits162 to bolster requests for mandatorylabeling and have also spurred legislative activity at the state-level.The FDA has, to date, remained largely immune to this pressure. In January 2001, itreaffirmed its product-oriented approach and submitted two proposals for strengthening itsregulatory oversight.163 First, it proposed to make hitherto voluntary consultation betweenGMO-producers and the FDA in the approval process mandatory, thus reactivating a similarproposal made in 1993 but dropped due to opposition from scientific and professionalgroups.164 This new measure would not, however, substantially change the FDA’s permissiveapproval policy: all GM-foods or feeds marketed in the United States have already gonethrough the voluntary consultation program installed in 1994.165 Second, the FDA proposed

24

guidelines for voluntary labeling of GM-/non-GM foods, similar to the policy currentlyapplied to organic and kosher foods.166

In summary, US regulatory policy on GMOs has focused almost exclusively on products. Ithas operated within pre-existing legislation and generally assumes that genetic engineeringper se does not pose a particular risk that requires specific regulation of this productionmethod. Approval of new GM-crops and GM-foods by the USDA and the FDA respectivelyhas been relatively swift and uncomplicated. The EPA has, on several occasions, proposedprocess-oriented regulations for pest-resistant GM-crops, but these proposals have not beenadopted. The majority of US consumers appear to favor mandatory labeling of GM-foods.NGOs have taken legal action in pursuit of this cause, and two bills on labeling have beenintroduced in Congress. So far, the FDA has not given in to this public pressure. It hasmaintained its permissive approval and no-labeling policy.

Explanation

In this section I account for the permissive GMO approval regulations and the absence ofmandatory labeling of GM-foods in the United States in terms of the theoretical argumentoutlined above. Regulatory laxity in the United States has been installed and sustained by abroad and cohesive pro-GMO coalition. This coalition comprises the large majority ofupstream and downstream producers as well as farmers. Anti-GMO lobbying on the part ofenvironmental and consumer groups has remained relatively weak.

Environmental and Consumer Groups

US environmental/consumer groups have engaged in GMO-issues much later than theirEuropean counterparts. Only since the late 1990s have they lobbied on a larger scale formandatory labeling of GM-foods and stricter environmental and health risk assessments in theapproval process.167 This time-lag is surprising, notably because it is hard to explain in termsof lesser openness of the US regulatory system. The US approval process, reaching fromscientific discovery to marketing of a GM-product, provides for at least five publicintervention points.168 It allows for just as much access by NGOs as the approval process inthe EU.In comparison to Europe, the proportion of US NGOs not opposed to GMOs in principle ismuch larger. Consumers Union, for example, the largest US consumer interest group, hasadvocated mandatory labeling but has not fundamentally questioned the safety of agriculturalbiotechnology as such.169 The same holds for the Environmental Defense Fund (EDF) and theUnion of Concerned Scientists (UCS), both of which are among the few NGOs active inbiotech issues already since the mid-1980s. Again, both are not per se opposed to agriculturalbiotechnology.170 They have concentrated on criticizing the FDA’s labeling policy and havelobbied for mandatory labeling.171 The Center for Science in the Public Interest, anotherimportant US consumer NGO, has been largely supportive of GMOs.172 In contrast to Europe,most NGO activity on GMOs in the United States has consisted of lobbying of policy-makerson the basis of scientific arguments, and lawsuits against regulatory agencies, such as theFDA, and individual biotech firms. Fundamental opposition to GMOs has remained the realmof smaller and more radical NGOs.173 Compared to Europe, more „hands-on“ NGO activity,such as demonstrations or destruction of GM-crops, has remained relatively rare.

25

In light of the evidence discussed in this section, it appears that substantial public support forGMOs and the broad and cohesive pro-GMO coalition comprising producers and governmentagencies have made US environmental/consumer groups very reluctant to campaign againstGMOs. While the interaction effects of public risk perceptions and NGO campaigns aredifficult to assess174, the evidence suggests that NGOs are willing to invest heavily in anti-GMO campaigns only when they perceive a substantial public outrage potential or whenpublic adversion is already high.As shown in Figures 3 (above) and 5 (below), US consumers view GM-foods more favorablythan European consumers.175 Public support for all applications of biotechnology hasremained at rather high levels over time.176

Figure 5: Public Acceptance of GM-foods in the United States

0

10

20

30

40

50

60

70

80

90

1995 1996 1997 Feb 99 Oct 99 Jan 00 May 00 Oct 00

% w

illin

g to

buy

Source: Hoban 2000. Consumers were asked whether they would buy GM potatoes ortomatoes protected from insect damage and requiring fewer pesticides.

The most evident difference between EU countries and the United States, which helps inaccounting for cross-national differences in perceptions, is the degree of consumer trust inbiotech information sources.177 US consumers trust the American Medical Association, theFDA, university scientists, registered dietitians, and farmers most, whereas Europeans putgreater trust in consumer/environmental organizations and the medical profession.178 The trustof US consumers in regulators and scientists has been sustained in the absence of major foodsafety scandals in the 1990s, in stark contrast to the EU where BSE and other food safetycrises have dealt a huge blow to the credibility of regulators and scientists. Even the recentcontroversy over Starlink corn (see below) has not significantly affected public support foragricultural biotechnology.179

As noted above, US NGOs have focused predominantly on the labeling issue, while surveyresults suggest that US consumers are somewhat ambivalent in regard to mandatorylabeling.180 Social scientists have not yet been able to show whether NGO activity hasincreased public demand for labeling, or whether NGOs demands for labeling have increaseddue to greater public demand. However, both developments combined have, in the late 1990s,begun to influence regulatory policy and the behavior of producers. I discuss the former atthis point and examine the latter in the subsection on producers.Biotech policy in the United States has, until very recently, been shaped almost entirely by thescientific community, industry, and the three US government agencies responsible forregulation (USDA, FDA, EPA). NGOs have, for example, not been substantially involved inthe development of the FDA’s approval and labeling policy for GMOs. While NGOs began tolobby for mandatory labeling and stricter environmental risk assessments already in the late

26

1990s, NGO activity has been boosted by the StarLink problem, which emerged in September2000.Because StarLink corn contains a proteine (Cry9C) that my cause allergic reactions it hadbeen approved by the US EPA only for animal feed, but not human consumption. InSeptember 2000, consumer groups detected traces of StarLink corn in foods. StarLink wassubsequently found in over three hundred brand-name products. The ensuing controversyexposed weaknesses in the US regulatory system, in particular the inability of producers toeffectively prevent cross-pollination and segregate crops in the US grain handling system.181 Italso exposed as ineffective the EPA’s 1999 request to farmers to plant non-GMO buffer cropsaround GM-crops.182 Recalling products, testing activity, buybacks, and other measures havecost Aventis, farmers, and downstream producers millions of dollars.183

The StarLink controversy, for the first time, opened the door for a stronger challenge byNGOs of US regulatory policy on GMOs. A wide range of environmental/consumer NGOs,for example Greenpeace, Friends of the Earth, the Organic Consumers Association, theGenetically Engineered Food Alert, the Pesticide Action Network, Sustain, the Center forFood Safety, and many others have seized the opportunity.184 Most of these groups haveseverely criticized the FDA’s January 2001 proposal on notification and voluntary labeling.They have requested more rigorous safety and toxicological testing for all food additives andan end to what they have called the FDA’s „no labeling, not safety testing“ policy.185 Insupport of their criticism, NGOs have come up with evidence that, in several cases, includingthe FDA’s 1994 decision on the FLAVR SAVRTM tomato, controversial scientific evidencewas ignored.186 Similarly, NGOs have criticized the inadequacy of EPA-mandated measuresagainst cross-pollination of GM-crops as well as the EPA’s environmental risk assessmentpractices for GMOs.It appears unlikely, however, that these campaigns will be more effective in bringing about afundamental change in regulatory policy than earlier efforts by environmental/consumergroups.187 As suggested by its 2001 proposal, the FDA has stuck firmly to its permissiveapproval and product-oriented labeling policy. Aware that public opinion favors the labelingof GM-foods, it has sought to restrict labeling to a voluntary, market driven process, and tomake sure that it works orderly. As long as public support for agricultural biotechnology andpublic trust in regulatory authorities do not decline dramatically it is unlikely that thissituation will change.NGOs have thus concentrated on regulatory restrictions on GMOs at the state level, and onmarketplace pressure.188 I seems likely that the future of GM-crops and GM-food in theUnited States will be decided in the market rather than through the regulatory process. Theoutcome will ultimately depend on downstream producer and farmer reactions to consumerpressure and, in the case of farmers, also productivity gains that can be achieved throughagricultural biotechnology.

Producers

In contrast to Europe, producers in the United States have been able to form and maintain acoherent and well-organized pro-GMO coalition, including upstream and downstreamproducers and export-oriented farmers. Big agbiotech firms have been the leaders in thiscoalition. The latter has cooperated closely with regulators in what can be termed a"boardroom“ structure189 of biotech policy making.190 Only since 1999 has this system becomesomewhat more open to influence by the US Congress and environmental/consumer groups.

27

Upstream ProducersSpurred by large-scale government funding, a well functioning venture capital market, andstrong cooperation between industry and universities the US biotech industry has developedinto the world’s largest.191 As noted in the introduction, the US firm Monsanto controlsaround 80 percent of the world market for GM-seeds. To recuperate its enormous R&Dinvestments, the US biotech industry needs biotech-friendly regulations.The large majority of upstream producers in the United States are organized in a singleassociation, the Biotechnology Industry Organization (BIO).192 In addition, large biotech firmssuch as AgrEvo, DuPont, and Pharmacia have also individually become involved in theregulatory process.193 BIO and its members have persistently supported permissive approvalregulations and have strongly opposed mandatory labeling194, thus backing the FDA’sregulatory policy.195

Homogenous preferences of its membership, substantial scientific expertise, and strongfinancial and other support from large biotech firms (e.g. AgrEvo, Monsanto, Pioneer Hi-Bred, Zeneca Plant Sciences, Calgene, Novartis) have equipped BIO with a large collectiveaction capacity.196 In addition, BIO maintains a close relationship with the FDA and other USregulatory agencies. A large part of BIO’s staff consists of former government officials. Manyobservers have also pointed to a „revolving door“ policy in which employees move back andforth between the FDA, the EPA, and biotech companies.197 Industry views of biotechnology,notably in regard to economic competitiveness and pressure for fast commercialization torecoup R&D costs, have thus quickly found their way to US biotech regulators.198 To enhanceconsumer acceptance of GM-products seven large biotech companies and BIO formed theCouncil for Biotechnology Information. They plan to invest around 50 million USD per year,and possibly up to 250 million USD over three to five years for building public support forbiotech foods.199

The situation just described can, in Gormley’s terms, by characterized as „boardroom“ policy-making.200 This pattern of policy-making is the most susceptible to regulatory capture byindustry. Indeed, Henry Miller, a former FDA official, stated that „in this area, the USgovernment agencies have done exactly what big agribusiness has asked them to do and toldthem to do“.201

Growing public opposition to GM-foods in the United States since the late 1990s has, thusfar, affected the cohesion of upstream producers only at the margin. Most notably, it has (inaddition to a variety of other factors) encouraged several large upstream producers (e.g.Novartis, AstraZeneca, Pharmacia, Aventis) to divest the agbiotech parts of their firms. It has,therefore, also contributed to a divergence of interests between large firms engaged inagricultural biotechnology and those, often smaller, firms engaged in medical biotechnology –with the latter blaming the former (notably Monsanto and Aventis) for negatively affectingthe overall biotechnology market.202 Moreover, the StarLink controversy has, for the firsttime, introduced a rift in the coalition of export-oriented farmers, and downstream andupstream producers. The former two have launched a series of class-action lawsuits againstAventis, requesting compensation for losses. As the scandal broke, Japan cut its US cornpurchases by 50%, Korea, the second largest US corn export market, banned imports of UScorn entirely.203

FarmersUS farmers have adopted the new technology at an almost breathtaking pace. The first GM-crops were planted in 1995/96. By the year 2000, the percent of acreage planted with GM-crops had grown to around 20 percent for corn and more than 50 percent for soybeans (Table4). For 2001, the USDA predicts further growth for soybeans (from 54 to 63 percent) andcotton (from 61 to 64 percent), but not for GM-corn.204

28

Table 4: Percent of US acreage planted with GM-crops1995 1996 1997 1998 1999 2000

Bt corn 1 6 18 26 19Bt cotton 12 18 23 32 39Bt potatoes 1 2.5 <4 <4 2-3RR cotton 4 21 37 54RR soybeans 2 13 37 47 54BXN cotton 0.1 0.1 1.2 5.8 7.8 7.2Source: Carpenter/Gianessi, National Center for Food and Agricultural Policy, 2001:3.

The crops shown in Table 5 are genetically modified to be more resistant to insects and/orherbicides. In principle, the benefits of GM-crops include increased yields, lower productioncosts and easier management, and reduced pesticide use.205.

The extent of benefits associated with this technology is still contested, however, and seemsto vary substantially across crop varieties. For example, Carpenter/Gianessi (2001) estimateaggregate gains in net revenues in 1999 of 35 million USD for Bt corn, 99m USD for Btcotton, and 216m USD for RR soybeans - rather modest profits in the overall picture. Asurvey of GM-crop performance by the USDA (Table 5 shows the data for cotton and

Table 5: Effects of Adopting Herbicide-tolerant and Insect-Resistant Field Crops

(a) Small increases or decreases are less than 1 percent change for a 10 percent change in adoption(b) Increases or decreases are less than 5 percent change for a 10 percent change in adoption(c) Coefficients are not statistically different from zero

Herbicide-tolerantsoybean 1997

Herbicide-tolerantcotton 1997 Bt cotton 1997

Change in yields (a) small increase (b) increase (b) increase

Change in profits 0 (b) increase (b) increase

Change in pesticide use

Acetamide herbicides 0

Triazine herbicides 0

Other synthetics (b) decrease 0

Glyphosate (b) increase 0

Change in insecticide use

organophosphates 0

Pyrethoid insecticides 0

other (b) decrease

Source: Fernandez-Cornejo, Klotz-Ingram & Jans (1999), Farm-level Effects of Adopting GeneticallyEngineered Crops in the US, Paper presented at International Conference “Transitions in Agritech: Economicsof Strategy and Policy, NE-165, Washington, DC, June 24-25, 1999, as cited in: USDA, Economic ResearchService, Value-Enhanced Crops: Technology’s Next Stage, Agricultural Outlook, March 1999

29

soybeans), on the other hand, indicated that yields from biotech crops were not consistentlyhigher. The USDA data suggest some gains for cotton, both in yields and profits – thoughprofits do not seem to be a function of reduced pesticide and insecticide use, which weresupposed to be the main benefits. GM-soybeans, on the other hand, were not associated withincreased yields or reduced pesticide/insecticide use. These conclusions hold also whencontrolling for differences in soil, weather, and regional farming practices. Foster (2001)concludes that “from the perspective of the farmer, it is not at all clear that the current furorover GMO food products is something that poses a critical problem apart from the issues ofdealing with foodstuffs already in the pipeline. Cotton is not a food stuff and the use ofGMOs in this crop reduces to the still-important issues relating to possible genetic effects onother crops resulting from cross-pollination with other plants.” 206

The available data suggest, however, that middle to large-size industrialized farms haveembraced biotechnology more rapidly than other farms.207 I thus expect the former to favorpermissive approval regulations, oppose mandatory labeling208, and support US trade policiesthat open world markets to GM-products. Smaller family farms, for which GM-crops havebeen less beneficial, would benefit from regulations that impose costs on larger producers. Ithus expect owners of smaller farms to favor stricter approval regulations and mandatorylabeling. This variation in preferences, driven by variation in industrial structure, should bemost pronounced in the case of corn and soybeans, which account for the largest share ofGM-crops. I now review the positions of major US farmers groups and commodityassociations to assess this proposition.The American Farm Bureau Federation (AFBF), the largest US farm organization, isdominated by large agricultural producers. It has fully supported the FDA’s approval andlabeling policy.209 The National Family Farm Coalition (NFFC), a network of grassrootsorganizations working on family farm issues, opposes the vertical integration of agricultureand corporate agriculture more generally. It has criticized US approval policy for GMOs andhas requested that “consumers in the US and around the globe have the right to know whethertheir food is genetically engineered”.210

The American Soybean Association (ASA) and the National Corn Growers Association(NCGA) represent export-oriented producers with a large stake in GM-crops.211 Becauseproducers of GM-soybeans and GM-corn are more export-dependent than other members ofAFBF, ASA and NCGA have adopted even stronger pro-GMO and anti-GMO labelingpositions. Predictably also, concerns of ASA over problems of consumer acceptance andregulatory restrictions in other countries have been stronger than concerns of NCGA: US soyproducers are more dependant on export markets than US corn producers: in 1996 37 percentof US produced soybeans or soybean products were exported, compared to 19 percent of USproduced corn.212 Facing the inevitable, the ASA and the NCGA have been negotiating withthe EU and EU member states separately in order to preserve US exports to Europe througheffective segregation and labeling. Nonetheless, both associations remain strong supporters ofpermissive approval policy and oppose mandatory labeling in the United States.213 The USGrains Council, which develops export markets for US crops, has adopted very similarpositions.214

The American Corn Growers Association (ACGA) represents smaller family farms. Incontrast to the ASA and NCGA, it has recommended stricter approval policies and mandatorylabeling of all GM-foods, both domestic and foreign „to fulfill the consumer’s right to knowwhat foods they and their children eat”.215 Pointing to decreasing demand for GM-foods andGM-feeds since 1999/2000, uncertainty over the costs of segregation, and liability problemswith GM-crops it has encouraged farmers to opt for non-GM crops.216 Predictions by the

30

USDA for 2001 show, however, that only GM-corn acreage is likely to stagnate, whileacreage of other GM-crops continues to grow.Farm size and export-orientation thus account rather well for positions with regard to approvaland labeling policy. Interest groups representing predominantly larger and export-orientedfarmers oppose stricter approval policy and mandatory labeling. Groups representing smallerfarmers oriented towards the domestic US market have not fundamentally opposed GM-foods, but have lobbied for stricter approval regulations and mandatory labeling of GM-foods.Diverging interests due to differences in industrial structure, but also the fragmentation of theUS agricultural association system more generally, have limited the collective action capacityof farmers. More than 200 interest groups are involved in agricultural policy issues. Theyspecialize in different issues (general commodity associations, single commodity associations,and specialized associations), and their general policies are marked by substantial ideologicaldifferences.217 US GMO regulations have been congruent with the preferences of groupsrepresenting larger and export-oriented farm interests largely because these groups have beenable to align themselves with biotech firms and food industry associations.218 Smaller farmersproducing for the domestic market and favoring stricter approval regulations and mandatorylabeling have only recently gained in influence in alliance with environmental and consumergroups.Calculations of benefits from GM-crops, such as those shown in Table 5 above, ignore theinfluence of consumer acceptance, which has had a growing impact on the rate of adoption ofGM-crops by US farmers. Consumer opposition to GMOs in the EU, Japan, and othercountries has increased, and so has the stringency of approval and labeling regulations outsidethe United States. As a result, demand for non-GM corn and non-GM soybeans has grown andUS soy and corn exports to countries with stricter regulations have declined.219 In addition tothis general trend, the StarLink controversy has led to a collapse of major export markets forUS corn, such as Japan, Korea, and Europe. Slowing or even declining world demand forGM-crops has contributed to a leveling out of global GM-crop acreage since 1999/2000.220

Due to high costs of segregation and identity preservation, export-oriented US farmers haveresponded only very slowly to this change in public risk perceptions.221

Whether US farmers will, in the medium to long term, engage or desert the market for GM-crops depends largely on the behavior of downstream producers. It also depends on whetherUS farmers continue to believe that GM-crops, and GM-soybeans in particular, create anarrow margin of profitability that keeps them afloat. There is growing evidence (see Foster2001) that this belief may be questionable. GM-soybeans are the key issue in this regard,because they are the most important GM-crop exported to Europe. Foster, for example, notesthat, apart from questions of economic benefits from GM-soybeans and trade with Europe,GM-soybean production in the United States has been supported by massive increases insubsidies. He claims that in the year 2000 38 percent of net receipts by US soybean farmerswere government subsidies. Estimates for 2001 are between 28 and 34 percent. In otherwords, US production of GM-soybeans may have grown not because of greater profitabilityor exports, but because of subsidies.222

Downstream ProducersUntil the late 1990s, US downstream producers encountered very little consumer oppositionto GM-foods, which had appeared on the US market on a massive scale since the mid-1990s.They have not benefited much from the first generation of GMOs, which involves modifiedinput rather than output traits. However, large downstream producers in particular have

31

expected greater benefits from the second generation of GMOs, which involves modifiedoutput rather than input traits (e.g. enhanced nutritional qualities)223.The vast majority of large downstream producers have thus supported the FDA’s approvaland labeling policy, arguing that labeling would stigmatize their products and brands, and thatit would be costly to implement. Grocery Manufacturers of America (GMA) and The NationalFood Processors Association (NFPA), the two largest associations of the US food processingindustry, have been the most influential representatives of large downstream producerinterests. GMA represents primarily processors of branded foods, NFPA represents firms thatprocess and package fruits, vegetables, meat, fish, specialty food, and beverage products.224

In contrast, many smaller, specialized food processors and retailers, notably those in theorganic food business, have supported stricter approval and labeling regulations. In theirlobbying efforts, they have frequently associated themselves with consumer NGOs, forexample, Consumers Union, the Center for Food Safety, or the Alliance for Bio-IntegrityIn response to growing consumer concerns in the United States over GM-foods GMA andNFPA have, since the late 1990s, sought to reverse that trend through intensified lobbying ofpolicy-makers and public campaigns.225 Pro-GM-food manufacturers and retailers have alsofounded the Alliance for Better Foods (ABF), which so far has invested several dozen millionUSD in public relations activities and election campaign contributions.226 These downstreamproducers, as well as upstream producers, have learned from Europe and Japan that labelingcan have enormous effects on consumer behavior. This assumption is supported by a May2001 survey by the Center for Science in the Public Interest, a consumer NGO. This surveyshows that, while the intensity of consumer preferences for mandatory labeling is rather low,consumers would still hesitate to buy GM food if it were so labeled.227

These efforts by downstream producers have been weakened by defections from the pro-GMO coalition of several large and many smaller US food processors, retailers, andwholesalers. In spring 1999, for example, Greenpeace sent letters to major baby foodcompanies, asking if they had taken any steps to ensure that they were not using GM-ingredients. In response, Gerber and Heinz, two large baby food producers, switched to non-GM-ingredients.228 Firms such as McDonalds, McCain Foods, Frito-Lay, IAMS pet foods,Whole Foods Market, Wild Oats Markets, Seagram, and a range of others have also reducedor eliminated GM-foods from their US business in response to NGO campaigns.229 TheStarLink problem has led to another wave of defections. Archer Daniels Midland, forexample, which purchases around 30 percent of all corn, sobeans, and wheat in the UnitedStates announced that it expected US farmers and grain elevators to start separating andsegregating GM- and non-GM crops. In mid-to-late 2000, large firms, such as ConAgra,ADM, and Cargill have initiated measures to segregate GM-crops from non-GM crops.230

Some analysts have thus regarded the StarLink crisis as the ‚beginning of the end’ of GM-crops.231

The evidence is too ambiguous to justify such a conclusion. On the one hand, a cautiousinterpretation of survey data (see above) suggests that US consumers do, in principle, valuemore choice and the right to know what is in their food. But they do not appear to befundamentally unhappy with the current system that requires labeling in exceptional cases andallows for voluntary labeling. Public trust in regulatory agencies and public support for GM-foods do not seem to have suffered substantially from the StarLink controversy. Consumerviews in the United States are, in comparison to Europe, still far away from a situation ofpublic outrage.232 In addition, US downstream producers which have moved towards non-GMfoods and/or segregation and identity preservation have discovered the technical difficultiesand high costs of doing so. Hence they are unlikely to cave in easily to somewhat increasedconsumer demand for non-GM foods. On the other hand, the fact that several downstreamproducers have turned to non-GM products and others, such as ConAgra, one of the world’slargest food companies, are preparing to segregate products, suggests that the US food

32

industry has become increasingly worried about a possible consumer backlash, similar to theone in Europe.233 In the extreme case fears of NGO campaigns and public adversion mighteven lead to a self-fulfilling prophecy – that is, major changes in firms’ GMO policy even inthe absence of a fundamental shift of public option against GM-foods.

Conclusion

Permissive approval practices and the absence of mandatory labeling in the United States aresufficiently explained by the key variables outlined in the theoretical argument above:collective action capacity of consumers and producers. Low public concern about GM-foodsand high public trust in regulatory authorities and scientists have made it very difficult forNGOs to mobilize their members and other latent supporters on GMO issues. Most USenvironmental/consumer groups have become active on GMO issues much later than theirEuropean counterparts. Moreover, most US NGOs have not fundamentally opposed GMOs,but have requested stricter approval regulations and mandatory labeling. They have reliedmostly on the lobbying of policy-makers and legal action against regulatory agencies andagbiotech firms, rather than the more activist strategies favored by many their Europeancounterparts.Agricultural biotechnology has been supported by a relatively homogeneous, well-organized,and well-funded coalition of export-oriented farmers and their commodity associations, theupstream industry, and downstream producers. Producers have, moreover, enjoyedpriviledged access to US regulatory agencies, particularly the FDA.Growing public concern over GMOs since the late 1990s, and the StarLink controversy inparticular, have produced several faultlines in the pro-GMO coalition. First, several largebiotech firms have divested their agbiotech branches and interests between firms engaged inmedical and firms engaged in agricultural biotechnology have begun to diverge. Second,some large and many small downstream producers have moved towards GMO-free productsand labeling, weaking the pro-GMO campaigns of downstream producer groups and dividingdownstream producers as a whole. Third, the StarLink crisis has led to conflicts betweenfarmers and upstream producers over damages and their compensation.Whether these faultlines will eventually lead to a disintegration of the producer coalitionremains an open question. If they do, and if such disintegration coincides with a completeclosure of European and Asian markets to US GM-crops, most downstream producers andfarmers in the United States may decide to desert the market for GM-crops. Technicaldifficulties and costs of segregating and preserving the identity of non-GM-crops throughoutthe food chain and labeling products may, under the circumstances mentioned, contribute tothis tendency.

GMO Policy of Japan

Until spring 2001, Japan’s regulatory policy on GMOs was very similar to US policy. Itaddressed GMOs under pre-existing product-oriented legislation. GMO approval regulationswere permissive and labeling of GM-foods was not mandatory. In contrast to the UnitedStates, however, no GM-crops have been grown in Japan. In the late 1990s, public concernsover GMOs began to rise and new regulations that provide for stricter conditions for approval

33

and mandatory labeling were adopted. These measures are much laxer than those of the EU,putting Japan inbetween the United States and the EU in terms of the overall stringency ofGMO regulations.Regulatory outcomes in Japan can, again, be explained in terms of the collective actioncapacity of consumers and producers. Until the late-1990s Japan’s regulatory system forGMOs corresponded to the „board-room“ pattern.234 In this system, industry and governmentbureaucracies controlled the regulatory process. A 1997 decision by the Japanese governmentto involve a range of private actors in the regulatory process opened the latter toenvironmental/consumer group influence. Thus environmental/consumer groups alreadyenjoyed substantial access to the regulatory process when, in the late 1990s, they decided tobecome involved in GMO issues on a larger scale. Growing public concerns about GMOs andparticularly the spillover of the StarLink controversy from the United States to Japan havehelped Japanese environmental/consumer groups in mobilizing their supporters and exertingpressure on regulators and producers.Consumer pressure has divided downstream producers and has driven many of them intosupporting stricter approval regulations and mandatory labeling. The change of position onthe part of downstream producers has been facilitated by the fact that the new regulations donot cover 70 – 80 percent of the soy and corn in Japan’s food chain, and that the new approvalregulations are little more than a formalization of already existing voluntary practices.Upstream producers, who have favored the US-model of GMO regulation, have not engagedin larger-scale efforts to prevent stricter policies because their economic stakes in agriculturalbiotechnology are small. In contrast to what many analysts expected, Japanese farmers havenot joined the anti-GMO camp. They could not achieve economic rents through stricter GMOregulation. Moreover, unlike in Europe, agricultural markets in Japan have not suffered frommajor food safety crises. Thus farmers did not view stricter GMO-regulations as aninstrument for reducing general consumer fears.

Regulations

Biotech regulation in Japan started comparatively late. First guidelines for rDNA experimentswere issued in 1979. In 1986, the Ministry of International Trade and Industry (MITI) adoptedguidelines for industrial applications of rDNA technology. Later in the same year, theMinistry of Health and Welfare (MHW) adopted guidelines for biotech pharmaceuticals. In1989, Japan’s Environment Agency and its Ministry of Agriculture, Forestry and Fisheries(MAFF) followed with policy statements and guidelines in areas under their respectivejurisdictions. Like in the United States, this activity was based on existing legislation fordrugs, worker health and safety, agriculture, and environmental protection.235 Several authorssuggest that much of this regulatory activity was motivated primarily by individual agencies’concerns about their turf.236

In contrast to the United States, no government agency has been empowered to coordinate thedifferent ministries involved in biotech regulation. As a result, Japanese GMO regulations aremarked by overlapping regulatory activity. The Ministry of Education, Science and Cultureauthorizes biotech research in universities. The Science and Technology Agency hasessentially the same function for companies and other research institutions. Depending on theproduct, separate guidelines issued by the MAFF, the MHW, and the MITI apply toexperimentation. The MHW and the MAFF are responsible for approving GM-foods.237 In1996, when Japan began to import GM-products, the MHW required, on a voluntary basis,proofs from suppliers that the imported GMOs met the ministry’s safety standards. Similar tothe US FDA, it argued that mandatory labeling of GM-foods was not necessary because thesefoods were not significantly different from their conventionally produced counterparts.

34

Because of its permissive GMO policy, and because a large share of its agricultural importscome from the United States, the largest producer of GM-crops, Japan became the world’largest importer of GM-crops and GM-foods. In 1999, around 80 percent of soybeansconsumed in Japan, and almost 90 percent of corn were imported from the United States. Byspring 2001, Japan had approved 29 varieties of seven genetically altered crops for import andsale under its food safety guidelines, including soybeans, corn, rapeseed, potatoes, cotton,sugar beet, and tomatoes. No official data on imports of GM-products have been published. Itis estimated that in 1999 around 30 percent of Japan’s 4.8 million tonnes of soybean importsand around 16 million tonnes of its corn imports were genetically modified.238

In August 1999, the MAFF announced a major change in its GMO policy. In early August, agovernment advisory panel, established by the MAFF in 1997, had proposed that Japanintroduce mandatory labeling of GMOs. On August 10, 1999, the MAFF adopted the advisorypanel’s report and released a list of 28 GM-products for which mandatory labeling was to berequired from April 2001 on. The list includes a range of soy products (e.g. tofu, soybeandrinks, eda-mame, canned soybeans), corn products (e.g. cornstarch candy, popcorn, cannedcorn), and unprocessed potatoes and tomatoes. Interestingly, the list adopted by the MAFFincludes many GM-foodsv that are already on the domestic market and are regarded as safeby the Japanese Ministry of Health and Welfare. The new labeling requirement is based onthe Japanese Law for Optimization of Standardization and Quality Labelling of AgroforestryMaterials. It has been justified largely in terms of helping consumers in making moreinformed choices while generally assuming that GM-foods are safe.239

In comparison to the EU’s labeling regulations, the labeling requirement issued by the MAFFis rather lax. It applies only to foods in which DNA or protein produced through geneticengineering is detectable, and to foods which are sold directly to consumers. Foods that areproduced from GMOs but do not contain DNA or protein produced through geneticengineering, e.g. vegetable oils or corn syrup, are not on the list. Also excluded are alcoholicbeverages (e.g. beer made from GM cornstarch), intermediate products, and feedstuffs.Moreover, the list includes only products in which GMOs are among the top three rawmaterials and/or account for more than five percent of total weight. The content of labels isrestricted to „genetically modified“, „inseparable“, or „no GMOs present.“240

In parallel to the introduction of mandatory labeling, Japan has also modified its approvalregulations. Revisions to the Food Sanitation Law, in force since April 2001, make it illegal toimport or sell GMOs that have not been inspected and approved by Japanese regulators. Thegovernment has also installed legally binding safety guidelines that allow the authorities totake legal action or issue orders to recall products.241

In summary, Japan’s GMO regulations are, in terms of their stringency, located between thoseof the United States and the European Union. No GM-crops are commercially grown inJapan, but Japan imports a large share of its soybean and corn supply from the United States.A large share of these imports contain GMOs. Japanese approval rules for GMOs have beenlax, similar to those of the United States. Since 1999, however, opposition to GMOs hasgrown. The Japanese government has responded by installing a mandatory labelingrequirement, which, compared to EU standards, is relatively lax. Japan has also strengthenedits approval regulations, has made it illegal to import GMOs not inspected and approved bythe government, and has improved its enforcement capacity.

35

Explanation

The regulatory process in Japan was, until the late 1990s, marked by a „board-room“ pattern.Regulations were developed and implemented by a closed network of government authoritiesand up- and downstream producers. GMO policy in this network was based on informal andvoluntary rules. Growing public concerns over GMOs and anti-GMO activity byenvironmental/consumer groups since the late 1990s have disrupted this pattern of policy-making.Upstream producers have, because their economic stakes in agricultural biotechnology arerather low, not engaged in substantial efforts to prevent stricter GMO-regulations – thoughthey have favored the US-model of product-oriented GMO-policy. Japanese farmers, whommany analysts expected to join the anti-GMO side, have remained largely inactive on GMO-issues. Their passivity is explained by the fact that they could not achieve economic rentsthrough stricter GMO-regulations, and that, unlike their European counterparts, they have notsuffered from a market characterized by major food safety crises. Consumer pressure hasdivided downstream producers and has driven the majority into supporting relatively laxlabeling and approval regulations. Downstream producer support for stricter regulation hasbeen facilitated by the fact that labeling regulations are rather lax, and that the new approvalregulations have merely formalized current practises.

Environmental and consumer groups

Various surveys show that public concerns in Japan over GM-foods have grown since themid-1990s.242 In 1995 and 1998 surveys, 82 and 75 percent of respondents respectivelysupported agricultural biotechnology, while 16 and 20 percent opposed it. These approvalrates were higher than those in the United States. When asked about the acceptability ofspecific biotech applications, more than 50 percent of respondents on average answeredpositively or neutrally. The same applies to questions about purchasing decisions. The majoryheld that the use of biotechnology would have a positive effect (28 percent) or no effect (33percent) on their purchasing decisions.243

By the year 2000, public support for agricultural biotechnology had dropped. An internationalpoll conducted in 2000 found Japanese consumers to top the list of those seeing the trendtowards GM-foods as negative (82 percent). Answers to what the types of benefits/advantagesof agricultural biotechnology were, as in the case of most other countries, ambiguous.Consumers regarded producers as the main beneficiaries of the new technology, whileperceiving some consumer benefits – e.g. better quality of foods, less pesticides used, feedingthe hungry. When asked about specific types of risk associated with GM-foods, Japaneserespondents perceived most risks to be higher than consumers in Europe and the UnitedStates.244 Moreover, Japanese consumers were among those least likely to buy a product ifthey knew that it contained GMOs (70 percent).245 Similar to Europe, growing opposition toGM food has gone hand in hand with low confidence in government regulators and scientificorganizations.246

The establishment of a GMO labeling panel by the MAFF in 1997 opened a channel fortransmitting these increased public concerns directly into the policy process. This panelincluded private citizens, notably industry representatives, scientists, and „ordinary“ people,thus opening the process to non-scientific influence. After substantial resistence, industry

36

representatives accepted labeling in May 1998, provided the absence of GM-ingredients couldbe confirmed.247

Many Japanese environmental/consumer groups have regarded these concessions by industryand regulators as insufficient. They have strongly criticized the new labeling and approvalregulations, which reflect these concessions, and have requested much stricter regulations,similar to those of the EU. The Consumers Union of Japan, Shufuren (HousewivesAssociation), the Home Nutrition Research Society, the Food and Agricultural Network, theNo! GMO campaign, and many other groups have requested the labeling of all foodscontaining any GMOs.The critics of the new labeling regulations have pointed to the fact that the exemptions (e.g.for soy oil and soy sauce) imply that the largest part of imported GMOs will not be coveredby labeling regulations – soy oil and soy sauce, for example, account for around 75 percent ofJapan’s soybean imports. Corn for feed and syrup, which accounts for around 80 percent ofcorn imports, will also be exempt from labeling. Rapeseed and cotton are not used in foodproducts and thus also exempt. So are the around 700'000 tons of processed tomatoes, whichmay contain GMOs – Japan does not import GM-tomatoes as such.248

The US government has protested against the new Japanese GMO regulations, callingmandatory labeling unnecessary and trade disrupting, and warned that labeling must bescience based and in conformity with WTO rules.249 Rather than deterring requests for morestringent regulations, this criticism by the United States seems to have encouraged JapaneseNGOs to address American producers directly. In a November 1999 open letter to USfarmers, a large number of Japanese NGOs asked them not to plant GM-crops anymorebecause Japanese consumers expressed a rapidly decreasing demand for such crops. They alsorequested that non-GM-crops imported into Japan contain less than 0.1 percent of GMOs. Theletter was supported by the Consumers Union of Japan, Shufuren (Housewives Association),Consumption Science Federation, the Tokyo League of Regional Women’s Organizations, theHome Nutrition Research Society, the Food and Agricultural Network, the No! GMOcampaign, and many other groups.The ability of Japanese NGOs to mobilize their members and supporters and exert pressure onregulators and producers increased further when the StarLink controversy spilled over intoJapan. In October 2000, the Consumers Union of Japan reported that traces of StarLink, aGM-corn variety made by Aventis and grown in the United States, were found in cornimported for food and animal feed, and also in a wide range of processed foods and beveragesmade from this corn. Additional testing confirmed these findings. StarLink was, at that time,approved in the United States only for animal feed because of concerns over allergenicproperties (see above). In Japan, it was (and still is) not approved for either food or feed.The appearence of StarLink in the Japanese food chain led to a public outcry and purchases ofUS corn dropped sharply. The controversy centered on the incapacity of Japanese authoritiesto control crop imports and make sure that no unapproved GM-crops were imported. It alsorevealed the inadequacy of Japan’s non-legally binding guidelines for the approval of GMOs.The MAFF, for example, argued that it had not conducted an inspection because it had notreceived an application from a company to sell StarLink corn. In addition, because Japan’sapproval guidelines were non-binding, the Japanese government did not have an effectivelegal instrument to recall illegally imported GMOs. In the StarLink case, it was thus unable toban the use of already imported StarLink corn and recall contaminated foods. US andJapanese government authorities and the firms involved have engaged in joint efforts tostrengthen controls over corn imports into Japan. As noted above, Japan has also reformed its

37

approval regulations. Whether these initiatives will restore public trust in GM-foods remainsquestionable.250

Producers

Producer responses to increasing public demands for mandatory labeling and stricter approvalregulation can be explained in terms of economic interests. Upstream producers have investedvery little in agricultural biotechnology. They have resisted the anti-GM-foods trend only tothe extent that it might spill over into other areas of biotechnology where they have a largereconomic stake. Farmers could not expect economic rents through stricter GMO regulationsand have remained passive. Downstream producers have resisted mandatory labeling the mostbecause of the costs involved and problems in substituting GMO-free alternatives for GM-products imported from the United States. Consumer pressure has split downstreamproducers: some firms have moved towards a GMO-free supply, whereas others are relyingon the new labeling legislation that does not apply to around 70 – 80 percent of GMOs in thefood chain.Upstream Producers. Compared to the United States, Japan is a latecomer in biotechnology.Japanese firms, which have entered this field since the early 1980s, include traditional food,pharmaceutical, and chemical firms, as well as large firms with non-biotech core-businesses,such as steal and construction. Start-ups have been relatively rare. To compensate for thisdisadvantage, Japanese firms have engaged in a growing number of alliances with US firms.The key players in Japan’s plant breeding and seed sector have been Sumitomo Chemical(Japan’s largest pesticide producer), Mitsui Toatsu (a seed and pharmaceutical company),Suntory (a beverage firm), Japan Tobacco (the world’s second largest tobacco producer), thePlant Research Institute (run by Mitsubishi), Kririn (a brewer), and Zennoh (the nationalorganization of agriculture cooperatives).251

Japanese upstream producers have been organized in the Japan Bioindustry Association(JBA). JBA, which represents mostly well-established firms with close ties to the government,has been the industry’s main instrument for increasing support for biotechnology among thewider public in Japan, and for lobbying policy-makers. JBA has supported product-orientedGMO-regulations along the lines of the US FDA’s approach (permissive approval regulation,no mandatory labeling), and has stressed the safety of GM-foods.252 Other associations, suchas the Society for Techno-Innovation in Agriculture, Forestry and Fisheries, the Japan HealthSciences Foundation, the Japan Association of Bioindustries Executives, and the JapanSociety for the Promotion of Science have adopted similar positions.Despite the widely shared view among JBA members and members of other industryassociations that agricultural biotechnology should be supported and regulated in a product-oriented form, upstream producers have not fought hard to prevent the new GMO-regulationsin force since spring 2001. Their restraint can be explained by two circumstances. First, whilesome Japanese upstream producers had engaged in agbiotech research between the mid-1980sand early 1990s, there has been a clear trend away from investment in this field and towardsmedical biotech applications.253 Second, the regulations installed are, compared to Europe,rather lax and were regarded as necessary to prevent a spillover of consumer concerns intonon-agricultural areas of biotechnology. In brief, the present costs and opportunity costs ofstricter regulations to Japanese upstream producers were very small.Farmers. Because Japanese farmers have traditionally benefited from a range of highlyprotectionist regulations – e.g. on rice imports – many analysts expected a similar pattern of

38

behavior with regard to GMOs. However, Japanese farmers have, for two reasons, not playeda significant role in the regulatory process.First, they have not encountered direct competition by imported GM-products. Japan has, formany years and without resistance by domestic farmers, imported a large share of its soybeanand corn supply. In 1999, around 80 percent of the soybeans consumed in Japan, and almost90 percent of corn were imported from the United States. It is the share of GMOs in theseimports, rather than the ratio of imports to domestic production, that has grown substantiallysince the mid-1990s.254 Second, the domestic food market has, unlike in Europe, not been hitby major food safety crises. Japanese farmers could thus not benefit from a reduction, throughstricter GMO regulations, of general consumer fear. Farmers in Japan are likely to becomemore strongly involved on the anti-GMO side only if (a) GM-products with direct competitiveeffects on domestic agriculture, especially GM-rice, enter their market, and/or (b) if theconsumer backlash against GMOs coincides with a major food safety crisis. So far, neitherhas been the case.

Downstream ProducersUntil the late 1990s, the Japanese food industry opposed mandatory labeling and supported apermissive GMO approval policy because it was heavily dependent on imports of soybeansand corn from the United States. From 1996 on a large share of these imports includedGMOs. Downstream producers feared that GMO-labels would cut into their sales of GM-products and that they would be forced to find more costly substitutes.255

What many observers regarded as very unlikely in the mid-1990s256 has occurred since theend of the decade. Growing public concerns over GM-foods have forced many downstreamproducers to change their position and have divided the industry. Some firms have continuedto express concern that the new labeling requirement will unnecessarily stigmatize GMO-products and impose additional costs on the food industry. But, similar to Europe, many firmshave switched to GMO-free products or GM-ingredients below regulatory thresholds to avoidlabeling. Japan’s largest maker of soybean protein food products, Fuji Oil Co., announced inSeptember 1999 that it would stop using GM-soybeans. Similar policies have been adopted byItochu Feed Mills, Nippon Flour Mills, Nissin, Kanot Soybean Wholesale, Kirin and SapporoBreweries, and many other firms.257 Because downstream producers have been divided overthe issue, food producer organizations, such as the Japan Tofu Association, have neitherencouraged nor discouraged such steps.258

The economic implications for downstream producers are not entirely clear, but appear to berather modest. First, the new labeling requirement does not apply to 70 – 80 percent ofproducts containing or produced with GMOs. Most Japanese downstream producers that haveopted for “GMO-free” products have based their new policy on the assumption that GMO-free means GMO-content below the threshold for mandatory labeling. Moreover, mostJapanese trading firms have noted that the new GMO approval rules would not significantlychange import practices: most importers were, when the new regulations were introduced,already cooperating with the MHW in seeking approval for new GM-products.Second, most Japanese importers have quickly and efficiently re-organized their supply byestablishing segregation and testing systems in collaboration with US exporters.259 Whilesegregation and testing create some additional but clearly limited costs for downstreamproducers260, the evolution of premiums for GMO-free crops remain more uncertain.Premiums for non-GM corn have remained low because there is an abundant supply on worldmarkets. Premiums for non-GM soybeans, on the other hand, have in 1999 and 2000 beensubstantial (see Figure 6).

39

Figure 6: Prices of GM-Soy and Non-GM Soy

20000

21000

22000

23000

24000

25000

26000

27000

28000

May 0

0

June

00

July

00

Augu

st 00

Sept

embe

r 00

Octob

er 00

Nove

mber 0

0

Dece

mber 0

0

Janu

ary 0

1

Febr

uary

01

March 0

1

April 0

1

May 0

1

Time

Pric

e (Y

en p

er 1

.00

0 k

g)

Non-GMO Soybeans U.S. Soybeans

Source: Tokyo commodity exchange, www.tge.or.jp/english/market/chart.e/inputpram_m.e.html

How premiums will develop in future depends on whether US producers are willing and ableto re-orient production towards non-GM soybeans. It also depends on future demand for non-GM soybeans in Japan and in Europe.261 However, in view of widespread public adversionagainst GM-foods in Japan, it is likely that downstream producers will be able to pass theadditional costs on to consumers.262

Conclusion

Changes in the collective action capacity of consumers and producers provide an adequateexplanation for recent changes in Japan’s GMO-policy. Consumer influence on the regulatoryprocess has grown since the late-1990s due to institutional changes (composition of theMAFF’s labeling commission), growing consumer concerns about GMOs, and the spilloverof the StarLink controversy from the United States to Japan.Upstream producers have favored the US-model of GMO regulation but have invested little inpreventing stricter GMO-regulations in Japan. Their business priorities had, already in theearly to mid-1990s, shifted to medical applications of biotechnology. Farmers have remainedpassive because soy and corn imports have not displaced domestic producers on a significantscale. Moreover, in the absence of major food safety crises farmers did not see stricter GMO-regulations as an instrument for reducing general consumer fears. Consumer pressure hasmotivated most downstream producers to move towards products not subject to mandatorylabeling. This change in downstream producer behavior has been facilitated by the fact thatmost products containing GMOs or produced with GMOs are exempt from the labelingrequirement.Strong consumer support for strict GMO-regulations, potentially growing support by farmersfor such regulations, weak support by upstream producers for a relaxation of existingregulations, and differences among downstream producers make it unlikely that Japan willrevert to the status quo ante. Whether Japanese policy on GMOs will move further towardsthe European model depends on several conditions. Additional controversies over food safety

40

issues, similar to the StarLink problem, or larger food safety crises, such as the BSE problem,would boost public pressure for stricter GMO regulations. Such problems, and/or theappearance on the Japanese market of GM-products with direct competitive effects ondomestic producers (e.g. GM-rice) would result in stronger farmer opposition to GM-foods.Finally, growing supply of non-GM soy by US producers would motivate Japanesedownstream producers to increase their use of non-GMO alternatives in products thatcurrently do not require labeling. Such practices might, in turn, encourage Japanese regulatorsto tighten labeling and approval regulations.

Conclusion

In this section I summarize and compare the results of the case studies on the EuropeanUnion, the United States, and Japan. I also assess the implications of regulatory diversity forinternational trade.

GMO-Regulations in the EU, the United States, and Japan

Guided by a theoretical argument that focuses on the collective action capacityof consumer and producer interests I examined GMO-regulations in the EU, the United States,and Japan.Regulatory outcomesIn the mid-1980s, the starting point of the analysis, the biotechnology policies of the EU, theUnited States, and Japan were very similar. At the end of the 1980s, they began to diverge.From 1990 on, the EU has concentrated on process-oriented regulations and has adoped morestringent approval and labeling standards than the United States and Japan. As a consequence,very few GMOs have been approved for commercialization in the EU, commercial planting ofGM-crops is almost non-existant in EU countries, and the number of field trials is far lowerthan in the United States. The number of labeled products on the EU market has dropped toalmost zero as producers have chosen to give up rather than label GM-foods. The EU’smarket for GMOs has thus shrunk to (an also declining market for) GM food ingredients andanimal feed below the threshold for mandatory labeling.US regulatory policy has remained largely constant over time. Regulators have taken theposition that the application of biotechnology does not lead to substantially differentproperties of food products, and that GM-foods and feeds are generally safe. Following thisproduct-oriented policy, the FDA and the USDA have, through an informal notificationprocedure and without pre-market risk-assessments, approved most industry requests for fieldtesting and commercialization of GMOs. Producers may voluntarily label GM-foods but arenot obliged to do so. Around 50 GMOs are on the US market, several thousand GM-varietieshave been authorized for field testing, and GM-crop acreage increased dramatically between1996 and 2000.Until spring 2001, Japan’s regulatory policy on GMOs was very similar to US policy. GMOswere regulated under pre-existing product-oriented legislation. GMO approval standards werevery permissive and labeling of GM-foods was neither mandatory nor voluntarily practiced.Around 20 GMOs were approved by Japanese authorities. The only substantial differencebetween Japanese and US regulatory policy was that Japan did not approve commercial

41

planting of any GM-crops. In April 2001, however, new regulations on approval andmandatory labeling entered into force. These standards are laxer than those of the EU butstricter than those of the United States.

Collective action capacity of environmental and consumer groupsConventional theories of regulation predict that environmental and consumer groups, becauseof their large and heterogeneous membership, experience greater problems than producers inmobilizing supporters and influencing public policy. I found, however, that the collectiveaction capacity of environmental and consumer groups has varied substantially across the EU,the United States, and Japan. This variation can be traced back to differences in: (a) publicperceptions of agricultural biotechnology; (b) consumer trust in regulatory authorities; and (c)access to the regulatory process. Variation along these lines has had significant effects onregulatory processes and their outcomes.In the EU, negative public perceptions of agricultural biotechnology, low trust in regulators(mostly the result of major food safety crises), and substantial access to the regulatory process(principally via the European Parliament) have facilitated strong and sustained NGOengagement in the issue, particularly since the mid-1990s. NGOs were thus able to launchpowerful campaigns against upstream and downstream producers (market pressure). Thesecampaigns have also transformed the GMO issue into one of „highy saliency“ and (in the eyesof the wider public) „low complexity“ – which tends to produce a „hearing room“ pattern ofpolicy-making. This hearing room pattern combined with the multi-level nature of EU policy-making has opened the door to strong NGO influence on policy-decisions.In the United States, relatively positive consumer perceptions of agricultural biotechnology,trust in regulators and scientists (mostly due to the absence of larger food safety crises), andlimited access to the regulatory process have made it difficult for NGOs to mobilize theirsupporters on the GMO issue and influence producers and policy-makers. As a result, USenvironmental and consumer groups have become involved in agbiotech issues much laterthan their European counterparts. NGO activity on a larger scale began only in the late1990s/early 2000s. At that time, public opinion on GMOs turned more negative and theStarLink controversy emerged. In contrast to their European counterparts, most US NGOshave lobbied for stricter approval procedures and mandatory labeling rather than a ban onGMOs. So far, the FDA and the USDA have remained largely immune to this pressure.Japanese NGOs became involved in the GMO issue only in the late 1990s/early 2000s. Bythen, consumer perceptions of agricultural biotechnology had become more negative, publictrust in regulators had decreased (primarily due to the spillover of the StarLink controversyfrom the United States to Japan), and NGO’s access to the regulatory process had improved.Environmental and consumer groups in Japan have thus been able to influence the regulatoryprocess via lobbying of policy-makers and pressure on downstream producers (see below).

Collective action capacity of producersWhile conventional theories of regulation regard producers as a homogenous interest group, Idistinguished upstream producers (agbiotech firms), farmers, and downstream producers(wholesalers, retailers, food processors). I hypothesized that differences in industrial structureand the extent of NGO opposition to GMOs account for variation in the collective actioncapacity of pro-GMO producers.In the EU, negative consumer perceptions of GM-foods and NGO campaigns have dividedproducers and have thus diminished the collective action capacity of pro-GMO interests.Large upstream producers, such as Monsanto and Novartis, became major targets of NGOcampaigns and have never gotten out of the defensive. EU farmers, who throughout the 1990sand beyond have suffered from major disturbances in European food markets, have becomeincreasingly hostile to GMOs. This opposition is bound to grow further if GM-products

42

affecting the competitive position of EU farmers more directly enter their markets, and ifdownward pressure on prices in EU food markets continues. NGO pressure and negativepublic risk perceptions have driven downstream producers into supporting mandatory labelingand stricter approval standards. While most downstream producers expected that labelingmight stop the anti-GMO trend half way, the unintended effect of mandatory labeling wasexposure of labeled GM-foods to even greater consumer and NGO pressure. As a result,downstream producers have deserted the market for labeled GM-foods, hoping that theconsumer backlash will not spread to thousands of processed foods in the EU that containGMOs below the threshold for mandatory labeling.US policy on GMOs has, from the very beginning, been supported by a homogeneous, well-organized, and well-funded coalition of export-oriented farmers, commodity associations, theupstream industry, and (until the late 1990s) downstream producers. These producer interestshave enjoyed priviledged access to US regulatory agencies, notably the FDA. US policy-making in this area thus corresponds to the „board-room“ pattern, the opposite of the hearing-room pattern that has prevailed in the EU. The most serious challenge to US GMO-policycomes from increasing public concerns about agricultural biotechnology since the late 1990s,and the StarLink controversy, which broke out in fall 2000. In part motivated by thesedevelopments, several large biotech firms have divested their agbiotech divisions, thuscontributing to growing differences of interest between medical and agricultural biotechfirms. Increased consumer and NGO pressure has encouraged some large and many smalldownstream producers to move towards GMO-free products and voluntary labeling. Thisdevelopment has weakened pro-GMO campaigns by up- and downstream producer groups.The StarLink controversy has also led to conflicts between farmers and upstream producersover damages and compensation. It is unlikely, however, that these differences amongproducers will grow to the extent observed in the EU.Until the late 1990s, Japan’s GMO-policy corresponded to the „board-room“ pattern observedalso for the United States. Since 1999, consumer influence on the regulatory process hasincreased due to growing public concerns about GMOs, the spillover of the StarLinkcontroversy from the United States to Japan, and institutional changes (notably thecomposition of the MAFF’s labeling commission). Low collective action capacity due todifferences in industrial structure explains why producers were neither willing nor able tojointly resist growing consumer pressure and stricter regulations. Upstream producers, whohave supported the US-model of GMO regulation, have not put up much resistance becausetheir business priorities are in medical biotechnology. Farmers have remained passive becausesoy and corn imports have not displaced domestic production on a significant scale, andbecause, unlike their European counterparts, they have not been plagued by surplus problemsin a market disrupted by food safety crises. Finally, consumer pressure has motivated manydownstream producers to move towards GMO-free products and to support the new approvaland labeling regulations. This change in downstream producer behavior has been facilitatedby the fact that, because of many exemptions, most products containing or produced withGMOs are not subject to mandatory labeling.In summary, the three case studies support the theoretical argument that GMO regulationstend towards greater stringency when:

(a) the collective action potential of environmental/consumer interests is high – the latterbeing a function of public risk perceptions, public trust in regulators, and institutionalaccess to regulatory policy-making;

43

(b) the collective action potential of producer interests is low – the latter being a functionof industrial structure, NGO and consumer pressure, and institutional access toregulatory policy-making.

Trade implicationsDifferences in GMO-regulations, such as those described and explained in this chapter, havehad substantial effects on international trade in agricultural commodities as well as processedfoods and beverages.263 The costs of growing regulatory restrictions in the EU, Japan, andother countries have fallen primarily on the United States, by far the largest exporter ofGMOs.Many analysts have therefore predicted trade conflict over differences in GMO-regulations,particularly between the United States and the EU, on a scale that would dwarf othertransatlantic trade disputes, such as those over bananas, export subsidies, or growthhormones.264 Such predictions have been nurtured by a variety of statements by high USgovernment officials suggesting that an escalation of the dispute is likely. In 1999, forexample, Stuart Eizenstat, then US Under Secretary of State, argued before the US Congressthat the GMO-issue was „the single greatest trade threat that we face systemically with theEuropean Union“. He also claimed that „almost 100 percent of our agricultural exports in thenext five years will either be genetically modified or combined with bulk commodities thatare genetically modified.“265

Most predictions of whether the US government will or will not escalate the latent tradedispute by formally requesting a WTO panel have been informed by legal assessments ofwhether the United States would or would not win such a dispute.266 Those concluding thatthe United States would win (mainly the proponents of legal action) claim that GMO-regulations more stringent than those of the United States violate the WTO Agreement onTechnical Barriers to Trade (TBT Agreement, particlarly Article 2.2). This agreement holdsthat regulations that disrupt trade must fulfil a legitimate objective, and must not be moretrade restrictive than necessary. They argue that the EU’s GMO-regulations fail the test onboth accounts, because (a) they cannot be justified on deception-prevention or public healthgrounds, (b) voluntary labeling would achieve the same objective (satisfy consumer demandfor more information); and (c) they are not consistent, neither internally nor acrossagricultural and medial biotechnology.267

Those claiming that the United States would not win a WTO dispute over the EU’s GMO-policy point out that the EU’s measures could be defended as long as it was feasible forproducers to comply and they were applied equally to domestic and foreign products. Ifmandatory labeling rules were challenged under the WTO’s Agreement on Sanitary ofPythosanitary Measures (SPS Agreement), the EU could also defend its policy as a short termmeasure while more scientific evidence was being gathered to address uncertainties aboutconsumer safety. The defendent’s chances of success would be even higher under the TBTAgreement. It would have to show that mandatory labeling was intended and designed toachieve a legitimate objective and that the costs were proportional to the purpose. It couldthus argue that the objective was consumers’ right to know and that mandatory labeling wasan effective measure to that end.268

While legal opinion on whether or not the United States would win a WTO case on GMOsremains divided, the above analysis on regulatory policy in the EU and Japan suggests that alegal victory of the United States could be a mixed blessing. As in the growth hormones case,it is very likely that the defendents (the EU, possibly also Japan and other countries that haveintroduced mandatory labeling) would not make the concessions requested by the plaintiff.Regulatory authorities of the EU and its member states are under severe pressure fromenvironmental and consumer organizations. Most of these groups regard labeling as only an

44

intermediate step on the way to a GMO-ban. In their view, labeling solves problems ofinformation asymmetry but does not accomodate consumer demands for risk-shielding.Similarly, labeling does not solve European farmers’ problems of agricultural surpluses,market disruptions caused by food safety crises, and the prospect of growing import-competition through GM-products. Consequently, farmer preferences are likely to move frommandatory labeling to banning GMOs rather than from mandatory to voluntary labeling.Because of substantial consumer concerns over GMOs, anti-GMO campaigns by NGOs, andsevere collective action problems among downstream producers, support by the latter for arelaxation of GMO-regulations is very unlikely.269

Strong consumer support for strict GMO-regulations, potentially growing support by farmersfor more stringent regulations, weak support by upstream producers for a relaxation ofexisting regulations, and differences among downstream producers make it equally unlikelythat Japan will revert to laxer GMO-standards. If Japanese GMO policy moves away from thestatus quo, it is likely to be towards the European rather than the US regulatory model.270

It is very unlikely, therefore, that an escalation of the dispute would bring about internationalconvergence of GMO-regulations along the lines of current US policy. Neither are morecooperative efforts, notably those by the Codex Alimentarius Commission, likely to achievethat objective anytime soon – be it harmonization according to current EU or USpreferences.271 This does not necessarily imply that the United States will refrain fromformally carrying the dispute into the WTO at some point. There are a variety of reasons whyit may do so; for example, to obtain concessions by the EU in other trade disputes, to deterother countries from adopting stricter GMO-regulations, or to increase domestic politicalsupport by specific constituencies (e.g. upstream producers, GM-crop exporters).Nonetheless, because the defendants are very unlikely back down, it is highly questionablewhether such dispute escalation would in the end help US exporters of GM-products. Itwould, moreover, burden the world trading system with an additional dispute at a time whenWTO members are struggling to re-launch trade liberalization talks in the wake of the abortedmilennium round.

Thomas Bernauer is Professor of International Relations at the Swiss Federal Institute ofTechnology in Zurich, Switzerland. Information on his research and teaching activity isavailable at www.bernauer.ethz.ch .

45

Endnotes 1 Fagan defines recombinant DNA techniques as involving „the isolation and subsequentintroduction of discrete DNA segments containing the gene(s) of interest into recipient (host)organisms. The DNA segments can come from any organism (plants, animals, or microbes).”The application of this technique is commonly referred to as genetic engineering and theproducts as genetically engineered or genetically modified products. See Fagan, John 1995.“A Science-Based, Precautionary Approach to the Labeling of Genetically EngineeredFoods”, www.naturallaw.org.nz.2 The term genetically modified food is defined in this chapter as “foods and food ingredientswhich contain or consist of genetically modified material or which are produced from, but donot contain, genetically modified material“. See Bredahl, Lone et al. 1998. “ConsumerAttitudes and Decision-Making with Regard to Genetically Engineered Food Products - AReview of the Literature and a Presentation of Models for Future Research”, Journal ofConsumer Policy 21, 251-277:252.3 As of March 2001, around 50 GMOs had been approved by US regulators, around 25 inJapan, and around 9 in the European Union. Particularly in the form of food additives, theseGMOs are present in thousands of food products in most OECD countries.4 Hoban, Thomas J. 1998. “Trends in Consumer Attitudes about Agricultural Biotechnology”,AgBioForum 1,1,3-7:5.5 Pauker, Peter 1999. “Consumer Benefits of Biotechnology”, www.cid.harvard.edu.6 EuropaBio 1999. “GMO Fact Sheet”, www.europa-bio.be.7 The remaining one percent was grown in Australia, Mexico, South Africa, Romania,Ukraine, Spain, Germany, and France. Soybeans (58%), corn (12%), cotton (12%), and canola(7%, all 2000 figures) account for almost 100 percent of global GM-crop planting. The areasown to GM-seeds appears to have reached a plateau in 1999/2000, with weaker growthexpected for the next few years. 73 percent of these crops were modified for herbicidetolerance, 22 percent for insect resistance, and 5 pecent for both. In 2000, 34 percent ofsoybeans planted globally were genetically modified, 16 percent of cotton, 11 percent ofcanola, and 7 percent of corn.8 In 1997, Europe’s life sciences firms spent around 1.9 billion Euro and US firms around 8.3billion Euro on R&D for agbiotech purposes.9 Agbios 2000. „World Use of Genetically Modified Crops up 11% in 2000“, December 19,2000, International Service for the Acquisition of Agri-Biotech Applications, www.isaaa.org.See also www.rafi.org; Kalaitzandonakes, Nicholas 1999. “A Farm Level Perspective onAgrobiotechnology: How Much Value and For Whom?”, AgBioForum (2) 2: 61-64;EuropaBio 1999.10 David Vogel. 2001. The Regulation of GMOs in Europe and the United States: A Case-Study of Contemporary European Regulatory Policy. Manuscript; Sapolsky, Harvey (ed.)1986. Consuming Fears: The Politics of Product Risks. New York, Basic Books; Wildlavsky,Aaron 1985. But is it True: A Citizen’s Guide to Environmental and Safety Issues.Cambridge, Harvard University Press; Wilson, James Q. 1980. The Politics of Regulation.New York: Basic Books; Brickman, Ronald et. al. 1985. Controlling Chemicals. The Politicsof Regulationin Europe and the United States. Ithaca, Cornell University Press; Vogel, David1986. National Styles of Regulation : Environmental Policy in Great Britain and the UnitedStates. Ithaca, Cornell University Press; Jasanoff, Sheila 1993. „American Exceptionalismand the Political Acknowledgment of Risk“, in Burger, Edward J. (ed.). Risk. Ann Arbor,University of Michigan Press: 66. See also Stephen Breyer.1993. Breaking the Vicious Circle:

46

Toward Effective risk Regulation. Cambridge: Harvard University Press, and O’Riordan1979.11 See www.rafi.org and Allan Woodburn Associates 2001. In 2000, four of the seven largestagrochemical firms in the world were from Europe (Syngenta, 22% world market share;Aventis, 13%; BASF, 11%; Bayer, 7%), and three from the United States (Pharmacia, 13%;DuPont, 8%; Dow, 7%). The total market share of the four European firms was much largerthan the total market share of US firms.12 ibid.13 Stigler, George J. 1971. “The Theory of Economic Regulation”, Bell Journal of Economicsand Management Science (2):3-21; Baron, David P. 2000. Business and Its Environment.Upper Saddle River: Prentice-Hall.14 Lemaux, Peggy G. 1998. „Impact of Public Perception on Regulatory Policy forAgricultural Biotechnology“, presentation given during the Scientific Session at the NaraInternational Symposium held in Nara, Japan, November.15 Food and Drug Administration (FDA) 2000. „Report on Consumer Focus Groups onBiotechnology“, October 20, www.fda.gov.16 Moore, Elizabeth 2000. Food Safety, Labeling, and the Role of Science: RegulatingGenetically-Engineered Food Crops in Canada and the United States. Paper prepared forECPR Joint Sessions.17 Platzer 1995:78-79; Scharpf 1995.18 Baron 2000; Stigler 1971.19 Richards, John E. 1999. Toward a Positive Theory of International Institutions: RegulatingInternational Aviations Markets“, International Organization 53 (1): 1-37; Gormley, WilliamT. 1986. „Regulatory Issue Newtorks in a Federal System“, Polity 18: 595-620.20 Chen, Zhiqi and Alison McDermott 1998. „International Comparisons of BiotechnologyPolicies“, Journal of Consumer Policy 21: 527-550. Hoban 2000.21 Vogel 1986, 2001; Jasanoff, Sheila 1995. “Product, Process, or Programme: Three Culturesand the Regulation of Biotechnology”, in Bauer, Martin (ed.). Resistance to New Technology.Cambridge: Cambridge University Press; Echols, Marsha A. 1998. „Food Safety Regulationin the European Union and the United States: Different Cultures, Different Laws“, ColumbiaJournal of European Law (4) 52.22 See, e.g., Moore 2000.23 See, e.g., Stigler 1971; Baron 2000.24 To keep the argument simple, we ignore, at this point, that regulators may have interests oftheir own.25 Many authors have exposed the empirical flaws and theoretical deficiencies in Stigler’s(1971) producer dominance argument (e.g. Meier 1988; Gormley 1986). In the absence of aparsimonious alternative, most economists still subscribe to this model. See Baron, D. 1995.„The Economics and Politics of Regulation: Perspectives, Agenda and Approaches“ in Banks,J. S. and E. A. Hanushek (eds.). Modern Political Economy. Cambridge, CambridgeUniversitiy Press: 10-62; Baron 2000). Political scientists have mostly rejected Stigler’smodel and have offered more complex, more descriptive, and less generalizable models (e.g.Gormley 1986; Héritier 1993).26 Olson, Mancur 1965. The Logic of Collective Action. Cambridge, Mass., HarvardUniversity Press:48-65.27 In explaining why large voluntary associations are sometimes still able to form andinfluence public policies, Olson introduces vague notions of social incentives in “federal”groups and the mass media, as well as individual benefits such as prestige.28 Edward Groth, "Consumer Perceptions of and Responses to Potentially HazardousTechnologies," ICGFI (ed.). Document No. 18. Vienna: ICGFI Secretary: 4123-124; Jennifer

47

B. Wohl. 1998. Consumers’ Decision-Making and Risk Perceptions Regarding FoodsProduced With Biotechnology, in Journal of Consumer Policy 21:387-404.29 See also Bruno S. Frey and Gebhard Kirchgässner 1994. Demokratische Wirtschaftspolitik:Theorie und Anwendung. München, Vahlen:209; and Marwell/Oliver 1993.30 Friedman, M. 1991. „Consumer Boycotts: A Conceptual Framework and ResearchAgenda“, Journal of Social Issues 47:149-168; Putnam, T. 1993. „Boycotts are Busting Outall Over“, Business and Society Review 85: 47-51.31 If scientifically identifiable causes and consequences of risks fully determined theaforementioned issue properties, public outrage would, within the same issue and at the samepoint in time, be constant across countries. Public outrage could thus explain regulatorydifferences across issue-areas but not differences across countries within a given issue-area,such as GMOs. However, in many cases of environmental or consumer protection we observesubstantial variation in public risk perceptions within issue areas, both across countries andtime. We assume that variation in public risk perceptions (or public outrage) is influenced bya variety of factors, including the following. First, the greater scientific uncertainty about agiven risk is the more room there may be for cross-national differences in public outrage. Thisargument remains contested, particularly because there are cases where strong cross-nationaldifferences in risk perception coincide with little scientific uncertainty (e.g. in the case of foodirradiation). Second, local environmental or public health conditions and budget constraints.These conditions, which are key explanations in much of the political economy literature onregulation, are less interesting for the analysis in this chapter because they are largely constantacross the EU, the United States, and Japan. Third, public sensitivity to certain environmentalor health risks may vary due to cultural or historical differences. This argument can often befound in the public policy literature on regulation. Its great popularity notwithstanding, socialscientists have not been able to deliver a coherent and empirically substantiated theory for thisargument. In our view, it remains highly implausible why, on food issues for example,Europeans should be inherently more risk averse than North Americans or the Japanese.Finally, priorities of environmental or consumer interest groups may differ across countriesand time, and these groups may experience different degrees of success in raising publicawareness (or outrage). For the sake of simplicity we take the extent of public outrage asexogenously given at this point, but discuss possible explanations for cross-nationaldifferences in public risk perceptions in the empirical analysis.32 Several authors have proposed a further differentiation of this argument by distinguishingbetween regulations focusing on the quality of products and regulations focusing onproduction processes. They claim that product regulations are easier to instrumentalize forprotectionist purposes than process regulations. Consequently, product regulations, onaverage, tend to be more stringent than process regulations and also vary more acrosscountries. In many empirical cases, including GMOs, a straightforward distinction of productand process regulations is difficult. But the underlying argument of protectionist interests canbe employed in a broader fashion. See Dale D. Murphy 1995. Open Economics andRegulations: Convergence and Competition among Jurisdiction. Dissertation, MIT,Department of Political Science (Cambridge Mass., MIT, 1995; and Fritz W. Scharpf 1998.„Globalization: The Limitations of State Capacity“, Swiss Political Science Review(4) 1:92-98.33 Whether protectionist rent-seeking is in fact a strong driving force of environmental andconsumer regulation, as suggested by some free-trade protagonists, is an empirical question.Hathaway and others have shown that the increasing openness of countries to world trade has,on average, not led to increasing demand for protectionist regulation because there has been apositive feedback effect: increasing openness changes industrial structures in a direction thatreduces industry demands for protection against foreign competition. See Oona A Hathaway

48

1998. “Positive Feedback: The Impact of Trade Liberalization on Industry Demands forProtection”, International Organization (52) 3:575-612.34 See Stigler 1971; Baron 2000.35 Foster 2001.36 Whether these variables are, individually, necessary or sufficient conditions for stricterregulation is an empirical question. For example, there may be cases in which environmentaland consumer groups clearly dominate over a homogeneous and well-endowed coalition ofproducers. We assume, however, that this scenario is less likely in cases where regulationimposes high costs on producers.37 Caduff, Ladina 2001. Regulating the Use of Growth Hormones in Meat Production in theEuropean Union and the United States. Center for International Studies, ETH Zurich.; Hall, P.and R. C. Taylor 1996. “Political Science and the Three New Institutionslisms”, PoliticalStudies, XLIV: 936-957.38 See, e.g., Kitschelt, Herbert P. 1986. „Political Opportunity Structures and Political Protest:Anti-Nuclear Movements in Four Democracies“, Britisch Journal of Political Science (16).39 March, James G. and Johann P. Olsen 1996. „Institutional Perspectives on PoliticalInstitutions“, Governance: An International Journal of Policy and Administration, vol. 9 (3):247-263; Hall/Taylor 1996; Nordlinger, Eric 1981. On the Autonomy of the Democratic State.Cambridge, Harvard University Press.40 The early years of biotech regulation in the EU are well described in Patterson, Lee Ann2000. „Biotechnology Policy: Regulating Risks and Risking Regulation“, in Wallace, Helenand William Wallace (eds.). Policy-Making in the European Union. Oxford University Press,Oxford: 317-343; and Cantley, Mark F. 1995. „The Regulation of Modern Biotechnology: AHistorical and European Perspective: A Case Study in How Societies Cope with NewKnowledge in the Last Quarter of the Twentieth Century“, in Rehm, H.-J. et. al.Biotechnology: Legal, Economic and Ethical Dimensions. Volume 12,VCHVerlagsgesellschaft, Weinheim: 505-681.41 Other EU legislation relevant to GMOs has, for example, focused on worker health andsafety protection (90/679 and 93/88), and the contained use of GMOs for commercial andresearch purposes. All EU legislation on GMOs is available at www.europa.eu.int/eur-lex42 Whereas part B of directive 90/220 regulates the release of such organisms into theenvironment for purposes of research and development, part C covers the approval of GMOsor products containing GMOs for commercialization in the EU.43 Chen/McDermott 1998.44 According to directive 90/220, producers or importers of a GM-product first have to notifythe national authority of the EU country concerned and provide technical documentation and afull risk assessment. If the national authority approves the GM-product, the request moves tothe EU level, where the other 14 member states are consulted. If there are no objections, theGM-product is approved for sale in the EU. If there is opposition, the issue is dealt with bythe EU’s Scientific Committee, the Commission, and the EU’s Regulatory Committee. Ifthere is still no agreement, the issue is decided by the Council of Ministers with qualifiedmajority.45 GM-crops can be listed in a EU catalogue of common varieties that are approved for EU-wide planting. No GM-crops have so far been listed in that catalogue.46 www.transgen.de; see also Levidow et al. 1997.47 Competent Authorities are defined as government officials in EU member countries whoare responsible for implementing EU regulations and representing their countries in EUnegotiations and approval processes.48 The approval decisions were taken by the Competent Authorities of the EU.

49

49 See, for example, ICTSD, July 19, 1999. Article 16 gives any EU member state the right to“provisionally restrict or prohibit the use h50 By April 1998, 16 GMOs had been approved by the EU, another 11 applications werepending, including four that had been approved by EU scientific committees but were rejectedby member states.51 International Trade Reporter 1999, June 30: 1090; www.transgen.de.52 Biosafety&Biotechnology Service 1998, October 12, www.biosafety.ihe.be; Jany/Greiner1998:1-3; Food Manufacture 1999, March 1; Gately, Jennifer L. 1997. “Novel Food andNovel Food Ingredients.” Colombia Journal of European Law (3) Winter/Spring: 318.53 www.transgen.de.54 Behrens, Maria 1997. Genfood: Einführung und Verbreitung, Konflikte undGestaltungsmöglichkeiten. Berlin, Sigma.55 Chege, Nyaguthii 1998. “Compulsory Labeling of Food Produced from GeneticallyModified Soya Beans and Maize”, Colombia Journal of European Law (4) Winter/Spring:179-181: 179.56 Robert-Koch-Institut 1999. “Gesetzliche Grundlagen, Richtlinien, InternationaleRegelungen“, April 8, www.rki.de; Jany/Greiner 1998: 1-3; Chege 1998:179.57 www.transgen.de.58 EU News, 2001, February 16; International Trade Reporter 1999, June 30: 1090; MAFF,Ministry of Agriculture, Fisheries and Food 1999. “Factsheet: Genetically Modified Foodsand Crops”, www.biotechknowlegde.com; Neue Zürcher Zeitung (NZZ) 2001, February 15:63; www.transgen.de.59 As of late 1998, when the EU’s moratorium on GMO-approvals was put in place, thenumber of GM-food products evaluated in the United States was 34, in the EU it was 6. Theaverage time to evaluate products was 6 months in the United States, and between 18 and 19months in the EU. The acreage involved in field testing was 30 million in the US, and 20'000in the EU (Lemaux 1998).60 United States Department of Agriculture (USDA), 2001, February 16;Biosafety&Biotechnology Service 1998, October 12, www.biosafety.ihe.be; MAFF 1999;www.transgen.de.61 See McCormick, D. 1992. Letting Consumers Choose, Bio/Technology, 10: 717; Frewer,L.J. et al. 1997. ‘Objection’ Mapping in Determining Group and Individual ConcernsRegarding Genetic Engineering. In Agriculture and Human Values, 1997, p.1-13;62 For a more generic explanation of why the Olsonian problem of collective action affectsconsumer groups less at the European level than at the domestic level, see Alasdair R. Young,European Consumer Groups, in Justin Greenwod and Mark Aspinwall, Collective Action inthe European Union. London: Routledge.63 Nature 1997, 387: 821-836; Eurobarometer 1996, 1999; Braun, R. August 1999. „ThePublic Perception of Biotechnology in Europe: Between Acceptance and Hysteria“, Biolink,Communication on Biotechnology, Worb, Switzerland; Nature Biotechnology, Vol.18, Sept.2000:935-38.64 Angus Reid Group 1999, International Awareness and Perceptions of Genetically ModifiedFoods (www.angusreid.com). These polls were conducted in late November/early December1999. Awareness: Percentage of respondents who have read or heard something aboutgenetically modified foods. Negative view on GM-foods: Percentage of respondents whoview the trend towards GM-foods as negative. Would not buy GM-food: Percentage ofrespondents who would be less likely to buy a product if they knew it contained GMOs. Noadvantages in GM-food: Percentage of respondants aware of GM-foods perceiving noadvantages of GM-food. GM-foods pose health and safety problems: Percentage of

50

respondants aware of GM-foods stating that food safety, health concerns and allergies are themain risks or disadvantages associated with GM-food,65 See also Financial Times 1998, November 18:10. In another survey, 65% of Swedishrespondents, 69% in Austria, 50% in Germany, and 39% in the UK considered GM-foods aserious risk, whereas only 14% of US respondents thought so (Science 1998, July 30: 714).66 In a 1995 survey, 21 percent of US respondents considered genetic engineering a serioushealth hazard, while 85 percent in Sweden, 60 percent in Austria, 57 percent in Germany, 48percent in the Netherlands, 39 percent in the UK, and 38 percent in France did so. Hoban,Thomas. 1997. Consumer Acceptance of Biotechnology: An International Perspective, NatureBiotechnology, March:233; San Francisco Cronicle, August 16, 1999:A6; Science, July 30,1998:714.67 Eurobarometer data for 1996 and 1999 show that the percentage of respondents who feltthat a particular biotechnology application should be envouraged fell from 75 to 63% forhuman genetic screening, from 71 to 57% for new medicines, from 58 to 42% for insect-protected crops, and from 44 to 31% for genetically improved foods. See also NatureBiotechnology, Vol.18, Sept.2000:935-38.68 Fischhoff, Baruch et. al. 1989. Acceptable Risk. Cambridge, Cambridge University Press;Powell, D. A. 1998. From Frankenstein to Frankenfood: Talking about AgriculturalBiotechnology . Country Guide; Groth 1991.69 Hoban (2000), for example, lists the following influences: „greater trust in government andscientists“, „culture supports innovation and progress, and accepts risks and mistakes“,„ongoing proactive education for industry, leaders, media, and consumers“, low impact andcredibility of protest groups“, „relatively balaanced media coverage“, „different views onfarming and food“, and „recognition of benefits from first products.“ Zechendorf (1998)concentrates on cultural, religious, geographic, and other influences. In a logistic regression,Gaskell et al. (Nature Biotechnology, Vol.18, Sept. 2000: 932) find that trust, age, gender,education and knowledge each independently and significantly predict the probability ofopposition to GM-foods. See also Lone Bredahl et al. 1998. Consumer Attitudes andDecision-Making With Regard to Genetically Engineered Food Products, in Journal ofConsumer Policy 21:251-277; David Vogel. 2001. The Regulation of GMOs in Europe andthe United States: A Case-Study of Contemporary European Regulatory Politics. Manuscript.Vogel (2001) in particular provides convincing evidence against cultural explanations fortransatlantic differences of consumer perceptions and regulation.70 Many authors have proposed that public acceptance of biotechnology is a function ofknowledge level, awareness of the benefits of the technology, confidence and trust inregulatory agencies, and several other variables. See Kamaldeen, Sophia and Douglas A.Powell 2000. Public Perceptions of Biotechnology. Food Safety Network Technical Report.Department of Plant Agriculture, University of Guelph, Canada; Hoban 1996; Lemaux 1998.Most of these explanations are difficult to support empirically. For example: Consumerawareness of GM-foods, measured in very general terms, is very high in Europe (see Figurex). But it is impossible to determine whether greater awareness causes greater risk adversionagainst GM-foods, or vice versa. Different types of questions about awareness are usuallyassociated with different results. When people were asked to assess statements such as„Ordinary tomatoes do not contain genes while genetically modified ones do“ or „By eating agenetically modified fruit, a person’s genes could also be changed“responses did not suggestthat awareness or knowledge of biotech issues was generally higher in Europe than in NorthAmerica. Moreover, the level of awareness does not covary with the level of support forGMOs. Negative perceptions of GM-foods may also be explained by a lack of benefitsconsumers associate with these products. In Europe, the percentage of people perceiving no

51

consumer benefits in GM-food is particularly high (Figure x). Also in this case, however, it isimpossible to determine whether benefit perceptions affect risk adversion or vice versa.71 Kamaldeen/Powell 2000; Dittus, K. L and Hillers, V. N. 1993. „Consumer Trust andBehaviour Related to Pesticides“, Food Technology, 477: 87-89; van Ravenswaay. E. O.1995. „Public Pereceptions of Agrichemicals“, Council for Agricultural Science andTechnology Task Force. Ames, Iowa. CAST 1995.72 The Eurobarometer data show that between 16 and 26% of the respondents trusted thatconsumer organizations, environmental organizations, and the medical profession were tellingthe truth about biotechnology. The percentage trusting universities, the government or themass media ranged from 4 to 9%. See Nature 1997, 387: 831-836; Eurobarometer 1996, 1999;Einsiedel, E. F. 1997. Biotechnology and the Canadian Public: Report on a 1997 NationalSurvey and some International Comparisons. University of Calgary, Alberta; Braun 1999.73 Kamaldeen/Powell 2000; Hoban, Thomas J. 1998. “Trends in Consumer Attitudes aboutAgricultural Biotechnology”, AgBioForum 1(1): 3-7; Einsiedel 1997. While this argument isvery plausible, it is difficult to find empirical evidence directly supporting the causal claim. Inparticular, it is impossible to demonstrate whether adversion against biotechnology causeslow trust in regulators, or whether the causality runs in the opposite direction.74 Vos 200075 Vos 2000; Committee of Independent Experts 1999. „First Report on Allegations RegardingFraud, Mismanagement and Nepotism in the European Commission“, March 15.76 This new body strongly resembles the European Agency for the Evaluation of MedicinalProducts. The latter assists the Commission in the authorization of pharmaceuticals.77 BSE was first detected in British cattle in 1982, but the EU accepted British assurances thatthere was no risk to humans. In the late 1980s, British authorities notified EU member statesthat their food safety might be at risk. In 1989–1990 a massive outbreak of BSE occurred andprompted the EU into banning human consumption of BSE infected meet. Growing publicconcerns over food safety were still largely ignored by the British government and the EU.The scandal finally broke when, in 1996, the British government announced the occurrence often cases of Creutzfeld-Jakob disease and related these cases to the consumption of BSEcontaminated meet.78 See Jasanoff 1997; Hoban, Thomas 1997. „Trends in Consumer Attitudes aboutBiotechnology“, Journal of Food Distribution Research(27)1: 1-10; Eurobarometer 1999; Vos2000; Braun 2000; Powell 1998; Lok, C. and D. Powell 2000. „The Belgian Dioxin Crisis ofthe Summer 1999: A Case Study in Crisis Communications and Management“, TechnicalReport, February, www.plant.uoguelph.ca/safefood; Nelkin, Dorothy 1992. Controversy:Politics of Technical Decisions. Newbury Park, Sage Publications.79 See Joly, Pierre-Benoit and Stéphane Lemarié 1998. “Industry Consolidation, PublicAttitude and the Future of Plant Biotechnology in Europe”, AgBioForum (1)2: 85-90: 88.Kalaitzandonakes and Marks found (by means of a content analysis) that unfavorablereporting on agrobiotechnology in The Daily Telegraph (UK) increased significantly due tothe BSE crisis. Kalaitzandonakes, Nicholas and Leonie A. Marks 1999: “Public Opinion ofAgBiotech in the US and UK: A Content Analysis Approach”, in AAEA National Meetings.They also showed that the coverage of agbiotech in The Daily Telegraph increased fromaround 30 articles per year to almost 90 in 1998. In the same time-period (1996-1998) thepercentage of respondents strongly opposing GM-foods increased from 29 to 40%.80 Caswell, Julie A. 2000. „Labeling Policy for GMO's: To Each his Own?”,AgBioForum(3)1: 305-309, http://www.agbioforum.org.81 Survey data show that media coverage and opposition to biotechnology have been morepronounced in countries where public perceptions of biotechnology are more negative, e.g. inGermany, Austria, Sweden, and Denmark (Hoban 1996, 1998). This result does not allow for

52

determination of the direction of the relationship – from media coverage and NGO campaignsto negative public perceptions, or vice versa. Surveys found that anti-GMO views werestronger in Europe already before GM-crops were approved by the EU, thus questioning theinfluence of NGOs on public option. In a 1995 survey 21% of the respondents in the UnitedStates, 85% in Sweden, 60% in Austria, 57% in Germany, 48% in the Netherlands, 39% in theUK, and 38% in France answered that genetic engineering constituted a serious health hazard(Hoban, Thomas 1997. „Consumer Acceptance of Biotechnology: An InternationalPerspective“, Nature Biology, March). The latter result may suggest that NGOs weresuccessful in framing the GMO agenda even before food scandals drove the general publiceven more strongly in an anti-GMO direction. It may also suggest, however, that NGOs havehad rather limited impact on public risk perceptions, for the most substantial anti-GMOcampaigns by NGOs have taken place only from the mid-1990s onward. Assessing theinfluence of NGOs on public opinion is difficult also because, in 1996, the year when RRsoybeans and Bt-corn received EU approval and NGOs began massive campaigns, the BSEcrisis broke with the British announcement that BSE contaminated meet was the cause ofCreuzfeld-Jakob cases.82 Patterson, Lee Ann 2000. „Biotechnology Policy: Regulating Risks and RiskingRegulation“, in Wallace, Helen and William Wallace (eds.) Policy-Making in the EuropeanUnion, Oxford University Press, Oxford: 317-343;83 See Alasdair R. Young, European Consumer Groups, and Ruth Webster, EnvironmentalCollective Action, both in Justin Greenwood and Mark Aspinwall, Collective Action in theEuropean Union, London: Routledge, for a more generic analysis of consumer andenvironmental groups at the EU level.84 BEUC 1999. “Annual Report 1998“; BEUC 1999: “GM Food and the EuropeanConsumer”, The 8th Annual European Food Law Conference, June 23; Egdell, Janet M. andKenneth J. Thomson 1999. “The Influence of UK NGOs on the Common AgriculturalPolicy.” Journal of Common Market Studies (37)1:121-131: 124; www. greenpeace.org;www.foe.org; www.beuc.org.85 The Novel Foods Regulation, for example, was discussed in an „Expert Working Group“(member states only), the Foodstuff Advisory Committee (agricultural interests – COPA,retailers – EuroCommerce, producers – CIAA, consumers – BEUC, and trade interests), andthe Consumer Consultative Committee (Behrens 1997).86 See for example Patterson 2000 and Vos 2000.87 By the mid-1980s, France and the United Kingdom in particular had established systems ofmonitored self-regulation. The EU’s Directorate General for Science, Research, andDevelopment advocated a product-oriented and strongly science-based approach, similar tothe one used in the United States. In contrast, Directive 90/220 is clearly modeled afterDenmark’s Environment and Gene Technology Act, which was adopted in 1986 and isprocess-oriented (Patterson 2000: 321). Patterson (2000: 328) notes that „what distinguishedthe debate about biotechnology regulatory policy from other examples of bureaucratic politicswas that there was little room for compromise, trade-offs, and side-payments because of theexistence of very strong and widely divergent world-views about the potential harm thatbiotechnology posed for humankind and the environment in general.“88 Biliouri, Daphne 1999. “Environmental NGOs in Brussels: How Powerful are TheirLobbying Activities?” Environmental Politics (8 )2: 173-182: 177; Barling, David 1995. “TheEuropean Community and the Legislating of the Application and Products of GeneticModification Technology”, Environmental Politics (4 )3: 467-474: 473; Alasdair R. Young,European Consumer Groups.89 See Barling 1995: 469; Patterson 2000; Behrens 1997; Jany/Greiner 1998; Gately 1997; andChege 1998. “Compulsory Labeling of Food Produced from Genetically Modified Soya

53

Beans and Maize”, Colombia Journal of European Law (4)Winter/Spring:179- 181. In 1989and 1990, NGOs heavily lobbied the European Parliament to eliminate the many GM-products exempted in the Commission’s draft for Directive 90/220. Forceful opposition by theEP’s Committee on the Environment, Public Health, and Consumer Protection led to majorrevisions in the draft. In the Directive eventually adopted, all products containing GMOs mustmeet the EU’s relevant product legislation and also process-based, environmental risk-assessment standards (Hodgson 1992; Patterson 2000). NGO lobbying has also motivated theEuropean Parliament, which had been divided in earlier readings of the Novel FoodsRegulation, to pass a Biosafety Resolution with a large majority, requesting mandatorylabeling for all foods containing GM-soy, and to push for further regulation of GMOs(Behrens 1997).90 In more general terms, the sheer complexity and the associated openness of the EU’sregulatory process to the influence of public interest groups is exemplified by the approval ofGM-soy and GM-corn in 1996 under Directive 90/220. The Commission first voted in favorof approval. The Competent Authorities in Belgium, Finland, France, Ireland, Spain andPortugal agreed with the Commission, those in Austria, Sweden, Denmark and the UnitedKingdom voted against, those of the other member states abstained. Because of thisopposition, the issue was referred to the Environmental Council. Lack of unanimity thereprevented the Council from amending the Commission’s proposal. The issue thus went backto the Commission, which consulted three scientific committees and then authorized the twoGM-crops (Bradley 1998 in Patterson; Patterson 2000: 337-338).91 According to the Novel Foods Regulation, which entered into force in May 1997, labelingwas mandatory only for products containing live GMOs or GMO-derived products that werenot „equivalent“ to conventional products. GM-soy and corn were judged to be equivalent andthus no label was required by the Commission.92 For example, Denmark required full disclosure of any GMO ingredient, whereas the UnitedKingdom did not require labeling of substantially equivalent foods.93 For example, Austria, Italy, and Luxembourg invoked the safeguard clause in 90/220 to banthe import of Bt corn into their countries. In another case, the French government firstprohibited the cultivation of Bt-corn, although it had been approved by the EU. In 1998, itapproved the planting of Bt corn, but under opposition from Greenpeace and other NGOsreversed its decision.94 Greenwood/Ronit 1995; Patterson 2000:329.95 Patterson 2000.96 Smaller biotech firms coordinated their efforts through the European Secretariat of NationalBio-Industry Associations (ESNB). In September 1996, SAGB and ESNBA merged.97 Scharpf (1985, 1995) in particular has provided a systematic explanation for thisphenomenon, which he termed “Politikverflechtungsfalle” (policy integration trap).98 Assouline, Gérald 1996. “European Industry Strategies in Biotechnology”, Biotechnologyand Development Monitor (26): 10-12; Chen/McDermott 1998: 543.99 See, e.g., The Economist 1999, February 20:75-77, 1999, June 19:15-16.100 Jasanoff 1997; Science 1998, August 7: 768-771.101 See, e.g., www.greenpeace.org, www.foe.org.102 Caduff 2001. In 1999, the United States imposed retaliatory tariffs on several EU countries.103 Harlander, Susan 1991. “Social, Moral, and Ethical Issues in Food Biotechnology”, FoodTechnology (45)5:152-161; Covello, Vincent T. et al. 1991. “Guidelines for CommunicatingInformation about Chemical Risks Effectively and Responsibly”in Mayo, Deborah G. andRachelle D. Hollander (eds.). Acceptable Evidence: Science and Values in Risk Management.New York: Oxford University Press: 95-118.

54

104 Monsanto, for example, invested more than two million USD in a campaign in the UnitedKingdom and France, only to find out that negative opinions about GMO increased from 44 to51% in that time-period (Financial Times 1997, October 23: 3; Wall Street Journal 1999, May11: A1, A10; Bernauer 2000 in NZZ; New York Times 2001, January 25 and Business Week,June 12, 2000 for analyses of the Monsanto case).105 EuropaBio, for example, stated that it “[...] supports the right of the consumer to choosefood products with or without GM (genetically modified) ingredients. Consumer choice mustbe supported by appropriate transparent legislation, i.e. clear guidelines and criteria forlabelling” (EuropaBio 1999).106 Chen 1998: 530, 543; Lynch, Diahanna and David Vogel 2000. „Apples and Oranges:Comparing the Regulation of Genetically Modified Food in Europe and the United States“,paper prepared for the Annual Meeting of the Ameican Political Science Association,Marriot-Wardman Park, August 31-September 3.107 ACGA 2000. „Corn Growers Question Need to Sacrifice Export Markets Due toGenetically Modified Crops”, www.acga.org; Feedstuffs 1999. “Attention on GM Grains,Oilseeds Becomes Issue of Acceptance”, October 4. For example, between 1996 and 1999,the EU’s soybean imports even rose from 14.3 million MT to 16.6 million MT. In that sametime-period, the EU’s corn imports rose from 9.6 million MT to 11 million MT. See FAOstat2001, www.appsl.fao.org; Dann, Christine 2000. “From Boom to Bust in Three Seasons –The Rapid Rise and Fall of GE Markets”, www.connectotel.com/gmfood;108 Vogel (2001) claims that the EU has had no economic interest in discouraging farmersfrom using GM-seeds because these lower costs rather than increase output. This argument isdisputed among agricultural economists.109 The regulatory capture argument would predict that France in particular should have beenvery hostile to domestic production of GM-crops. The evidence shows, however, that thegovernments and important farmer interests in France and the UK were among the strongestsupporters of agricultural biotechnology. Only under pressure from consumer interest groupsdid they reverse their position.110 See Egdell 1999:123; Cummins 1998; www.cpe.org; Agence Europe 1997, January 16.111 Vos 2000.112 In May 2000, for example, it became public that 4'700 hectares of canola from GM-contaminated seeds imported from Canada had been planted in the United Kingdom. NGOs,such as Greenpeace, requested that the seed company be sued for violating EU and Britishregulations on GMO release, and that all crops affected be destroyed. See Financial Times2000, May 29: 6; ENDS Environment Daily 2000, May 29; New York Times 2000, May 19.113 Foster 2001.114 See, e.g. the statements by the European Confederation of the Food and Drink Industries.See also Erik Millstone, Analysing Biotechnology’s Traumas, New Genetics and Society,Vol.19, No.2, 2000:117-132. Millstone shows that only when mandatory labeling legislationwas on the horizon did downstream producers more systematically consult consumers, e.g.through focus group discussions. When they found out that consumers, when offered a choice,would quite invariably purchase non-GM foods, they turned to non-GM supplies rather thanlabeled GM-products. The policies of food manufacturers and retailers are compiled atwww.greenpeace.org.115 The controversy among different types of producers resulting from changing interests ofdownstream producers is well reflected in Joyce A. Nettleton. 1999. Food industry retreatsfrom science. Institute of Food Technologists, September 9, 1999.116 Millstone 2000:120.117 See European Report 1997. “Transgenic Maize: Green MEPS condemn Companies forRefusing to Label GMO Presence”, December 13; and Food Labeling News 1998. “Future of

55

GM Food in Europe Said to Rest with Food Industry Persuasion”, July 1. Nestlé stated that“although in our opinion there is no justification for such labelling in terms of either health orscientific concerns, we recognise consumers’ desire for such information and therefore,although we are not legally required to do so, we will indicate whether a productcontains genetically-engineered ingredients.”118 Food Labeling News 1998.119 RAFI Communique 1998. “Seed Industry Consolidation: Who Owns Whom?”.July/August, http://www.rafi.org/communique.120 www.ciaa.be; Edgell 1999: 124; www.transgen.de.121 Reuters, August 3, 2000.122 See, e.g., Financial Times 1998, January 8: 2, 1998, March 18; Wall Street Journal 1999,May 11: A1, A10.123 A listing of individual downstream producer’s labeling policy can be found at:www.greenpeace.org/~geneng/reports/food/food004.htm.124 The concerns of wholesalers and retailers over problems in harmonizing approval andlabeling regulations are reflected in many statements by EuroCommerce, the EuropeanAssociation for Retail, Wholesale and International Trade (www.EuroCommerce.be). InJanuary 1998, for example, European food producers declared that they would begin to labelGM-foods voluntarily to address consumer fears (Financial Times 1998, January 8: 2). Thisexample suggests that food processors, wholesalers, and retailers have reacted more toconsumer concerns than to regulation.125 www.transgen.de; www.greenpeace.org.126 See also Millstone 2000:121.127 See Millstone 2000.128 Growing premiums for non GM-soybeans (around 5-25 cts/bushel in 1999/2000)demonstrate that world demand for non-GM soybeans exceeds the supply (seehttp://www.tge.or.jp/english/market/chart.e/inputpram_m.e.html). While large downstreamproducers have found it easier to use their market power to purchase GM-free soybeans,smaller retailers in particular have often been found to sell GM-soy contaminated productswithout labeling. The BSE crisis and the EU’s ban on animal proteine feed have compoundedthese problems, because demand for soy feed has risen sharply. The ban on animal proteinefeed implies that around 1.32 million metric tons of animal proteine needs to be replaced bythe same amount of vegetarian proteine. This is equivalent to 3.7 million tons of soybeans, or6.2 million tons of sunflowers or 6.8 million tons of canola. Growing scarcity of non-GM soymay also have motivated downstream producers to pressure farmers on the issue: becauserising premiums for non-GM soy cut directly into downstream producers’ profits, motivatingfarmers to increase the supply of (and reduce premiums for) non-GM-soy is of direct interestto downstream producers. Once downstream producers have switched to non-GM products,they can, particularly for reasons of reputation, not reverse their policy.129 As of March 2001, 11 varieties of GM-corn were approved in the United States, and only 4in the EU.130 The EU’s corn imports from the United States decreased from around 2.7 million tons in1996/97 to 1.3 in 1998/99 to around 0.8 in 1999/2000. For 2001 a decline to almost zero wasexpected.131 www.transgen.de.132 Since 2000, many large food processors, wholesalers, and retailers in the EU have movedeven further against GMOs by restricting their meet supply to animals raised on non-GMOfeed. Even McDonalds UK and major retailers in the United Kingdom, so far the EU countrywith the most GMO friendly policy, have changed sides, making it ever more likely thatEuropean markets are definitively turning against GMOs (Organic Newsline 2001, February

56

1). This development is also crucial because the EU’s GM-soybean imports are almostexclusively used for animal feed. After US corn exports to Europe have collapsed, soybeanexports may thus follow the same trajectory.133 www.transgen.de maintains a database of such GMOs on the EU market.134 Economic theories of regulation have focused largely on issues of low saliency and highcomplexity that are often associated with „boardroom politics“ and regulatory capture (seeGormley 1986).135 Large and vertically integrated downstream producers in mass product markets (such as thefood market) can, in principle, benefit from increasing product differentiation and labeling atthe expense of smaller producers – large producers can controll their entire supply-chain moreeffectively and at lower cost. See Foster 2001. In practice, large downstream producers areunlikely to embark on such a strategy. In addition to NGO, farmer, and consumer opposition,larger downstream producers would also encounter opposition from smaller downstreamproducers. Under these circumstances, large upstream producers would, among themselves,face an important collective action problem. Each firm would want to wait for the others tomove first and bear the costs of re-establishing the GM-food market and follow as a free-rideronce the market has re-opened.136 www.bio.org; www.transgen.de; www.isaaa.org.137 Jasanoff, Sheila 1995. “Product, Process, or Programme: Three Cultures and theRegulation of Biotechnology”, in Bauer, Martin (ed.). Resistance to New Technology.Cambridge: Cambridge University Press. See also Kraus 1996; Bud 1995.138 In 1984, the EPA had proposed to regulate biotechnology under the Federal Insecticide,Fungicide and Rodenticide Act and the Toxic Substances Control Act.139 Holla, Roger 1995. “Safety Issues in Genetic Engineering: Regulation in the United Statesand the European Communities”, in Fransman, Martin et. al. The Biotechnology Revolution?Oxford: Blackwell: 414-415.140 www.bio.org; Jasanoff 1995:314.141 Chen/Dermott 1998; Moore 2000, „Food Safety, Labeling and the Role of Science“:Annex. Government&Policy, Vol.78, No.21, May 22, 2000.142 National Research Council. 1989. Field Testing Genetically Modified Organisms –Framework for Decisions. This report stated that „...the product of genetic modification andselection constitutes the primary basis for decisions...and not the process by which the productwas obtained...“ It confirmed conclusions of a 1987 report by the National Research Council.143 Jasanoff 1995: 315.144 These crops were chosen because they had undergone the most extensive field-testing, andbecause none of these crops had wild relatives in the United States.145 One indicator for this relaxation of approval practices is, for example, that the number offield releases increased from 8 in 1987 to 1105 in 1998 (Beach 1998; Lemaux 1998).146 Federal Register, May 29, 1992. In its 1992 policy statement, i.e. its interpretation of theUS Federal Food, Drug, and Cosmetic Act in regard to foods derived from new plantvarieties, FDA stated that it would require GM-foods to meet the same rigorous standards asall other foods. It also announced that it would treat substances intentionally added to foodthrough genetic engineering as food additives if they were significantly different in structure,function, or amount from substances currently in the food. (FDA Backgrounder 1994, May18) The FDA thus ignored a 1991 proposal by the Environmental Defense Fund (nowEnvironmental Defense) to apply the Food Additives Amendment to the Food, Drug, andCosmetic Act to all biotech foods.147 The FLAVR SAVRTM tomato is designed to remain on the vine longer before being picked(it softens slower). In 1994, Calgene, the producer of the tomato, engaged in consultationswith the FDA’s Center for Food Safety and Applied Nutrition (CFSAN). Based on an

57

evaluation of the data provided by Calgene, FDA determined that the tomato had not beensignificantly altered compared to other varieties of tomato. It was thus declared as safe as anyother commonly consumed tomato. (FDA Backgrounder 1994, May 18; Belinda Martineau,Food Fight: The short, unhappy life of the Flavr Savr tomato, The Sciences, Published by theNew York Academy of Sciences, Spring 2001:24-29)148 See FDA. 1994. Backgrounder, May 18; Krimsky, Sheldon and Roger P. Wrubel 1996.Agricultural Biotechnology and the Environment: Science, Policy, and Social Issues. Urbananand Chicago, University of Illinois Press: 106-107. Technically, the FDA does not considernew genes and their protein products to be food additives. If regulated as food additives, suchproducts would fall under section 409 of the Food, Drug and Cosmetic Act, which wouldforce the FDA to use pre-market approval processes. In that case, producers would bear theresponsibility (and costs) of evaluating the new products’ safety.149 www.bio.org; Caswell 1998: 23-24; Moore 2000, „Food Safety, Labeling, and the Role ofScience“, Annex; Marshall 1998: 35; National Center for Biotechnology Education (NCBE)1999. “US Regulations”, www.ncbe.reading.ac.uk/NCBE/GMFOOD.150 For example: Higher than normal levels of lycopene, which can lower the risk of cancer orheart disease, would imply a change in composition. Higher levels of vitamin C would implya change in nutritional value. The presence of a protein from peanuts in GM-food might implythe presence of a peanut allergen, and thus a change in safety.151 International Food Information Council (IFIC) 1999. “Backgrounder – FoodBiotechnology”, www.ificinfo.health.org.152 www.bio.org.153 Exceptions include, for example, high laurate canola and high oleic acid soybeans. In thosecases, the FDA confirmed the safety of these products but required labeling to inform aboutthe new ingredients.154 Jasanoff 1995: 316-317.155 Council for Agricultural Science and Technology. 1998. „The Proposed EPA PlantPesticide Rule“, October, Issue 10; Lemaux 1998;156 Nature 1999, May.157 New York Times 2000, January 17.158 Bonior, David. 1999. “Letter to FDA”, www.purefood.org/ge/cong49label.cfm.; TheColumbus Dispatch 1997, February 27; Center for Science in the Public Interest 1998; NCBE1999; Hoban 2000.159 Bonior 1999.160 www.foodsafetynow.org.161 These bills were introduced by Dennis Kucinich (D-Ohio), Patrick Moynihan (D-NewYork), and Barbara Boxer (D-California). See www.thecampaign.org.162 A compilation of past and ongoing legal action on GMO issues is maintained by the Centerfor Food Safety (www.centerforfoodsafety.org). Most lawsuits have been launched against theFDA, EPA, and individual biotech companies. So far, the FDA’s policy on GMOs has beenfully protected by court decisions (see, e.g., www.centerforfoodsafety.org, October 3, 2000;New York Times, October 4, 2000).163 In November and December 1999, the FDA held public hearings on whether or not GMOsshould be defined as food additives, which would require mandatory labeling, and on whetheror not more stringent testing procedures should be required to ensure consumer safety. (FDA,Report on Consumer Focus Groups on Biotechnology, 2000, October 20; Neegaard 1999) Aninfluencial study by the National Academy of Science, published in May 2000, supported inprinciple the FDA’s product-focus. However, it proposed three additional measures: long-term studies on risks posed by the consumption of GM-foods containing substances other thanthose already on the market; the adoption of the EPA’s 1994 proposal for regulations on pest

58

resistant GM-plants; regulation of GM-crops designed to resist viruses. The January 2001FDA proposal reflects the proposals of the NAS rather than the results of the FDA’s publichearings.164 Chen/McDermott 1998: 531. If formally adopted, food or feed developers will have tonotify the FDA at least 120 days before a GM-food or animal feed is marketed. They wouldalso have to provide information demonstrating that the product is as safe as its conventionalcounterpart. This measure would transform the existing voluntary consultation procedure intoa mandatory and more transparent one.165 GM foods on the US market are tracked, for example, by the Union of ConcernedScientists (www.ucsusa.org).166 www.fda.gov. The FDA’s proposal may be regarded as an attempt to cope with growingdiversity in voluntary labeling practices. Driven by consumer concerns and pressure byenvironmental/consumer groups, several large US food processors (for example Frito-Lay,McDonalds, Heinz, Gerber, and McCain Foods) have reduced or eliminated GM-foods fromtheir supply (New York Times 2000, June 3: 1, 25; Washington Post Weekly 2000, February14: 18). The USDA stated that it would subject the Department’s biotech approval process toindependent scientific review. It would promote the standardization of tests for detectingsmall amounts of GM-soybeans, corns, and other grains. The aim of this initiative is tosupport food-processing firms that wish to use GMO-free products or introduce voluntaryGMO-labeling (ICTSD 1999, July 19).167 Moore 2000b. Turning points are marked by the cloning of a sheep in 1997 and by thediscovery in 1999 that Bt-corn may not kill not only pests but also the Monarch Butterfly(Nature 1999, May). The latter quickly became a symbol for the environmental risksassociated with GM-crops.168 These include the Biosafety Committee review according to US National Institutes ofHealth Guidelines; USDA determination of non-regulatory status (permission to grow, test,and use for traditional crop breeding without further USDA action); EPA experimental usepermit approval; EPA determination of food tolerance or exemption from tolerance; EPAproduct registration; FDA review process. (See www.consumeralert.org: Regulation ofAgricultural Biotechnology in the United States: How the Process Works)169 Moore 2000, “Food Safety, Labeling and the Role of Science”; Moore, Elizabeth 2000:“Science, Internationalization, and Policy Networks: Regulating Genetically-Engineered FoodCrops in Canada and the United States, 1973-1993. Department of Political Science; FDA.1999. “Biotechnology in the Year 2000 and Beyond: Public Meeting”, December 13,www.fda.gov/oc/biotech/default.htm.170 Moore 2000, „Science, Internationalization, and Policy Networks“ .171 Environmental Defense Fund (EDF). 1999. Statement of Rebecca Goldburg,Environmental Defense Fund, At the FDA Public Hearing on Genetically Engineered Foods,November 30; EDF 2000; UCS 2000.172 www.cspi.org.173 E.g. Jeremy Rifkin from the Foundation on Economic Trends.174 E.g., social scientists have not yet examined whether public support for mandatory labelinghas preceded or followed NGO campaigns for mandatory labeling.175 Only three percent of respondents mentioned biotechnology or GM-foods when asked whatfood safety issues they were most concerned about (GMA, October 2000). No one mentionedbiotechnology or GM-foods when asked what, over the past few months, they had beenavoiding or eating less of. Only five percent stated that, during the past few months, they haddone something or taken any action because of concerns over GM-foods. (GMA 2000,October).

59

176 Survey results reported in Priest (2000) and Hoban/Miller (1998) show that between 1998and 2000 support for human genetic screening increased from 70 to 84 percent, for newhuman medicines it declined from 80 to 79 percent, for insect protected crops it increasedfrom 66 to 71 percent, for improved foods it increased from 58 to 60 percent (percentage ofrepondents agreeing that the respective application should be encouraged). The percentage ofrespondents stating that particular biotech applications were unacceptable was 21 percant forhuman medicines, 27 for insect protected crops, 37 for lower fat foods, 39 for disease resistantanimals, and 57 for faster growing fish. (Angus Reid. 2000, October) General consumeracceptance of plant biotechnology has remained approximately constant and high compared toEurope: survey results from 1992, 1994, and 1998 show constant support from 70 perdent ofsurveyed Americans (Hoban 1998: 4). The results of the FDA’s public hearings in late 1999provide a similar picture, though they highlight growing public concerns over GM-foods(FDA. 2000. „Report on Consumer Focus Groups on Biotechnology“,, October 20). See alsothe Wirthlin Worldwide Survey on food biotechnology, www.ificinfo.health.org. Surveysreported in Nature Biotechnology, Vol.18, Sept. 2000:939-942 and a compilation of varioussurvey results by the Center for Food Safety (www.centerforfoodsafety.org) suggest slightlyeroding consumer support for GM-foods and declining trust in US regulators.177 These public perceptions, which differ from those in Europe, are difficult to explain.Awareness levels, for example, cannot account for consumer acceptance, notably becauseawareness indicators vary enormously depending on the question posed, making it almostimpossible to claim that US consumers are generally less aware of GMOs than their Europeancounterparts. (Hoban 2000; Hoban 1998; Kalaitzandonakes/Marks 1999; Wirthlin GroupQuorum Surveys; www.ificinfo.health.org/)178 Hoban 1994; Eurobarometer 1996, 1999; Enriquez, Juan and Ray Goldberg 2000.Transforming Life, Transforming Business: The Life Science Revolution“, Harvard BusinessReview, March-April.179 Hoban 2000; www.cspi.org, survey in May 2001.180 In constrast to the FDA’s current policy, consumers seem to be more interested in labelingof how a food was produced than in the compositional effect of the process on the foodproduct. This preference seems to be driven by concerns about unknown long-term effects. Italso reflects consumers’ desire to be given an opportunity to register their views about thewisdom of food biotechnology, and the disappointment that food biotechnology has becomeso pervasive in the US food supply without consumers knowing much about this. (FDA.2000, October 20). Public survey data on labeling provide a somewhat contradictory picture.In several surveys, a majority of US consumers (around 60 to 90 percent) have stated theirpreference for labeling of GM-foods. (Bonior 1999; Novartis survey, cited in The ColumbusDispatch 1997, February 27; www.cspi.org; NCBE 1999). At the same time, the intensity ofthose preferences and the willingness of consumers to pay for labeling appear to be low(www.cspi.org, survey conducted in May 2001). Moreover, surveys have found that theFDA’s labeling policy (which requires no mandatory labeling of GM-foods) enjoys largesupport among consumers (Hoban 1998: 5).181 Fortune 2001. „The Gene Food Segregation Fantasy“, February 19; New York Times,October 14 and December 11, 2000.182 The primary reason for this failure is said to be incomplete control by Aventis and Garst(the key distributor of the seeds) over seed companies, farmers, and downstream producers.Some analysts suspect that Aventis speculated on the approval of StarLink for humanconsumption – its application was pending – and thus did not fully control segregation.(Fortune, February 19, 2001)183 Aventis licensed the StarLink technology to 11 seed companies. Garst Seed Co., located inIowa, was by far the largest. Approximately 315’000 acres were grown with StarLink seeds

60

by around 2’070 farmers. Around nine million bushels had entered commercial channelswhen the recall began. New YorkTimes, October 14, 2000; Houston Chronicle, October 22,2000; BioDemocracy News 2001, March , 2000, November; Washington Post, March 1,2001:A01.184 See, e.g., Biodemocracy News, Issue 32, March 2001 (www.organicconsumers.org);Washington Post, November 26, 2000:B02; New York Times, July 19, 2000;www.organicconsumers.org, May 25, 2000; BioDemocracy News 27, May 2000.185 Washington Post, February 13.186 New York Times, December 11, 2000; Los Angeles Times, January 7, 2001.187 Already in 1998, a coalition of NGOs filed a lawsuit against the FDA for failing to fulfillits regulatory duties. In March 2000, more than 50 consumer/environmental groups filed alegal petition with FDA requesting pre-market and environmental testing as well asmandatory labeling. These activities, as well as other NGO campaigns on GMO issues, havehad virtually no effect on the FDA’s policy.188 E.g., BioDemocracy News 32, March 2001 (www.purefood.org); Organic ConsumersAssociation, April 28, September 21, 2000.189 Gormley 1986.190 Holla 1995: 414-415; Moore 2000, „Science, Internationalization, and Policy Networks“.191 For an overview of firms in the agribusiness sector, see Agribusiness Examiner 2001,Supplement, Issue 1, January 15. See also Lehne 1995; Bijman 1995.192 In addition, plant biotechnology companies are also represented by the American SeedTrade Association (ASTA).193 Moore 2000, „Science, Internationalization, and Policy Networks“: 8-10; www.bio.org;Moore 2000b: 8-10; Rizzo 1999.194 Upstream producers have also expressed concern over voluntary GM-free labels becausethese could imply superior quality. BIO has supported the FDA’s labeling policy and lhasobbied heavily against an extension of this policy. It fears that labeling might stigmatize GM-food and encourage downstream producers to opt for GM-free food to avoid complicatedregulations, particularly when different markets impose different labeling requirements.195 FDA. 1999; www.bio.org; Caswell 2000.196 Moore 2000, „Science, Internationalization, and Policy Networks“: 37; Epstein 1996.197 Moore 2000, „Food Safety, Labeling and the Role of Science“: 14, 25, 37; IATP FoodSafety & Health 1999.198ibid: 22, 25.199 Reuters, July 6, 2000; www.whybiotech.com.200 Gormley 1986.201 See New York Times 2001, January 25.202 The Economist, November 18, 2000: 95, November 14, 2000: 93. In 2000, Burrill’s indexof medical biotech firms rose by more than 50 percent, whereas agbiotech firms experienced afall of more than 10 percent (Economist, 2000, November 4: 93; Business Week, November6, 2000).203 Growing opposition to GMOs has also motivated agbiotech firms to slow down themarketing of new products. For example, in October 1999 Monsanto declared that it wouldnot commercialize the so-called terminator gene, a controversial technique to make seedssterile (Kilman 1999; Food Chemical News 1999; Feedstuffs 1999; BioDemocracy News2001, January 31).204 www.transgen.de.205 Carpenter, Janet E. and Leonard P. Gianessi 2001. „Agricultural Biotechnology: UpdatedBenefit Estimates“, National Center for Food and Agricultural Policy, http://www.ncfap.org;Baker, Ken 1999. “Encouraging Innovation in European Agricultural Biotechnology”, Agro-

61

Food-Industry Hi-Tech, September/October 29-32. Insect resistant Bt corn varieties haveenabled farmers to control the European Corn Borer, an insect pest that is hard to controlthrough pesticides. Benefits include primarily increased yields and slightly reducedinsecticide use. Insect resistent Bt cotton varieties control the three most destructive pests(tobacco budworm, cotton bollworm, and pink bollworm). Benefits include reduced pesticideuse and increased yields. Herbicide resistant Bt cotton allows farmers to use a broaderspectrum of herbicides with smaller crop injury. Insect and virus resistant potato varietieshave been used to reduce insecticide use. The introduction of a highly effective conventionalinsecticide and problems of marketing GM-potatoes have limited their adoption. Herbicidetolerant soybeans alllow for more effective weed control at lower cost.206 See also Baker 1999; Dann 2000. Other data are discussed in James L. Foster. 2001.Globalization, New Technology and the Competitive Use of Regulation: EnvironmentalRegulation and the Global Food Market. April. Manuscript. See also www.ers.usda.gov(“Agricultural Outlook”).207 Baker 1999; www.rafi.org; www.isaaa.org; Carpenter/Gianessi 2001.208 Opposition to labeling is motivated by the costs of segregating GM- and non-GM-crops,and by the risk of GM-products being stigmatized through labels.209 FDA 1999; Moore 2000, „Science, Internationalization, and Policy Networks“ : 40;www.afbf.org.210 National Family Farm Coalition (NFFC) 1999. “Farmers' Declaration on GeneticEngineering in Agriculture”, www.nffc.net/bio1.htm; www.nffc.org; Moore 2000, “Science,Internationalization, and Policy Networks“.211 Moore 2000, „Science, Internationalization, and Policy Networks“: 14.212 ibid.213 Food Labeling News 1998. “Monsanto Changes Stand on Labeling Genetically ModifiedFood in EU”, May 13; National Corn Growers Association (NCGA) 2000. “Biotechnologyand Labeling”, www.ncga.com/04growers/issues/biotechlabeling.htm. Food Labeling News1998d; NCGA 2000; ASA 2000.214 US Grains Council 2000. “Public Policy Positions“, www.grains.org/about/policies.shtm.215 FDA. 1999. ACGA, together with consumer and environmental organizations such as theCenter for Food Safety, Consumers Union, Friends of the Earth, and the Institute forAgriculture and Trade Policy, has also signed the 1999 letter to the FDA by 49 Congressmenrequesting mandatory labeling (Mendelson 1999).216 www.acga.org.217 Moore 2000, „Science, Internationalization, and Policy Networks“: 38-39.218 ibid: 13, 39.219 US corn exports to Europe decreased from 2 million tons in marketing year 1997/98 to137,000 tons in 1999, while in the same time-period soybean exports dropped from 11 milliontons to 6 million tons (ACGA 2000; FAOstat 2001; Feedstuffs 1999). Figures from FAO-Stat,which are available until 1998, also show a sharp decline in US corn and soybean exports tothe EU from 1996 on. In comparison, US wheat exports to the EU increased in that time-period. Whereas 27 percent of all US soybean meal exports went to the EU in the 1997/98marketing year, the EU only accounted for 7 percent of all US soybean meal exports a yearlater (Feedstuffs 1999). The value of US exports decreased by almost half (from 2.1 bUSD to1.1 bUSD) between 1996 and 1999 (Dann 2000). In parallel to this development, Braziliannon-GM soy exports to the EU increased from 10.1 million tons in 1996 to 15.1 million tonsin 1998 (Dann 2000).220 www.isaaa.org; www.rafi.org; www.acga.org. An informal survey among 400 US farmersin January 2000 indicated that GM soy crops would drop by 15 percent and GM corn cropswould drop by 22-24 percent in 2000 (Dann 2000). This result was backed up by a US

62

Department of Agriculture survey indicating that GM corn crops would decrease by 20% andsoybeans by 6 percent in 2000 (Dann 2000).221 For example: a survey among 100 Midwest grain elevators, conducted in September 1999,found that only 8 percent of the elevators were segregating non-GM soybeans and only 11percent were segregating non-GM corn (Feedstuffs 1999).222 Communication with James Foster, June 3, 2001.223 Second-generation GM-foods are expected to have decreased allergenicity, lower fatcontent, and improved freshness and taste. Some biotech foods might also deliver medicines,e.g. vaccines or compounts that help prevent cancer (www.bio.org).224 Food Labeling News 1999; Moore 2000, „Food Safety, Labeling and the Role of Science“:7; FDA 1999; www.gma.org; www.nfpa.org.225 Moore 2000, „Food Safety, Labeling, and the Role of Science“; Washington Post, August15, 1999; www.sustain.org/biotech/news.226 Dann 2000; www.abf.org.227 www.cspi.org.228 Wall Street Journal, July 30, 1999.229 New York Times 2000, June 3: 1, 25.230 Fortune 2001, February 19; Reuters, November 27, 2000; www.cropchoice.com,September 29, 2000; www.innovasure.com.231 E.g. BioDemocracy News, January 31, 2001.232 Fortune 2001, February 19; Vorman 2000; Hoban 2000.233 www.cropchoice.com; New York Times, June 3, 2000.234 Gormley 1986.235 Chen/McDermott 1998.236 Tanaka, Masami 1991. „Government Policy and Biotechnology in Japan“, in Wilks,Stephen and Maurice Wright (eds.). The Promotion and Regulation of Industry. New York,St. Martin’s Press, 110- 131. For earlier regulatory activity in the biotech area, seeChen/McDermott 1998 and Brock, M.V. 1998. Biotechnology in Japan. London: Routledge.See also Brock 1989: 88.237 Commandeur 1995.238 Japan Times, January 16, 1999:1; OECD 1999; Kafka 1999.239 Reuters, September 30, 1999.240 ICTSD, Weekly Trade Digest, August 16, 1999; ISSHO, January 6, 2000; ReutersSeptember 30, 1999, April 26, 2000; Japan Times, August 7, 1999: 3, October 26, 1999: 3;August 5, 1999: 1, October 21, 1999: 3.241 Reuters World Report, April 26, 2000.242 Commandeur 1995. See also Japan Times January 27, 2000: 2, October 18, 1999: 18.243 Hoban, Thomas 1996. „How Japanese Consumers View Biotechnology“, FoodTechnology(50) 7: 85-88.244 Nature Biotechnology, Vol.18, September 2000: 945-947.245 Angus Reid Group 2000. „Significant Knowledge Gap in Debate over Modified Foods“,www.angusreid.com/Media/Content/.246 Nature Biotechnology, Vol.18, Sept.2000: 947.247 USDA Reports 1998.248 Japan Times August 7, 1999: 3.249 Reuters July 27, 1999; Japan Times, August 7, 1999.250 Financial Times, February 2, 2001; BBC Monitoring Service, February 21, 2001,December 27, 2000; Asia Intelligence Wire/The Yomiuri Shimbun, November 28, 2000.251 Heissler 1995.252 www.jba.or.jp; Commandeur 1995.

63

253 Heissler 1995; www.jba.or.jp.254 OECD 1999; Kafka 1999; Japan Times, July 16, 1999: 1.255 Bridges Weekly Trade News Digest, May 10, 1999, www.sustain.org/bulletins.256 E.g. Commandeur 1995.257 Reuters, September 1, 1999.258 Japan Times, August 27, 1999: 3; Reuters, August 23, 1999.259 The Japanese trading firm Marubeni Corp., for example, has collaborated with the US firmArcher Daniels in this regard. Most other Japanese importers, e.g. Takara Shuzo Co. and thetrading house Mitsubishi Corp., have also set up procedures for testing and certification ofGMO-free supplies.260 It is estimated that around half of Japan’s one million ton market for food soybeans isimported without segregation on the basis of GMO content. The other half is eitherdomestically grown or identity preserved imports.261 It is unclear, e.g., whether the Japanese crushing industry (e.g. producers of soy oil) willturn to non-GM soybeans. Many of its products, such as soy oil, are exempt from the newlabeling requirement. Of Japan’s total soybean market of 5.2 million tons in 1999, around 3.6million tons were crushed for oil and proteins. (Inform 10/11, November 1999: 1051)262 Indian Express, July 15, 1999; Japan Times, August 27, 1999.263 E.g., from 1996 to 1999, EU corn imports from the United States, the largest producer ofGM-corn, fell from 305 million USD to 1 million. Those of US soybeans decreased from 2.6billion USD to 1 billion (New York Times, March 14, 2000: A1, A10). The principal reasonsfor this drop are the refusal or inability of US producers to effectively segregate non-GM fromGM-crops.264 E.g. Robert Paarlberg. 2000. The Global Food Fight”, Foreign Affairs, May/June 2000,Vol.79, No.3; ICTSD, August 16, 1999; International Trade Reporter, June 30, 1999: 1089;Patterson 2000: 340.265 Nomination of Stuart Eizenstat To Be Deputy Treasury Secretary, Before SenateCommittee on Finance, 106th Congress, 1999. See also International Trade Reporter, June 30,1999: 1089.266 One GMO case has already entered the WTO system. In 2000, Thailand challenged animport ban by Egypt on Thai tuna. Egypt has claimed that this tuna is packed in soybean oilmade from GM-soybeans. Thailand argues that there is no legal basis for the ban because oflacking scientific evidence on health risks posed by GM-soybeans. This case has, so far,remained at the consultation level.267 Keller and Heckman LLP 2000. „Comments on Non-Tarrif Trade Barriers for ProcessedFoods and Beverages: Mandatory Bioengineered Food Labeling Requirements“, InvestigationNo.332-421, submitted to the US International Trade Commission.268 E.g. Caswell 2000; Runge, C.F. and Jackson L.A. 2000. „Labeling, Trade, and GeneticallyModified Organisms (GMOs): A Proposed Solution“, Journal of World Trade.269 Large and vertically integrated downstream producers in mass product markets (such as theGM-foods market) can, in principle, benefit from increasing product differentiation andlabeling at the expense of smaller producers – large producers can controll their entire supply-chain more effectively and at lower cost. In practice, large downstream producers are unlikelyto embark on a strategy of re-establishing markets for GM-foods. In addition to NGO, farmer,and consumer opposition, they would also encounter opposition from smaller downstreamproducers. Under these circumstances, large downstream producers would, amongthemselves, face an important collective action problem. Each firm would want to wait for theothers to move first and bear the costs of re-establishing the GM-food market and follow as afree-rider once the market has been re-opened.

64

270 In the case study on Japan we noted that the likelihood and pace of such a trend depends onat least three conditions. Additional controversies over food issues, similar to the StarLinkproblem, or larger food safety crises, such as BSE, would boost consumer demand for stricterGMO-regulations. Such problems, and/or the appearance on the Japanese market of GM-products with direct competitive effects on domestic producers (e.g. GM-rice) would result instronger farmer opposition to GM-foods. Finally, growing supply of non-GM soy by USproducers would motivate Japanese downstream producers to increase their use of non-GMOalternatives in products that currently do not require labeling. Such practices might, in turn,encourage Japanese regulators to tighten labeling and approval regulations.271 CODEX negotiations on this issue have been stalled because of fundamental EU-USdifferences over GMO-policies. See also Hans-W. Micklitz. 2000. International Regulation onHealth, Safety, and the Environment. Journal of Consumer Policy 23:3-24.