ARTICLE IN PRESS

0743-0167/$ - se

doi:10.1016/j.jru

Abbreviations

Jakob disease; F

GMO, Genetica

NGO, Non-gov

VOSL, Value of�CorrespondE-mail addr

Journal of Rural Studies 22 (2006) 29–37

www.elsevier.com/locate/jrurstud

Investigating GM risk perceptions: A survey of anti-GM andenvironmental campaign group members

Clare Hall�, Dominic Moran

Land Economy Research Group, SAC, West Mains Road, Edinburgh, EH9 3JG, UK

Abstract

This study investigates how members of anti-GM campaign groups and environment groups perceive the risks and benefits of

genetically modified (GM) technology in food and agriculture. The study targeted these groups as the most risk-averse sector of

society when considering GM technology. Survey respondents were asked to rank the current and future risks and benefits of GM

and to rank GM risks against other health risks. Respondents appear to be unconvinced by the claims that future GM technologies

will provide additional consumer (or environmental) benefit, since perceived future risks were ranked more highly than future

benefits. Results support the claim that there is an inverse relationship between perceived risk and perceived benefit. Results also

suggest that among the respondents there are differences of opinion regarding the degree of risk to health posed by the technology.

Women and people living in rural areas, on average, ranked risks more highly than men and people living in urban areas. This study

serves to fill a void in the understanding of the perceptions of anti-GM campaigners.

r 2005 Elsevier Ltd. All rights reserved.

1. Introduction

Surveys and polls continue to point to consumeropposition to genetically modified (GM) food (see forexample, DEFRA (2002); MORI (2003)). In manymedia reports particular emphasis has been placed onthe role of anti-GM campaign groups and environ-mental action groups such as Greenpeace and Friends ofthe Earth. It has been claimed that opposition existsbecause people perceive the technology to be risky tothemselves, their families and the environment, withoutoffering any consumer benefits. Further, there isexpected to be an inverse relationship between risksand benefits, and therefore, if consumers perceive GMtechnology as offering personal benefit (or environmen-

e front matter r 2005 Elsevier Ltd. All rights reserved.

rstud.2005.05.010

: ANOVA, Analysis of variance; CJD, Creutzfeldt-

SE, Farm-scale evaluation; GM, Genetically modified;

lly modified organism; NEP, New ecological paradigm;

ernmental organisation; SHS, Scottish household survey;

statistical life.

ing author. Fax: +44131 667 2601.

ess: clare.hall@sac.ac.uk (C. Hall).

tal or social benefit) their perception of the risks willdiminish.

Accordingly, the aim of this paper is to investigate thenature of GM risk perceptions held by anti-GMcampaigners and environment group members in Scot-land. The paper begins with a discussion of the nature ofconsumer concerns relating to GM foods. We considerthe role of campaigners in resistance to GM foods andthe nature of anti-GM activities. We also address the useof the New Ecological Paradigm (NEP) scale toinvestigate environmental values (Dunlap et al., 2000),and the extent to which this may influence attitudes toGM food. The literature review and description of thesurvey methodology follow the introduction. Theremainder of the paper contains the presentation anddiscussion of results, and conclusions.

1.1. GM technology and consumer concerns

Since the widespread introduction of GM food andcrops in some countries, a large number of studies haveinvestigated consumer attitudes to the technology.

ARTICLE IN PRESS

1See for example the report from GM-Nation? (http://www.

gmnation.org.uk/docs/gmnation_finalreport.pdf) which places at least

as much emphasis on the results from ten discussion groups involving a

total of 77 people as it does on the results from public meetings

attended by an estimated 20,000, and from 36,000 feedback forms.

C. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–3730

Previous studies have revealed a diversity of consumerconcerns relating to GM food, including unpredictablehealth risks (Lemkow, 1993; Olubobokun et al, 2001;Isaacs, 2001; Subrahmanyan and Cheng, 2000; Ver-durme et al., 2001a), environmental safety risks (Olu-bobokun et al., 2001; Isaacs, 2001), and the structure ofagri-business (Isaacs, 2001). There are fears that theremay be long-term environmental effects (Lemkow,1993), risks to future generations (Poortinga andPidgeon, 2003; Rosati and Saba, 2000), and long-termfood safety issues (Grove-White et al., 1997). Theselong-term effects are expected in many instances to belargely unpredictable.

1.2. Environmental attitudes and GM foods

As much of the resistance to GM foods is fromenvironmental groups, an understanding of environ-mental values of acceptors and non-acceptors of GMproducts may be useful (Isaacs, 2001). An individual’sviews about the environment may contribute to will-ingness (or unwillingness) to accept GM food. As GMfood is alleged to have both positive environmentaleffects (for example, reduced pesticide use), as well asnegative impacts (for example, impact on non-targetspecies), an ‘environmentalist’ may accept or reject GMtechnologies depending on which element of theirenvironmental concerns is stronger. The NEP scale(Dunlap et al., 2000) was developed to examineenvironmental values, and has also been used inprevious studies into consumer attitudes to the risks ofGM food (Poortinga and Pidgeon, 2003; Isaacs, 2001).

There have been a number of studies that haveemphasised the role of environmental action groups inthe anti-GM debate. Purdue (2000) examined the role ofanti-GM campaign groups in the UK in some depth.Although he refers to the role of small, single issueNGOs, such as Genetics Forum, he also emphasises thesignificance to the anti-GM debate of large membershipNGOs such as Greenpeace. Environmental organisa-tions are recognised as being predominant in thecampaign against genetic engineering of food, withgroups such as Friends of the Earth and Greenpeacetaking public positions against GM technology inagriculture (Reisner, 2001).

1.3. Anti-GM activities in Scotland

A range of anti-GM activities has taken place inScotland in recent years, underlining the significance ofanti-GM sentiment and the need to understand themotivations of those involved in such activities. First,there was a long-running anti-GM campaign againstcrop trials that took place in the Black Isle. As reportedin the Guardian ‘‘y under cover of darkness, someoneentered the field at Roskill Farm and ripped out some

five acres of plants. Andy five people were arrestedafter tearing up more plants’’ (Scott, 2002).

The Friends of the Earth Scotland website (www.foe-scotland.org.uk) details a number of ways in which itsmembers might consider turning their anti-GM senti-ments into action. The website calls on members to jointhe campaign for a GM-free Britain and suggests thatpeople email their local authority with this message. Thesite further suggests that members should support newGM laws to prevent GM contamination, sign a globalGM petition asking the WTO to throw out the UScomplaint on GM foods, make their area GM-free bycontacting their local authority, and express zerotolerance of GM food by challenging their supermarket.

Further, in 2002 a petition was submitted to theScottish Parliament Health and Community Carecommittee, demanding an end to the GM field trialson the grounds of potential health risks. This was signedby 6500 people.

In June 2003, many Scottish residents were alsoactively involved in the government-sponsored GM-Nation debate. As well as the official ‘tier one’ meetingorganised in Glasgow, there were many other localmeetings held throughout the country. Forty eightmeetings in 11 regions are listed in the GM-Nationfinal report as having taken place with 30 or moreattendees, at least 1440 people in total (Heller, 2003). Itis reasonable to suggest that people who could betermed ‘anti-GM’ campaigners because of their mem-bership of specific groups (environmental or GM) mayhave been involved in at least some of these activities.

Most previous studies have identified the concerns ofthe general population towards genetic engineering.Where this study differs is in the specific targeting ofanti-GM campaigners and environment group membersthat have been vocal in their opposition to adoption ofthe technology, as outlined by the discussion above.Specifically, the aim was to investigate, in depth, theattitudes of this group towards the risks and benefits ofGM food.

2. Methodology

We conducted a survey of anti-GM campaigners andenvironment group members, perhaps the most risk-averse group in society when addressing the issue ofGM. This section of society has been dismissed as beingan irrational minority and their views discounted,1 yetthere is currently a gap in the understanding of their

ARTICLE IN PRESSC. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–37 31

perceptions of risk relating to GM. We wanted toinvestigate in some depth the nature of the concerns ofthis group towards the technology in food andagriculture. Specifically, we wanted to determinewhether respondents displayed a consistent lexico-graphic stance in considering the possible technologicaltrade-offs, or whether it is possible to identify differentsegments within the group that might influence accept-ability.

The survey addressed a number of questions. First,how campaigners might rank the (perceived) current andfuture risks and benefits of GM technology in food.Second, whether anti-GM and environmental campaigngroup members could be segmented into groups withdifferent attitudes towards the human–environmentrelationship, and if so, whether this was correlated withattitudes to GM technology. Third, whether attitudestowards GM technology varied between different socio-demographic segments of our target group, as claimedby previous studies with the general population, andfinally, how respondents ranked the perceived riskinessof GM food relative to other health-related risks.

2.1. The survey

In part one of the survey, statements based on theNEP scale were used to investigate the degree to whicheach respondent held an ecocentric or anthropocentricworldview, and the extent to which there was a range ofattitudes among campaigners. The purpose of using theNEP scale was to discover whether respondents whosevalues relating to the human–environment relationshipwere strongly ecocentric, were more or less risk aversethan respondents who were weakly ecocentric.

Parts two and three of the survey presented respon-dents with 24 questions relating to applications of GMtechnology in food and crops. The questions addressedperceived current and future risks and benefits and weredrawn up after a review of the literature. Respondentswere asked to indicate on a five-point scale the extent ofthe current or future risk or benefit arising from thescenario presented by the question (based on their ownperceptions). Results provided mean values of currentand future risks and benefits.

The survey also investigated the perception of theriskiness of GM foods relative to other established risks,through the use of a risk ranking exercise. Using anexisting list of health risks such as cancer, road accidentsand variant CJD (Sykes, 2003), we constructed a ‘riskladder’ with the most risky activity at the top and theleast risky at the bottom. Items on the list were rankedaccording to how many deaths arose from each, in theUK in 2000 or 1999, depending on the risk. Respon-dents were not told why the risks were in that order,only that the most risky was at the top. They were askedto place GM food somewhere among the list. Responses

allowed us to gain an understanding of how peopleperceived the health risks of GM food relative to otherknown risks.

Socio-demographic information was used to investi-gate the extent to which personal details affectedvaluation of risks and benefits, as well as ranking ofrisk and environmental values.

3. Results and discussion

Sixteen organisations in Scotland were contacted in2003 and asked to help with the distribution of thesurvey to their members. Groups contacted includedlocal Friends of the Earth (FoE) and Greenpeacegroups, as well as specifically anti-GM groups such asLothian and Borders Biocheck, Highlands and IslandsGM and Munlochy GM Vigil. As discussed above, thedecision to include FoE and Greenpeace groups in thestudy was a valid one as these organisations have high-profile campaigns against GM technology. Five organi-sations agreed to help with distribution, 31% of thosecontacted. Total number of usable responses was 38.

When the decision to target anti-GM and environ-ment groups was taken it was assumed that the level ofknowledge and interest in the topic, by at least some ofthis population, would result in a good response rate.However, this was not the case. The degree of suspicionand scepticism of some of those targeted is evident fromthe comments received, as shown below:

‘‘We feel it is in our interests to not aid you in yourresearch. This is because; whatever your personalviews on GM, the research that goes back to thepolicy makers will be used for the interests ofpromoting and furthering ‘‘scientific research’’ andin the interests of profit. Detailed information on theperception of those against GM can and are easilyused to assist in greenwashing the GM industry’’.

‘‘They (other members of the group) didn’t see howfilling in the forms could benefit our cause, and were abit suspicious of what the SE might use them for’’.

‘‘How will the results be used?’’

‘‘When and how will the results be published, and willthe respondents be forwarded copies?’’

‘‘Considering GM nation has already taken place,what is the rationale for this study, and why is theScottish Executive involved?’’

‘‘As taxpayers, who have already supported GMNation, some people are curious over the level offunding attached to this survey?’’.

‘‘I am really sorry to see public funds (and your owntime) used in this way. The risks of the technology aresufficiently clear; the need for it sufficiently tenuous;and the commercial motivation sufficiently strong

ARTICLE IN PRESS

Table 1

Respondent details

Respondent details Percent

Gender

Female 47.4

Male 42.1

No response 10.5

Education

O grades, O levels etc 5.3

Highers, A levels etc 15.8

Undergraduate degree 26.3

Postgraduate degree 39.5

Other 13.2

Householda

No children 28.9

With children 55.3

Unknown 15.8

Place of residencea

Urban 68.4

Rural 21.1

No response 10.5

GM_Trial

Near trial 28.9

Not near trial 34.2

Don’t know 36.8

aThe variables ‘household’ and ‘place of residence’ were reclassified

to reduce the number of categories.

C. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–3732

that I am really distressed at this attempt to find waysto discredit those opposed to it’’.

These comments demonstrate a lack of trust ingovernment and industry, a recognised social issue notrestricted to the GM debate (see for example, Poortingaand Pidgeon, 2003). This distrust has, however, beenespecially noticeable in the GM debate and is particu-larly strong among the target population. It becomespertinent to policy makers when it leads to disengage-ment with the policy debate.

3.1. Respondents

The typical respondent was female, aged between 30and 49 yr old, with a postgraduate degree, and living inan urban area in a household with children (Table 1).Typically, respondents did not know whether or notthey were living near the site of a GM Farm ScaleEvaluation (FSE) trial. Although our aim was not totarget a general population sample, it is neverthelessinstructive to compare the demographics of our samplewith results from the Scottish Household Survey (SHS)(Dudleston et al., 2002). Fifty-five percent of householdsin our population contained children. This compares tothe SHS where just 27% of all households were familieswith children. The gender composition of our sample issimilar to the SHS. Our population contained 47%females and 42% males, with 11% unknown. The SHSresults show 52% females and 48% males. Finally, 66%of our population held an undergraduate or postgrad-uate degree. This compares to 14% of the SHSpopulation who held a first or higher degree. Clearly,respondents were more highly educated than the generalpopulation and more likely to be from a household withchildren.

3.2. New ecological paradigm scale

Results from the NEP scale were scored as follows:Lowest possible score (10) represented very stronganthropocentric; highest possible score (40) representedvery strong ecocentric. Statements used in the NEPincluded ‘human behaviour is damaging the environ-ment’, with which 89.5% of respondents stronglyagreed, and ‘humans are meant to have stewardshipover the rest of the environment’ with which 50%strongly disagreed. The average score from results was34, strong ecocentric. The majority of respondents couldbe classified as being ecocentric, based on the results ofthe NEP scale. Only 8% were neutral—i.e. neitheranthropocentric nor ecocentric, and none, unsurpris-ingly, were anthropocentric. However, the resultsrevealed ‘shades of green’ among the respondents, with13% classed as weak ecocentrics and 39.5% as bothstrong and very strong ecocentrics. This demonstrates

that the target population should not be viewed as anhomogeneous group of environmental protesters but adiverse group with different levels of concern about theenvironment.

The women in our target population were morestrongly ecocentric than men, with an average NEPscore of 36 as opposed to 32 (Table 2). Previous studiesusing the NEP scale have suggested that there is a rangeof socio-demographic variables that are significant indetermining a respondent’s environmental values. Dun-lap et al. (2000) found that environmental concern wascorrelated with variables such as age, education,political party, occupational sector and others. How-ever, none of the other socio-demographic variables inour study had any significant effect on the NEP score.

3.3. Risk and benefit questions

Recall that in part two of the survey there were sixrisk-related questions and six benefit-related questions,all concerned with the situation now. This was followedby part three that also featured six risk-related questionsand six benefit-related questions, but concerned with thesituation in the future. Responses to these questionswere on a scale numbered 1–5, from no risk (or benefit)to lots of risk (or benefit). Scores from these answersenabled an average score to be derived for four

ARTICLE IN PRESS

Table 2

Respondent variables and NEP score

Variable Mean values T-test statistic Degrees of freedom F statistic P-value

Gender 3.51 32 — 0.001a

Female Male

35.94 32.38

Place of residence �1.81 32 — 0.079

Urban Rural

33.69 36.13

Group membership 0.47 16 — 0.646

Anti-GM Environment

34.80 33.85

Household �1.17 30 — 0.251

Without children With children

33.18 34.67

Education — — 1.90 0.134

Near trial? — — 0.32 0.725

aSignificant at the 0.01 level.

C. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–37 33

categories (risks now and in the future, and benefits,now and in the future). Overall results from the risk andbenefit questions revealed that, of the four categories,the highest average score was for future risks (3.86),followed by current risks (3.79) and then current benefits(2.30). Lowest of all were future benefits (1.67). Theseresults correspond with previous findings demonstratingthat more people thought the risks of GM foodsoutweighed the benefits than vice versa (Poortinga andPidgeon, 2003).

Note that current risks were valued lower than futurerisks and that current benefits were valued more highlythan future benefits. There is some suggestion of arelationship between risks and benefits—when risks arevalued more highly, benefits are valued lower. Putanother way, the greater the perception of risk, thesmaller the perception of benefit.

We expected to find that the value of future benefitswould be greater than current benefits, if respondents

believed the claims made about the future benefits of GM

technologies. This expectation was based on the claimsthat future GM technologies will present clearerconsumer, environmental and/or social benefits. How-ever, current benefits were valued greater than futurebenefits. Thus it appears that the respondents to thisstudy do not expect many of the claims made about thefuture benefits of GM technologies to materialise.Consider for example, the question ‘‘to what extent doyou think people in developing countries will benefitfrom the production of GM crops with added vita-mins?’’. Of respondents 50% said ‘not at all’ or ‘verylittle’. Only 5% said ‘some’. This corresponds with thescepticism and suspicion revealed by the commentsnoted above from non-respondents.

There have been claims that some environmentalistsmight be more inclined to accept some GM technologies

if they believe that environmental benefits will arise fromthe application of the technology (Isaacs, 2001). How-ever, the responses to specific questions within oursurvey suggest this is unlikely to be the case among thosemost pro-environmental and/or anti-GM. For example,in response to the question ‘‘to what extent do you thinkgrowing commercial GM crops with built-in pesticide(in other countries) reduces the need for agri-chemi-cals?’’ only 8% of respondents answered ‘some reduc-tion’. The majority (45%) answered ‘no reduction at all’or ‘very little reduction’. Similarly, in response to thequestion ‘‘to what extent do you think growingcommercial herbicide-resistant GM crops in othercountries (which leads to better targeting of herbicides)reduces agri-chemical use?’’, 45% of respondentsanswered ‘no reduction at all’ or ‘very little reduction’.If people do not believe these claims about GMtechnologies they are unlikely to consider they couldoffer any environmental benefits. As regards theenvironmental benefits that GM technologies mightoffer in the future, a similar story is revealed. Fiftypercent of respondents did not expect yield-increasingGM species to help to preserve or re-instate ‘natural’habitats.

There was a degree of consensus relating to questionsabout environmental impacts. In contrast, responseswere more ambiguous regarding health risks. Thissuggests that respondents have made up their mindsregarding GM and the environment but are still unsureabout the health effects. For example, consider thequestion ‘to what extent do you think there will behealth risks for UK consumers from eating GM foodproducts and ingredients?’. 24% said ‘very little risk’,30% said ‘some’ or ‘a lot’. Similarly, when asked aboutthe health risks for people living in the vicinity of fieldscale trials, 37% said they thought there had been ‘no

ARTICLE IN PRESSC. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–3734

risk at all’ or ‘very little risk’ while 18% thought therewas ‘a lot’ or ‘some risk’.

The study revealed that women considered the risks ofcurrent GM technology to be greater than menconsidered them to be (average value of risks underscenario one was 4 as opposed to 3.5) (Table 3). Thisfinding corresponds to surveys conducted with membersof the general public that women considered GM foodto be more risky than men (Subrahmanyan and Cheng,2000; Moon and Balasubramanian, 2001). A study byIsaacs (2001) investigating acceptance of GM foodsfound no significant difference in willingness to buy orunwillingness to buy among many socio-demographicvariables. The exception was that a larger portion of the

Table 3

Respondent variables and ‘now’ scenario risks and benefits

Variable Meana

Now scenario risks

Gender

Female Male

4.04 3.52

Place of residence

Urban Rural

3.66 4.22

Group membership

Anti-GM Environment

4.06 3.70

Household

Without children With children

3.70 3.82

Education

O grade Higher U/deg P/deg

4.42 3.84 4.06 3.26

Near trial?

NEP group

Weak eco Strong eco V strong eco

3.21 3.59 4.22

Now scenario benefits

Gender

Female Male

2.33 2.26

Place of residence

Urban Rural

2.41 1.92

Group membership

Anti-GM Environment

1.89 2.27

Household

Without children With children

2.24 2.30

Education

Near trial?

Near trial Not near DK

1.80 2.45 2.42

NEP group

aMeans are included for significant variables only.bSignificant at the 0.05 level.cSignificant at the 0.01 level.

‘unwilling to buy’ sample were female than the ‘willingto buy’ sample.

Level of education and the extent to which respon-dents are ecocentric, were also shown to have an effecton how great they believe the risks of current GMtechnologies are (Table 3). Specifically, respondents witha postgraduate degree valued current risks of GMtechnology lower than those respondents whose highesteducational qualification was O grade, O level orequivalent, or undergraduate degree. Again, this reflectsfindings from other studies with the general population(Moon and Balasubramanian, 2001; Gaskell et al.,2003). The more strongly ecocentric a respondent wasthe greater the value they placed on risks of current GM

T-test statistic Degrees of freedom F statistic P-value

2.11 25 — 0.045b

�1.88 25 — 0.072

0.75 10 — 0.473

�0.40 23 — 0.695

— — 3.88 0.023b

— — 1.67 0.209

— — 5.88 0.008c

0.29 25 — 0.776

2.07 25 — 0.049b

�1.16 10 — 0.271

�1.16 10 — 0.271

— — 0.63 0.606

— — 3.90 0.034b

— — 0.60 0.559

ARTICLE IN PRESSC. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–37 35

technologies. The study by Isaacs (2001) also used theNEP scale and found that those who opposed GM foodwere more ‘pro-environmental’ than those who sup-ported it.

The study revealed that people living in urban areasconsidered the benefits of current GM technology to begreater than those people living in rural areas (benefitsvalued at an average 2.4 as opposed to 1.9) (Table 3).There was also evidence that individuals living near to aFSE trial site placed a lower average value on thebenefits of current GM technology than those not livingnear an FSE trial site or those who did not knowwhether there was an FSE trial nearby. This suggeststhat heightened concern may have arisen out of closerpersonal experience of GM crops.

There was evidence that a range of socio-demographicvariables influenced the value people placed on the risks

Table 4

Respondent variables and ‘future’ scenario risks and benefits

Variable Meana T-te

Future scenario risks

Gender 2.

Female Male

4.12 3.58

Place of residence �2.

Urban Rural

3.68 4.50

Group membership 2.

Anti-GM Environment

4.66 3.83

Household �0.

Without children With children

3.75 3.94

Education —

Near trial? —

NEP group —

Future scenario benefits

Gender 0.

Female Male

1.69 1.65

Place of residence 2.

Urban Rural

1.80 1.23

Group membership �1.

Anti-GM Environment

1.17 1.66

Household �0.

Without children With children

1.65 1.68

Education —

Near trial? —

Near trial Not near DK

1.23 1.95 1.63

NEP group —

aMeans are included for significant variables only.bSignificant at the 0.05 level.cSignificant at the 0.01 level.

of future GM technologies. The average value of risksunder the future scenario was 4.1 for women as opposedto 3.6 for men; the average value was 4.5 for peopleliving in rural areas as opposed to 3.7 for people living inurban areas; the average value was 4.7 for people whowere members of anti-GM campaign groups as opposedto 3.8 for people who were members of environmentgroups (Table 4). Overall, women, people living in ruralareas and members of anti-GM campaign groupsconsidered the value of future risks of GM technologyto be greatest. The other variables did not have asignificant effect on the value of risks under the futurescenario.

People living in urban areas were shown to value thefuture benefits of GM technology higher than those inrural areas (average value 1.8 as opposed to 1.2)(Table 4). Those who lived near to FSE trial sites did

st statistic Degrees of freedom F statistic P-value

18 25 — 0.039b

88 25 — 0.008c

22 10 — 0.05b

68 23 — 0.502

— 2.11 0.128

— 1.08 0.355

— 2.96 0.071

13 25 — 0.894

32 25 — 0.029b

32 10 — 0.216

14 23 — 0.887

— 0.40 0.751

— 3.70 0.040b

— 0.67 0.520

ARTICLE IN PRESS

Table 5

Results from risk ranking exercise

Deaths per year Percent

C. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–3736

not consider the potential future benefits of GMtechnology to be as great as those who did not livenear to a site where there were FSEs.

More than 10,000 18.4

4,000–10,000 5.3

501–3,999 5.3

21–500 15.8

20 or less 26.3

No response/don’t know 28.9

Table 6

ANOVA—Risk ranking exercise and NEP score

Variable F statistic P-value Means

Risk

ranking

4.46 0.019a More than 501 deaths per

year

36.09

Up to 500 deaths per year 32.19

aSignificant at the 0.05 level.

3.4. Results from risk ranking exercise

The risk ladder variable was reclassified from aninitial 18 categories to six categories in order toovercome the problem of very small numbers ofrespondents in some categories. Results in Table 5 arethe reclassified results. It is important to bear in mindthat respondents were effectively completing this ex-ercise ‘blind’. They were not told the reason for theranking of items on the scale (other than risk posed tohealth) and hence were not given the number of deathsper year figures. The majority of respondents placedGM at a point on the scale that suggested that it (GMtechnology) was expected to lead upto 20 deaths peryear. This corresponded with variant CJD. The secondlargest response was that GM was expected to causemore than 10,000 deaths per year.2 This correspondedwith the category ‘accidents’.

Before reclassifying this variable, the majority ofrespondents to the risk-ranking exercise who did notanswer ‘don’t know’ placed GM at a point on the scalesuggesting that respondents believed GM presented anamount of risk equivalent to causing between 21 and 499deaths per year. However, this was the point on the scalebetween food-borne illness and variant CJD. It isreasonable to assume that the food and agriculture linkof these risks influenced peoples’ responses. This resultsuggests that, on average, respondents viewed the healthrisks to be less than those posed by six out of eighthealth risks presented on the scale. Similarly, Poortingaand Pidgeon (2003) found that GM food, whencompared with four other risky items, was consideredto be among the least risky. Overall, responses werewidely spread along the scale, suggesting little consensusabout the scale of the health risk posed by GM food.This is understandable, given the amount of uncertaintysurrounding the specific nature, probability and scale of

2It is possible, using the value of statistical life (VOSL), to elicit an

economic value for the health risk of GM food from these results. One

of the most widely used VOSL figures across government is that from

the former Department of Transport (Andrews and McCrea, 1999).

This is £850,000 and is based on individual’s willingness to pay to

reduce risks of road accidents. With risks relating to health, a different

value is sometimes used. For example, for substances where exposure

can result in cancer, the Department of Transport’s VOSL is doubled

to account for individual’s dread factor associated with this disease. If

the uncertain health risks associated in some peoples’ minds with

exposure to GM food encompass the same dread factor it might be

appropriate to use the same approach. However, for the purpose of

this exercise we assume a VOSL of £850,000. Assuming up to 20 deaths

per year equates to the perceived risk of GM food, we can say that the

value of this risk to the most risk-averse potential consumers is up to

£17,000,000 per year.

risk. It is also supported by the findings above thatresponses to the health questions were more ambiguousthan those relating to the environment.

ANOVA was conducted on the NEP score against therisk-ranking exercise to see whether the degree to whichrespondents could be said to be ecocentric influencedtheir ranking of the risks of GM against other healthrisks. Results showed there was no relationship whenthere were six categories of risk but the means suggestedthere might be a relationship, so the risk-rankingvariable was reclassified into three categories andANOVA conducted again (Table 6). This time resultsshowed that respondents with a higher average NEPscore (more ecocentric) expected GM food to result in alarger number of deaths per year, hence considered it tobe potentially more risky to health.

4. Conclusions

The results enable us to draw a number of conclusionsregarding the construction of risk perceptions of anti-GM campaigners in relation to GM technology in foodand agriculture. First, there does appear to be a linkbetween the perceived risks relating to GM food, andperceived benefits. When benefits are expected to begreater, risks are expected to be smaller. It could beconcluded that this finding lends weight to the expecta-tion that the public are likely to be more accepting ofGM food in the future when consumer benefits areoffered by the products. However, respondents appearto be sceptical about the claims of future benefits of GMtechnology, suggesting that actually the perception ofhigh risks will remain among the most risk-averse

ARTICLE IN PRESSC. Hall, D. Moran / Journal of Rural Studies 22 (2006) 29–37 37

section of society. Opposition and protest are thereforelikely to continue.

Nevertheless, the results of this study show that anti-GM campaigners and members of environment groupscan be segmented into groups with different perceptionsof the risks and benefits of GM food, and should nottherefore be viewed as an homogeneous group. Thissuggests that some campaigners may be less inclined toremain staunchly opposed to GM food than others. Thispossibility is borne out by the fact that certain socio-demographic variables affect campaigners’ perception ofthe risks and benefits of GM food. For example, there isevidence that, in some cases, women and those living inrural areas consider the risks of GM to be greater, andthe benefits less, than men and people living in urbanareas. Perhaps, therefore, men and people living inurban areas, even those opposed to the technology atthis stage, may be prepared to purchase GM food atsome point. However, there is also evidence that peopleliving near to an FSE trial site expected the futurebenefits of GM technology to be less than those whowere not living near an FSE trial site. This suggests thatcloser personal experience of GM crops may in fact leadto greater resistance.

Overall, this study contributes to the GM policydebate by adding to the understanding of the percep-tions of this very specific population, whose views haveplayed a high profile role in the development of the GMissue across Europe. The challenge remains for policymakers to incorporate these views into future policydecisions, whilst also acting in accordance with Eur-opean legislation and the economic interests of thebiotechnology and agricultural industries.

Acknowledgements

This paper is part of a research project funded by theScottish Executive—The social construction and eco-nomic value of food-borne risk perceptions: the case ofgenetically modified food ingredients.

References

Andrews, R., McCrea, P., 1999. UK Health and Safety Executive

approach to exposure limit setting. In: Proceedings of the OECD

Workshop on the Integration of Socio-Economic Analysis in

Chemical Risk Management Decision-Making, January 7–9, 1998,

London. OECD, Paris, pp. 193–202.

DEFRA, 2002. Survey of public attitudes towards the environment

and to quality of life—2001. DEFRA, London.

Dudleston, A., Hope, S., Littlewood, A., Martin, C., Ormston, R.,

NFO System Three Social Research & MORI Scotland, 2002.

Scotland’s people: results from the 2001 Scottish household survey,

vol. 5: Annual Report. Scottish Executive National Statistics

Publication, Edinburgh.

Dunlap, R.E., Van Liere, K.D., Mertig, A.G., Jones, R.E., 2000.

Measuring endorsement of the New Ecological Paradigm: a revised

NEP scale. Journal of Social Issues 56 (3), 425–442.

Gaskell, G., Allum, N., Stares, S., 2003. Europeans and biotechnology

in 2002, eurobarometer 58.0. A report to the Directorate General

for Research from the project ‘Life Sciences in European Society’.

European Commission, Brussels.

Grove-White, R., Macnaghten, P., Mayer, S., Wynne, B., 1997.

Uncertain world: Genetically Modified Organisms, Food and

Public Attitudes in Britain. IEPPP, Lancaster University, Lan-

caster.

Heller, P. (Ed.), 2003. GM Nation? The Findings of the Public Debate.

DTI, London.

Isaacs, J.C., 2001. Acceptance of Genetically Modified Foods and

Environmental Attitude. Louisiana Dept of Wildlife and Fisheries,

Louisiana.

Lemkow, L., 1993. Public Attitudes to Genetic Engineering: Some

European Perspectives. Office of the Official Publications of the

European Communities, Brussels.

Moon, W., Balasubramanian, S., 2001. A multi-atttribute model of

public acceptance of genetically modified organisms. Selected paper

for presentation at annual meeting of American Agricultural

Economics Association, Chicago, Aug 5–8, 2001.

MORI, 2003. Continuing Opposition to GM Foods. MORI, London

Available online at: http://www.mori.com/polls/2003/meb2.shtml.

Olubobokun, S., Phillips, P., Hobbs, J.E., 2001. Analysis and

differentiation of consumers’ perceptions of genetically modified

foods. Paper presented at the Fifth International Conference of the

ICABR on ‘Biotechnology, Science and Modern Agriculture: A

New Industry at the Dawn of the Century’, Ravello, Italy, June

15–18, 2001.

Poortinga, W., Pidgeon, N., 2003. Public perceptions of risk, science

and governance. Main findings of a British survey of five risk cases.

Centre for Environmental Risk, University of East Anglia,

Norwich.

Purdue, D.A., 2000. Anti-Genetix: The Emergence of the Anti-GM

Movement. Ashgate Publishing, Ltd, Aldershot.

Reisner, A., 2001. Social movement organisations’ reactions to genetic

engineering in agriculture. American Behavioral Scientist 44 (8),

1389–1404.

Rosati, S., Saba, A., 2000. Factors influencing the acceptance of food

biotechnology. Italian Journal of Food Science 4 (12), 425–434.

Scott, K., 2002. Trials and tribulations. The Guardian. Manchester,

May 1, 2002.

Subrahmanyan, S., Cheng, P.S., 2000. Perceptions and attitudes of

Singaporeans toward genetically modified food. The Journal of

Consumer Affairs 34 (2), 269–290.

Sykes, K., 2003. Introduction to Citizens’ Jury on GM food. Power-

point presentation for Food Standard’s Agency Citizens’ Jury on

GM Food, London, 4th April 2003. Available online at: http://

www.food.gov.uk/multimedia/presentations/gmjurykathysykes.ppt

Verdurme, A., Gellynck, X., Viaene, J., 2001a. Consumers’ accept-

ability of GM food. Contributed paper at 71st EAAE Seminar,

‘The Food Consumer in the Early 21st Century’, Zaragoza, Spain,

19–20 April, 2001.