Trust and willingness to pay for nanotechnology foodJ. Roosena,, A. Biebersteina, S. Blanchemancheb, E. Goddardc, S. Maretteb, F. VandermoeredaTUM School of Management, Technische Universitt M} unchen, Freising, GermanybUMR Economie Publique, INRA, Paris, FrancecDepartment of Resource Economics and Environmental Sociology, University of Alberta, Edmonton, CanadadDepartment of Sociology, Antwerp University, Antwerp, Belgiumarti cle i nfoArticle history:Available online 20 January 2015Keywords:EvaluationNanotechnologyOne-and-a-half-bound dichotomous choiceSelf-protectionTrustWillingness to payabstractWe analyze the role of trust in the evaluation of a new food technology, namely nanotechnology. A liter-ature review in the social and economic sciences reveals that many different trust concepts are available.The economics literature suggests that trust can lead to lower efforts of self-protecting behavior. Trans-lating this concept into the framework of willingness to pay (WTP) for food products allows for the der-ivation of hypotheses on the workings of trust. We show that WTP for new food characteristics increaseswith trust also when new information about the technology is revealed. The results are conrmed withonline survey data for Canada and Germany and experimental data in Germany. 2014 Elsevier Ltd. All rights reserved.IntroductionThedevelopmentofintegrated, longsupplychainsandsomefood technologies has made modern food businesses more vulner-abletosafetyincidentsandconsumertrustinthefoodindustryhas beenchallengedby a series of scandals. Consumers havebecomeskepticaloffoodinnovationsandindustrialization. Evenwhenrespondingtothegrowingdemandfor convenienceandsafety, thesuccessful introductionofnewfoodtechnologieshasbecome a major challenge.It has been shown that trust is a construct that helps people toaccept risks in the face of moral hazard. Hence, trust contributes toeconomicprogress(Arrow, 1974). However, despitethisrecog-nized importance of trust in the economics literature, most originalcontributionsandthosethat operationalizetrust asameasure-ment construct have come from the eld of other social sciences,such as sociology (e.g., Luhmann, 1968/2000; Giddens, 1990). Heretrust is a source of social capital that helps to reduce complexityandtofacilitateinteraction. Giddens(1990), forexample, positsthat trust in expert systems is a mechanism to reduce complexity:when a laypersons knowledge is inadequate, the person will retainhis or her ontological security by trusting experts. In a context offood safety crises and the development of new food technologies,trust is consideredtobeakeyconcept (Berg, 2004; deJongeetal., 2008;Freweretal., 1996;Kjrnesetal., 2007;RennandRohrmann, 2000; Sassatelli and Scott, 2001).The objectives of this paper are twofold. From extant literature,denitions of trust are reviewed with regard to food technologiesand an attempt is made to relate trust to consumers willingnessto pay (WTP) for altered food characteristics. Secondly, therelationshipbetweentrustandWTPisassessedforanewfoodattributethat is introducedusingnanotechnology. Weanalyzewhether trust correlates with the acceptance of a functional foodattribute(here: vitaminenrichment andprotection) andif thisevaluationchangeswhenconsumerslearnthattheattributehasbeen created by means of nanotechnology.Foodnanotechnologyis regulatedunder existinglegislation(European Commission, 2012; Government ofCanada, 2013) anda case by case approach is applied through a pre-market approvalsystem (European Commission, 2012). A review of approaches totheregulatorygovernanceofnanotechnologyupto2009canbefoundinPelleyandSaner (2009). However, theapplicationofnanotechnology in thefood domain is surrounded by high levelsof scientic uncertainty withseveral studies pointing to possiblenegativelong termeffects (Wangetal., 2006; Oberdrsteretal.,2005). Theapplicationofnanotechnologyinthefoodindustryisstilllimitedandnewandratherunknowntoconsumers. Inthiscontext of low knowledge, high complexity and high uncertaintynobody retains the authority of better knowledge (Luhmann,1993) and the safety of the food market increasingly depends onthe decisions of the responsible actors (Fischler, 1988). Consumeracceptance of such new, complex technologies is likely to dependon how much they trust these actors.http://dx.doi.org/10.1016/j.foodpol.2014.12.0040306-9192/ 2014 Elsevier Ltd. All rights reserved.Correspondingauthor at: TUMSchool of Management, MarketingandCon-sumerResearch, AlteAkademie16, 85350Freising, Germany. Tel.:+498161713318; fax: +49 8161 71 4501.E-mail address: jroosen@tum.de (J. Roosen).Food Policy 52 (2015) 7583ContentslistsavailableatScienceDirectFood Policyj our nal homepage: www. el sevi er . com/ l ocat e/ f oodpolFor the analysis we use data from online surveys conducted inCanadaandGermany. TheGermansurveywasaccompaniedbyan economic laboratory experiment. In all three studies, we proposetwotypesoforangejuicetotherespondentsandmeasureWTPundertwodifferentinformationscenarios. Whiledatafromtheonlinesurveys havebeenanalyzedinMatinet al. (2012) andVandermoere et al. (2010), results on the WTP data of the surveyshave not yet been published. Bieberstein et al. (2013) analyze theexperimental data used in this paper, however, the WTP data hasnot been linked to the question of trust. To our knowledge, this pre-sentsoneofthefewstudiesthatanalyzetheimpactoftrustonWTP. Nocella et al. (2010) analyze the impact of trust on consumersWTP for higher animal welfare standards. However, they measuretrust not asinstitutional trust, but byusingFishbeinsattitudemodel. OhandHong(2012) provideatheoretical analysisoftherole of trust on WTP, where trust presents a shift in the expectedvalue of an uncertain outcome. Meyer and Liebe (2010) use a mea-sure of generalized trust as an antecedent of WTP for environmentalprotection in Switzerland. Our analysis differs from previous anal-ysisinthatweuseinstitutional trustasadeterminantof WTPandthatwenotonlyanalyzethemeanimpact, butalsoaspectsof the distributional impact based on experimental data.Our results show that information about the use of a new tech-nology in the food domain leads to a welfare decrease for consum-ers. Furthermore, we can show that trust increases WTP and that itprotects WTP from bad news.Thepaperproceedsinfoursections. First, aliteraturereviewcovers thedifferent denitionsoftrust andtheiramenabilityforameaningful denitionof trust inthecontext of WTPstudies.Thefollowing sectiondescribesthesurveysandthemethods formeasuring trust and WTP. Results are presented next and the paperconcludes with implications for research and regulation.The role of trustLiterature reviewTrust is a construct that has beendeveloped in the social sci-ences. In this section we attempt to grasp the signicance of trustinthesocialsciencesanditsmeaningfortheeconomicsofcon-sumers dealing with new food technologies.According to Earle (2000) many trust researchers accept someversion of the denition offered by Rousseau et al. (1998): Trustis a psychological state comprising the intention to accept vulnera-bility based upon positive expectations of the intentions or behav-ior of another (p. 395). This type of interpersonal trust has beendened in atrustortrustee relationship and more specically asastateof expectationfromatrustor(Gambetta, 1988; Bradachand Eccles, 1989). An alternative approach to trust is less centeredon the individual and includes groups as trusting or trusted actors.Rotter (1967, p. 651), for example, denes interpersonal trust asexpectancy held by an individual or a group that the word, promise,verbal or written statement of another individual or group can bereliedupon. (seealsoNooteboom, 1996). Anotherdenitionofbasic trust goes back to Erikson (1953) who describes it as centraltoahealthypersonality. Trustintermsof personalityhasbeenassessed using attitudinal survey questions such as the one in gen-eral social surveysthatasksthefollowing:Generallyspeaking,would you say that most people can be trusted or that you cantbe too careful in dealing with people? (Glaeser et al., 2000, p. 812).Research on technology acceptance has focused on the impactof social trust on technological risk perception and on the accep-tance of a technology. Hereby, the denition of social trust relatesto interpersonal relationships and to relationships betweenindividuals and institutions (Kasperson et al., 1992). While inter-personal trust and institutional trust are often differentiated(Hudson, 2006), the latter is supposed to be important in complexsocietiessuchasours(Luhmann, 1968/2000)andimportantforunderstanding the acceptance of moderntechnologies. Institu-tional trust has alsobeenfoundtoplayanimportant roleinexplaining perceivedrisk(Earle andCvetcovich, 1995; Siegristet al., 2000; Slovic, 1999) and the acceptance of new food technol-ogies (Costa-Font et al., 2008; Visschers et al., 2007).Trust and willingness to pay (WTP)Calculative trust plays a central role in decisions under uncer-tainty and is based on the expected behavior of others. While trustis mostly measured by attitudinal questions in the social sciences,economists tend to use choice-based metrics (McEvily et al., 2012).Oneexampleof measuringtrust istheinvestment game(Berget al., 1995). Here two groups of players interact on an investment.A player in group A can share an amount x of his show-up fee, say 10, with a player in group B. By this investment the amount growsto ax, where a > 1, and players in group B can decide the share to bereturned (b). Foraplayer in group A thepay-offis 10 x + abx,and for the player in group B it is (1 b)ax. The full informationequilibrium of this game is to invest nothing. However, precludingsidearrangements, individualsstill invest andthis isbasedontrust. Indeed, it is this feature of trust in overcoming the transac-tioncostsof self-protectionthatmakestrustubiquitousineco-nomic relationships (Arrow, 1974). Trust means that anindividual is willing to forego self-protecting behavior that wouldcome at a cost. Hence McEvily et al. (2012) turn to a denition ofdistrustthatisexpressedintermsofthe(transaction)costthatsomeoneiswillingtobearinordertobelessvulnerabletotheaction of another party. Taking a positive perspective, trust avoidscosts that arise from measures of self-protection.Translating this idea into a utility maximizing framework, ourmodel starts on a money gamble. Suppose the payoff of an individ-ual isx Lwithprobabilitypandxwithprobability1 p. Theexpected utility of the outcome is hence:EU pux L 1 pux 1where u is the instantaneous utility function and L is the loss in thegamble. Now suppose the individual can invest an amount e in pre-vention or self-protection (Ehrlich and Becker, 1972). By thisamounttheindividualiscapableofdecreasingtheprobabilityofa loss, so thatmaxeEU peux L e 1 peux e 2where p decreases in e. It has been shown that the optimal amountof prevention effort is not monotonic in risk aversion (Dionne andEeckhoudt, 1985)andprudence(EeckhoudtandGollier, 2005). Incontrast to a mean-preserving contraction of the payoff distributionin the sense of Rothschild and Stiglitz (1970) preventioncauses acost that shifts the probability distribution function downward by e.Translating the expected utility maximizing framework to thedenition of trust by Rousseau et al. (1998) and using the conceptof distrust according to McEvily et al. (2012) leads to the conclu-sion that self-protecting effort e is lower with higher interpersonaltrust. Hence trust alters the perceived risk by changing the proba-bilities of negative outcomes.We now turn to the role that trust in food production may havewhen measuring WTP. The utility of a good that is consumed in xunits depends onconsumptionandfunctional product qualitydenoted by q. Utility also depends on income, w, such that u(x, q,w). For simplicity, we consider the case where the product is eithernot consumed (x = 0) or consumed in a single unit (x = 1). From thissetup we can dene the WTP for a quality q1 and q2 by the follow-ing set of equations.76 J. Roosen et al. / Food Policy 52 (2015) 7583u0; 0; w u1; q1; w WTP1 3u0; 0; w u1; q2; w WTP2 4Assumethatqualityq1isbroughtaboutbysometraditionaltechnology. A new technology then allows a different specicationof the attribute as q2. However, this new technology is associatedwith the risk of a loss L with probability p and yields a new WTPdened byu0; 0; w pu1; q2; w L WTP2 1 pu1; q2; w WTP25Combining (3) and (5) givesu1; q1; w WTP1 pu1; q2; w L WTP2 1 pu1; q2; w WTP2 6Three different cases arise:(a)The qualities of the products produced with traditional andnewtechnologiesareequal(q1 = q2). Inthiscase, thenewtechnology, 2, introduces a risk to consumers andWTP1 > WTP2.(b)The newtechnology leads to more effective functionalbenets(q1 < q1). Inthiscase, theriskperception, pandL,the benet perception, q2, as well as the preferences of theconsumer, u(), determinetherelationshipbetweenWTP1andWTP2. WTP2will increasewithtrustastrustreducesself-protectingeffort, e.g. foregoingthefunctional benetof the new technology.(c)A trusting consumer will be more ready to accept eventualuncertainty posed by the technology. We consider twocases:Good news: WTP increases with the introduction of a new tech-nology,i.e., WTP2 > WTP1. Ahighleveloftrustwillmake thisincreaseinWTPlargerandWTP2 WTP1will increasewithtrust.Bad news: WTP decreases with the introduction of a new tech-nology,i.e., WTP2 < WTP1. Ahighleveloftrustwillmake thisdecreaseinWTPsmaller andWTP2 WTP1will decreaseinabsolute value with trust.MethodsWe study the impact of trust on WTP for a new food technology.An online survey was conducted in Canada and Germany. The Ger-man survey was accompanied by an economic laboratory experi-ment. Inall threestudies, wepresenttwotypesofhypotheticalorange juice to the respondents and measure WTP under two differ-ent information scenarios. As no food products using openly nano-technology were available on the market at the time, we createdthenanocharacteristicsbasedonareviewofliterature, selectingthe most likely nanotechnology application in the food domain.The survey questionnaireThe questionnaire contains three parts. First, attitudinal ques-tionsareaskedregardingfoodchoicemotivesbasedonSteptoeet al. (1995) and institutional trust. The institutional trust questionprompts respondents to indicate their trust in the respective insti-tutionsregardingtheirresponsibilityoverthesafetyoffood. Sixinstitutions are considered: Agriculture, food-industry, science/research, pharmaceutical industry, government agencies/publicauthorities and consumer organizations. Responses were recordedon a 5-point rating scale from 1 = no trust to 5 = very high trust. Anexploratory factor analysis on the data of the six variables in eachcountry shows that trust in the rst ve institutions loads on thesame factor, while trust in consumer organizations loads on a sec-ond factor. Hence an overall trust variable was constructed as themean score for the rst ve trust items.Additional questions wereasked regarding health risks: Howdoyouconsiderthehealthriskposedtoconsumersbyregularcon-sumption of . . . and a list, as provided in Table 3, is used for a ratingscale from very low risk (1) to very high risk (5).Questions on WTP follow. The product proles presented to therespondents are two types of orange juice, one fortiedwithvitaminD, the second in a UV-light protecting bottle that protects the vita-min C in the orange juice. Questions in the online survey were posedin form of a one- and-one-half-bound (OOHB) dichotomous choicequestion (Cooper et al., 2002). The OOHB method is thought to over-come starting point bias related to the double-bound dichotomouschoice method (Hanemann et al., 1991). At the same time, it maypreserve efciency gains over single dichotomous choice questions.Based on a market search in local supermarkets two price levelsfor a1-literbottle oforangejuice werechosen(Ca-$ 1.75/0.90and Ca-$2.50/1.80) and introduced as the variation of the typicalmarket price.1Using different price levels limits the comparability oftheresultsforthetwocountries. However, it allowsbasingcon-sumer responses on a naturally observed anchor. Respondents wererandomizedintorst receiving the lower (upper) price. If theiranswer to the rst price was positive (negative), they received a sec-ond question with the upper (lower) price. In a rst set of WTP ques-tions people had not been informed about the use of nanotechnologyin these products. In the next step, consumers were informed aboutthe nature of nanotechnology and the particular use of changing thefunctional attributes of the products. The WTP questions were thenrepeated (see Fig. 1). Between these two WTP questions, the follow-ing information was given:Nanotechnology refers to materials, systems and pro-cesseswhichexist or operateintherangeof about 1100nm(nm). Onenanometer(nm)isonemillionthofamillimeter (mm). Materials at this scale show novel prop-erties that lead to novel applications in diverse elds suchas medicine, cosmetics, biotechnology, energyproduc-tion and environmental science. There is uncertaintyregardinghownanomaterialsmayinteract withhumanhealthand theenvironment.Nanotechnologyoffers newopportunitiesforfood indus-tryapplication. Manufacturednanomaterialsarealreadyused in some food products, nutritional supplements,and food packaging applications. Nanotechnology allowsfortheimprovementofbarrierfunctionsinfoodpackag-ingtoreduceUV-lightexposureormicrobialgrowthandthus extend the shelf-life of many food-products. Further-more, nano-biosensors are able to control the foods leveloffreshnessbyindicatingspoiledfoodtotheconsumersby means of color change. There is not much known abouttheeffects onhumanhealth andenvironment.Finally, socio-demographic variables were collected.Theexperimentproceededsimilarly, butWTPwasmeasuredusing a price list from0.90 and1.80 in 10 cents intervals for a1 lbottle oforangejuice. Participants could answer thequestion1Using an exchange rate of 1 = 1.38 Ca-$, 0.90 correspond to about Ca-$ 1.25, and1.80 correspond to Ca-$ 2.50.J. Roosen et al. / Food Policy 52 (2015) 7583 77fortheirwillingnesstobuythejuicesatthevaryingpriceswithyes, noormaybe. Onlyafrmativeyes-responseswerecountedin determining WTP.Analysis of trust and risk perceptionFirst descriptivestatisticsarecalculatedforthesampleasawhole. Then, thesampleisseparatedintolowandhighinstitu-tional trust, where low trust is dened as Trust < 3. The two subs-ampleswerecomparedwithrespecttotheirperceptionsoffoodrisks.Analysis of WTPFor the analysis of the WTP variables in the OOHB format, wefollowCooperetal. (2002). Letthetrue, unknownWTPofcon-sumer i be denoted by Ci. Ci can be cast in the random utility frame-work and is a function of personal characteristics of respondent i.I.e., if WTP follows a cumulative distribution function G(Ci, h), thentheparametervectorhdescribestheparametersofthedistribu-tion, such as the mean and variance.Twodifferent prices areproposedtorespondents: PU(=$2.5/1.80) and PL(=$1.75/0.90). There are three possibilities:(1)The respondent accepts the product neither at the low nor atthe high price. The probability for the observation ispnm GPLi ; h.(2)The respondent accepts the product at the low price but notat the high price. The probability is pym GPUi; h GPLi ; h.(3)Therespondentacceptstheproductatthehighprice. Theprobability is pyy 1 GPUi; h.Inthechoicequestion,amaybeoption wasincluded: thesemaybe responses havebeencountedas no answers intheeconometric estimation.The resulting log-likelihood function for the sample i = 1, . . . , NislnLh XNi1dnnilnGPLi ; h

dyniln GPUi; h GPLi ; hhdyyiln 1 GPUi; h iThe logistic function is commonly used in dichotomous choiceestimations due to its close resemblance to the normal distributionfunction and its computational ease. We use a logistic function forG(; h) and assume that the mean is a linear function of K personalcharacteristicsoftherespondent andtheparametervector tobeestimated. Hence, expected WTP above the price can be derived asWillingness to pay I In what follows we will present you information about two pure orange juices sold in one litre bottles. On the market, the average price of this type of orange juice varies between $1.75 and $2.50 per litre. Orange juice A This orange juice is fortified with vitamin D. According to scientific estimation, many Canadians have vitamin D intakes below recommendations as a result of inadequate intake and inadequate sunlight exposure. Orange juice B This orange juice is filled in a plastic bottle that is fabricated in a way to reduce the juices exposure to UV-light. Exposure to UV-light has an adverse effect on important food nutrients like vitamin C. Willingness to pay II Nanotechnology offers new opportunities for food industry application. Manufactured nanomaterials are already used in some food products, nutritional supplements, and food packaging applications. Two examples in development are the two orange juices than have already been presented to you above. (In the market, the average price of this type of orange juice varies between $1.75 and $2.50 per litre) Orange juice A Orange juice A is fortified with vitamin D by means of nanotechnology. The vitamin D is enclosed in a nanoscale capsule that allows a better absorption and mobilization of the vitamin.According to scientific estimations, many Canadians have vitamin D intakes below recommendations as a result of inadequate intake and inadequate sunlight exposure. Orange juice B Orange juice B is produced by means of nanotechnology. The bottle is imbued with nano titanium dioxide particles that reduce UV damage of food nutrients.Exposure to UV-light has an adverse effect on important food nutrients like vitamin C.Fig. 1. Product information within the WTP questions.78 J. Roosen et al. / Food Policy 52 (2015) 7583ECi a0 b1X1i . . . bKXKi cPiwith a0 as a constant and the parameter c to price measuring themarginal utilityof income. AverageWTPforthesampleaveragecanbeestimatedbyestimatingE[Ci] = a0 cPi(withoutexplana-tory variables), so that WTP = a0/c. The standard error of the WTPestimate is obtained by the delta method.Regardingtheexperimental WTPdata, weadopt astandardTobit model on the pooled observations. WTPi is used as the depen-dent variable as censored at PU= 1.80 and PL= 0.90, so thatWTPi Ciif0:90 6 Ci 6 1:800:90 ifCi< 0:901:80 ifCiP1:808>:ResultsAnonlinesurveywas conductedwith615Englishspeakersresiding in Canada. Table 1 provides descriptive statistics on sev-eral sociodemographics of the sample and general trust measures.About 50%of thesamplearemale, theaverageincomeis Ca-$70 578 and the mean age of the sample is 50.1 years. Comparedtothe Canadianpopulationsurveyrespondents are somewhatolder and better off (Matin et al., 2012).Amongthefoodchoicemotives accordingtoSteptoeet al.(1995) we consider importance of price (M_price), importanceof nutrient content (M_nutrition) andimportanceof natural-ness (M_natural) as relevant. Items are evaluated on a ve-pointLikert scale. We report summated scores with M_price containingtwo, M_nutrition six, and M_natural three items. The nutrient andthe naturalness motive are with average values of 3.45 and 3.39 ofabout equal importance. The price motive is slightly more impor-tant (3.69). The average institutional trust in food has a mean valueof 2.75 (on a 5-point scale) and 55% of the sample show low trust(point value below three).The German online survey was conducted with a sample of 750respondentsbetweenFebruaryandApril 2009. Table2providesdescriptive statistics. The average age is 45 years, 51% of the sam-ple are male. Net monthly income is on average 1 995. ComparedtotheGermanpopulationsurveyrespondentsarelesswell-off(Federal Statistical Ofceof Germany, 2013). Thepricemotiveturnstobemostimportant(4.12) comparedtonutrition(3.74)andnaturalness(3.72). Themeantrustscoreis2.87, yieldingalow level of trust for 48% of the sample.Table 2 also provides some summary statistics for the experi-ment conducted in Munich, Germany. The lab sessions took placebetweenJanuaryandFebruary2009with143participants. Themajority of participants (55%) are female and the average monthlynet income is above that of the online survey sample. Theexperimental participantsare ofsimilar ageas theonline surveyrespondents. Comparedtothepopulationofthecitytheexperi-mental sample is somewhat older and has a higher income.The importance of trust for perceiving being protected against foodrisksSplitting the sample using the variable low institutional trust,weanalyzetheresponsestothequestionHowdoyouconsiderthe health risk posed to consumers by regular consumption of. . . inTable 3. Overall we observe that trust is relevant to explain con-cernsaboutfoodsafetyrisksposedbyvarioustechnologies. Forbothcountriesandall riskslowtrustconsumersaremorecon-cerned about food safety risks posed by food technologies, henceconrming the institutional trust hypothesis. The two subsamplesshow signicant differences in concern for the foodsgrown withpesticidesandotherchemicals, GMOs, irradiatedfood, preserva-tivesandarticial coloring, meat/shcontaininghormonesandantibiotics and foods based on nanotechnology in Canada. In Ger-many, differences are additionally signicant for supplementsand enriched foods.WTP for functional attribute before and after information onnanotechnologyUsing the responses of the OOHB dichotomous choice questionsin the Canadian survey and the relevant individual characteristicsof the respondents, we obtain 510 valid responses. The parameterestimates of the likelihood maximization are given in Table 4.ThehypotheticalproductjuiceAisvitaminDenriched. First,respondents do not know details of how the benet was achieved(second column of Table 4). Institutional trust inuences WTP pos-itively. Men and women do notdiffer in their evaluation. Higherimportance of price in the consumption decisions decreases WTPwhile more importance attributed to nutrient value increasesWTP. Educationandother sociodemographicvariableshadalsoTable 1Descriptive statistics Canada online survey.VariableaDenition Mean Std. Dev.Male (N = 613) 1 = Male, 0 = Female 0.49Income (N = 610) Total household income in Ca-$ (annual) 70577.87 33937.12Age (N = 615) In years 50.10 13.79Food choice motiveM_price (N = 606) Summated scale on importance of price 3.69 0.91M_nutrition (N = 594) Summated scale on importance of nutrient content 3.45 0.85M_natural (N = 596) Summated scale on importance of naturalness 3.39 1.11Trust (N = 540) How much trust do you have in the following institutions regarding their responsibility over the safety of food?Mean value for agriculture, food industry, science/research, pharmaceutical industries, gov. agencies/publicauthorities (scores range from 1 = no trust to 5 = very high trust)2.75 0.76Low trust (N = 540) Dummy variable = 1 if Trust < 3 0.55aBecause of varying number of missing data, the number of observation is indicated in parentheses (N=).Table 2Descriptive statistics Germany online survey and lab experiment.Variable Websurvey (N = 750) Experiment (N = 143)Mean Std. dev. Mean Std. dev.Male 0.51 0.45Income 1995.43 1265.59 2 577.77 1 630.76Age 45.38 15.36 45.74 14.35Food choice motiveM_price 4.12 0.86 3.80 0.78M_nutrition 3.74 0.85 3.99 1.01M_natural 3.72 1.04 3.92 0.73Trust 2.87 0.63 2.90 0.62Low trust 0.48 0.48J. Roosen et al. / Food Policy 52 (2015) 7583 79been included; because all of them turned out to be insignicant,the more parsimonious specicationis presentedhere. Finallythe price coefcient is highly signicant and indicates that utilitydecreases with increasing price.When respondents are informed about the use of nanotechnol-ogy for increasing the bioavailability of vitamin D (third column ofTable 4), the constant reduces and becomes insignicant. The onlysignicant explanatory variables are now thefood choice motivenutrient content, institutional trust and price.ForjuiceB, packagedinabottleprotectingcontentfromUV-light to secure vitamin C content, we see that WTP before the infor-mationabout nanotechnologyisdependent onthefoodchoicemotivenutrientcontent, trustandprice(fourthcolumn). Afterinformationabout nanotechnology(fthcolumn), heterogeneitybetween the sexes increases and men respond more positively tothe informationas comparedtowomen. The nutrient contentmotive turns out to be less important and insignicant. Again trusthas a positive inuence on WTP. Finally price remains an impor-tant explanatory variable.UsingrelatedestimateswithoutexplanatoryvariablesallowsfortheestimationofWTP, asreportedinTable5. MeanWTPishigher for juice A with $1.900 before information about nanotech-nology and $1.655 after the information about nanotechnology, ascompared to juice B, where the WTP is $1.754 and $1.326, respec-tively. In both cases, information about the use of nanotechnologyis deemed bad news.It can be noted that the original benet perceived for product Ais higher. This may be related to the information given about prod-uct A (ghting ubiquitous vitamin D deciency in Canada), whileproduct B may be perceived as less benecial given the easily alter-nativesourcesofvitaminC. ThiscoincideswithresultsbyZhou(2013) that show a lower WTP for nanopackaging in a hypotheticalchoiceexperiment conductedonUSconsumers comparedtoafunctional health benet offered by nanodrops. The lower prefer-encefor nanopackagingmayalsoexplainwhyWTPfor juiceBdecreases by a larger amount as compared to juice A when respon-dents learn that the juice is produced by means of nanotechnology.Indeed, thedecreaseinmeanWTPis muchlarger for juiceB($0.428) than juice A ($0.245).Tables 6 and 7 show the results for the German online survey.The sample size reduces to 740 respondents due to missing data.Results arefairly similarcompared withCanada. Malecomparedto female respondents have a higher WTP (not signicant for JuiceAbeforeNanoinformation). PriceisafoodchoicemotivethatreducesWTPforJuiceBwithoutNanoinformationwhereasthemotivenutrition(M_nutr) increasesit for JuiceA. FinallytrustTable 3Risk perception in Canada and Germany, online survey.Canada GermanyTotal Low Trust High Trust Total Low Trust High TrustMean(Std.Dev.)Mean(Std.Dev.)Mean(Std.Dev.)Mean(Std.Dev.)Mean(Std.Dev.)Mean(Std.Dev.)Vitamin and mineral food supplements 2.21(0.84) 2.25(0.84) 2.14(0.84) 2.73(1.04) 2.86(1.08) 2.61(0.98)***Foods enriched with vitamins or minerals 2.17(0.84) 2.18(0.83) 2.14(0.85) 2.53(0.97) 2.63(1.02) 2.44(0.91)**Foods grown or treated with pesticides and otherchemicals3.84(1.02) 3.99(1.00) 3.65(1.02)***4.41(0.92) 4.50(0.93) 4.33(0.91)***GMOs 3.53(1.15) 3.74(1.11) 3.27(1.15)***3.95(1.12) 4.03(1.16) 3.87(1.08)***Irradiated food 3.52(1.15) 3.63(1.11) 3.37(1.17)**4.29(0.97) 4.35(0.99) 4.24(0.95)**Preservatives and articial coloring 3.54(1.00) 3.79(0.96) 3.23(0.98)***3.62(0.97) 3.75(1.04) 3.49(0.95)***Meat/sh containing hormones and antibiotics 3.83(0.99) 4.04(0.92) 3.57(1.02)***4.44(0.88) 4.56(0.83) 4.34(0.90)***Foods that are based on nanotechnology 3.13(1.13) 3.44(1.12) 2.82(1.05)***3.66(1.13) 3.82(1.13) 3.49(1.11)****Refer to the difference between high and low samples being signicant at the 10% level as tested by a MannWhitney U-test per country.**Refer to the difference between high and low samples being signicant at the 5% level as tested by a MannWhitney U-test per country.***Refer to the difference between high and low samples being signicant at the 1% level as tested by a MannWhitney U-test per country.Table 4Parameters of logistic one-and-a half bound estimation, Canada (N = 510).Juice A (Vitamin D enriched) Juice B (Vitamin C protected)Without Nano With Nano Without Nano With NanoConstant 2.747***(0.586) 1.010 (0.580) 1.777***(0.578) 0.443 (0.614)Trust 0.331***(0.116) 0.387***(0.115) 0.313***(0.116) 0.364***(0.128)Gender (male = 1) 0.068 (0.168) 0.124 (0.173) 0.198 (0.169) 0.339*(0.187)M_price 0.253**(0.106) 0.020 (0.111) 0.081 (0.105) 0.028 (0.121)M_nutr 0.500***(0.140) 0.326**(0.145) 0.342**(0.141) 0.245 (0.155)M_nat 0.030 (0.107) 0.107 (0.113) 0.032 (0.106) 0.169 (0.122)Price 2.265***(0.149) 1.706***(0.136) 1.996***(0.147) 1.512***(0.145)LogL 527.337 502.689 514.423 442.404Standard errors are in parentheses.*Refer to signicance at the 0.10 level.**Refer to signicance at the 0.05 level.***Refer to signicance at the 0.01 level.Table 5Mean WTP and condence intervals, Canada (N = 510).Juice A (Vitamin Denriched)Juice B (Vitamin Cprotected)WTP withoutNano1.900***(1.824; 1.975) 1.754***(1.665; 1.843)WTP with Nano 1.655***(1.545; 1.765) 1.326***(1.167; 1.486)Condence intervals are in parentheses.***Refers to signicance at the 0.01.80 J. Roosen et al. / Food Policy 52 (2015) 7583hasasignicant andpositiveimpact inall fourestimations. Incomparison to Canada it is interesting to note that the parameteralmost doubles from before to after information about the use ofnanotechnology. For example, the parameter for the trust variablefor Juice A (Juice B) changes from 0.331 (0.313) in the case withoutinformation about theuse ofnanotechnologyto0.387 (0.364)inthe case with information for Canada, while it changes from0.418(0.363)to0.709(0.672)forGermany. Thismeansthatinthe German sample trust is much more important for the accep-tance of nanotechnology. Table 7 shows the mean WTP estimatesfor Germany. Estimated condence intervals are lower in compar-ison to Canada. However, as different prices were used in the WTPelicitation, a direct comparison is not possible. Comparing the WTPfor the products, Juice A receives a higher WTP initially. However,after information about the use of nanotechnology WTP for Juice Bis larger than WTP for Juice A. This contrasts to results for Canadaand for results of a US study (Zhou, 2013).Distribution of WTP for functional attribute before and afterinformation on nanotechnology based on lab experimentsTo gain further insights into the interplay of trust with shifts inWTP, we analyze data collected in a laboratory experiment.Because WTP has been elicited using a price list, empirical distribu-tions of WTPcanbedescribed. TheWTPdatais describedinTable8. AverageWTPis1.154forJuiceAand1.004forJuiceB(including zero bids). WTP decreases withinformationabout theuseofnanotechnology. Hencealsointheexperiment, theuseofthetechnologyisevaluatedasbadnews. JuiceBisrejectedbymoreconsumers, however, thenumberofconsumersdecreasingtheir bid to zero is not as large (9) as in the case of Juice A (41).Splittingthesampleintoconsumerswithlowandhightrust, itcan be observed that the low trust consumers are more likely todecrease their bid to zero when they learn about the applicationof nanotechnology, even if only slightly. E.g. for juice A the numberof nonengaged consumers increases from 11 to 33 in the sample oflow trust respondents and from 7 to 26 in the sample of high trustrespondents.Fig. 2showstheempirical cumulativedistributionfunctions(cdf) of WTP before and after information about the use of nano-technologyforjuiceAandB. Valuesbelowthelowerboundofthe price list at0.90 are coded as a WTP of zero,whereas yes-responsestoapriceof 1.80arerecordedas1.80. Toexplainthe gure, we carefully discuss the graph for Juice A and the lowtrustsample(upperleft). Asshownbythered, solidline, about12.5%of thesamplehasaWTPbelow0.90. AstheWTPvalueincreases, theshareofrespondentswithaWTPofagivenvalueandbelowincreasesstepwise, asweaskedforWTPin10centsintervals. The blue, dotted line shows the cdf after the informationthat Juice A was manufactured by means of nanotechnology. Theupward and left shift of the step functions indicates the increasein theproportion of respondents denoting a zero WTP (now 50%of the sample). Indeed the maximum WTP observed is 1.60, whileit was1.80 before the information.In all cases the WTP distribution shifts leftwards after informa-tion about the use of nanotechnology. Comparing the low trust andTable 6Parameters of logistic one-and-a half bound estimation, Germany (N = 740).Juice A (Vitamin D enriched) Juice B (Vitamin C protected)Without Nano with Nano Without Nano With NanoConstant 0.406 (0.536) 0.836 (0.638) 1.334**(0.567) 0.287 (0.579)Trust 0.418***(0.118) 0.709***(0.131) 0.363***(0.122) 0.672***(0.127)Gender (male = 1) 0.182 (0.147) 0.603***(0.171) 0.369**(0.152) 0.579***(0.163)M_price 0.056 (0.094) 0.084 (0.107) 0.232**(0.099) 0.156 (0.104)M_nutr 0.204*(0.121) 0.138 (0.138) 0.033 (0.121) 0.127 (0.132)M_nat 0.010 (0.098) 0.147 (0.108) 0.152 (0.099) 0.118 (0.103)Price 2.583***(0.150) 2.315***(0.179) 2.791***(0.189) 2.276***(0.161)LogL 663.173 525.655 596.815 576.064Standard errors are in parentheses.*Refer to signicance at the 0.10 level.**Refer to signicance at the 0.05 level.***Refer to signicance at the 0.01 level.Table 7Mean WTP and condence intervals Germany (N = 740).Juice A (Vitamin Denriched)Juice B (Vitamin Cprotected)WTP withoutNano0.851***(0.793; 0.909) 0.738***(0.677; 0.799)WTP with Nano 0.502***(0.403; 0.601) 0.597***(0.510; 0.684)Condence intervals are in parentheses.***Refers to signicance at the 0.01.Table 8Description of WTP data, experiment, Germany.Juice A (Vitamin D enriched) Juice B (Vitamin C protected)Without Nano With Nano Without Nano With NanoAverage WTP incl. zero bidsa1.154 0.756 1.004 0.878 excl. zero bids 1.298 1.280 1.290 1.283Number of Zero/Non-zero bidsFull sample 18/119 59/80 36/101 45/93Low trust 11/54 33/32 22/42 26/38High trust 7/65 26/48 14/59 19/55aBids with no responses at all prices have been coded as 0.00. Bids with yes responses at all prices are coded as1.80.J. Roosen et al. / Food Policy 52 (2015) 7583 81high trust samples for juice A and juice B, it becomes obvious thatbefore information (solid lines) the distribution of WTP for the hightrust sample is to the right of the one for the low trust sample. Inall four graphsit isconsistentlyobservedthat theinformationabouttheuseofnanotechnologyleadstoaleftwardshiftinthecumulative distribution function. Based on this descriptive resultit appears that the distribution of WTP after information isrst-degree stochastically dominated by that frombefore the infor-mation. This means that information about the use of a new tech-nologyleadstoawelfaredecreaseforconsumers. Asbefore, theinformation is considered bad news. Secondly, we observe thathigh levels of trust not only increased (rightward shift) the initialWTP distribution of respondents but also reduced the size of theleftward shift in the distribution due to new information. Hence,trust seems to yield a double dividend, in the form of a higher ori-ginal acceptance of nanotechnology and less reduction after addi-tional information is provided about the technology.We analyze the experimental data further by estimating a tobitmodel of theWTPdistribution. Table9showstheresults. Hereobservationsbeforeandafter theinformationabout theuseofnanotechnology are pooled together and the variable Informationindicates a dummy variable that equals unity after the informationtreatment. The results in Table 9 show that the information lowersWTP in both cases (Juice A and B) but only signicantly so for JuiceA. This is similar to a study in Switzerland by Siegrist et al. (2007)wherefoodpackagingapplicationsofnanotechnologywereper-ceived more favorably compared to nanotechnology food applica-tions. Trust hasapositiveandsignicantimpactinbothcases.Againthefoodchoicemotivesprice andnaturalness lowersWTP(thoughnot alwayssignicantly). Thefoodchoicemotivenutrient content has a mixed impact on consumers WTP.ConclusionThe results of this paper show that the use of nanotechnologyraises concerns in consumers minds about either juice or packag-ingof juiceproducedwithnanotechnology. Itisunclearatthispoint whether these concerns are related to lack of awareness ofnanotechnology in the general public or lack of awareness of nano-technologyuses inthefoodindustry. DecimaResearch(2006)reported generally favorable attitudes toward the use of nanotech-nology, in general, in the Canadian society with approximately 47%of the population, at that time, having some familiarity with nano-technology. Similarly, the Federal Institute of Risk Assessmentreported based on a survey in 2008 that 50% of Germans have somefamiliarity with nanotechnology, but that the level of acceptanceforpackagingtechnologiesishigherascomparedtoforticationtechnologies (Zimmer et al., 2008).As shown in our results, trust can often ameliorate risk percep-tions in the presence of scientic uncertainty about a technology orproducts. In our case trust was shown to reduce the levels of con-cern and to restore condence in the market. As suggested in the00.20.40.60.810 0.5 1 1.5Aer Informaon A before Informaon AWTPCDFJuice A - Low Trust (N=64)00.20.40.60.810 0.5 1 1.5Aer Informaon B before Informaon BWTPCDFJuice B - Low Trust (N=63)00.10.20.30.40.50.60.70.80.910 0.5 1 1.5Aer Informaon A Before Informaon AWTPCDFJuice A - High Trust (N=75)00.20.40.60.81. 0 0 5 1 1.5Aer Informaon B Before Informaon BWTPCDFJuice B - High Trust (N=75)Fig. 2. Cumulative distribution of WTP for juices measured in experiment, Germany.Table 9Parameters of Tobit estimation, experiment, Germany.Juice A (Vitamin Denriched)Juice B (Vitamin Cprotected)N = 268 N = 267Constant 1.231***(0.199) 1.601***(0.216)Information 0.206***(0.042) 0.050 (0.045)Trust 0.069**(0.033) 0.061*(0.037)Gender(male = 1)0.020 (0.043) 0.028 (0.046)M_price 0.039 (0.030) 0.112***(0.032)M_nutr 0.070**(0.033) 0.025 (0.037)M_nat 0.085***(0.026) 0.030 (0.028)Sigma 0.322***(0.018) 0.350***(0.021)LogL 132.934 152.556Standard errors are in parentheses.*Refer to signicance at the 0.10 level.**Refer to signicance at the 0.05 level.***Refer to signicance at the 0.01 level.82 J. Roosen et al. / Food Policy 52 (2015) 7583sociological and economic literature, trust reduces the vulnerabilitythat consumers feel and therefore lowers the level of self-protec-tion measures such as foregoing the benets of a new technology.As for our conjectures stated at the end of the section The roleoftrust wecanconcludethattheinformationabouttheuseofnanotechnologyis consideredbadnews. This is aconsistentresultforCanadaandGermanyintheonlinesurveysandintheexperiment for Germany. It hasalsobeenfoundfor theUSbyZhou (2013) and for Switzerland by Siegrist et al. (2007).Concerningregulatorydiscussionsaroundmandatorylabelingof productscontainingnanomaterials, thisnanolabelingmeansbad news for many consumers and marketing of nanotechnologyinnovationsarelimitedtothosewhotrust theinvolvedactors.Against thebackgroundof limitedscienticcertainty, however,thoseconsumerswhoareratherskepticalserveaswatchdogsofthefoodsystem(Berg, 2004) andshouldthereforebeactivelyinvolved in the regulatory process by measures of citizen participa-tion. The importance of trust is also an expression of an increasingcomplexity in the food market, making individual consumersincreasingly dependent on the decisions of those bodies responsi-ble bodies for regulation and oversight of the food sector.AcknowledgmentThe authors thank Alberta Agriculture and Rural Development,the German Research Foundation (DFG) and the French ResearchAgency (ANR) for funding.ReferencesArrow, K., 1974. The limits of organization. WW Norton Co., New York.Berg, L., 2004. Trust in food in the age of mad cow disease: a comparative study ofconsumers evaluation of food safety in Belgium, Britain and Norway. Appetite24, 2132.Berg, J., Dickhaut, J.W., McCabe, K.A., 1995. Trust, reciprocity, andsocial history.Games Econom. Behavior 10 (1), 122142.Bieberstein, A., Roosen, J., Marette, S., Blanchemanche, S., Vandermoere, F., 2013.Consumer choices for nano-food and nanot-packaging in France and Germany.Eur. Rev. Agr. Econom. 40 (1), 7394.Bradach, J.L., Eccles, R.G., 1989. Price, authority and trust: From ideal types to pluralforms. Ann. Rev. Sociol. 15, 97118.Cooper, J.C., Hanemann, M., Signorello, G., 2002. One-and-one-half-bounddichotomous choice contingent valuation. Rev. Econom. Stat. 84 (4), 742750.Costa-Font, M., Gil, M., Traill, W.B., 2008. Consumers acceptance, valuation of andattitudes towards genetically modied food: review and implications for foodpolicy. Food Policy 33, 99111.deJonge, J., vanTrijp, J.C.M., vanderLans, I.A., Frewer, L.J., 2008. Howtrustininstitutions and organizations builds general consumer condence in the safetyof food: a decomposition of effects. Appetite 51 (2), 311317.Dionne, G., Eeckhoudt, L., 1985. Self-insurance, self-protectionandincreasedriskaversion. Econom. Lett. 17, 3942.Earle, T.C., 2000. Trustinriskmanagement. Amodel basedreviewof empiricalresearch. Risk Anal. 30 (4), 541574.Earle, T.C., Cvetcovich, G.T., 1995. Social Trust. TowardaCosmopolitanSociety.Praeger Publishers, Westport, CT.Eeckhoudt, L., Gollier, C., 2005. The impact of prudence on optimal prevention. Econ.Theor. 26, 989994.Ehrlich, I., Becker, G., 1972. Market insurance, self insurance and self protection. J.Polit. Econ. 80, 623648.Erikson, E.H., 1953. Growth and crises of healthy personality. In: Kluckhohn, C.,Murray, H. (Eds.), PersonalityinNature, SocietyandCulture, 2nded. Knopf,New York.European Commission, 2012. Communication fromthe Commission to theEuropean Parliament, the Council and the European Economic and SocialCommittee. SecondRegulatoryReviewonNanomaterials. 572Final, Brussels. (accessed 08.09.13).Federal Statistical Ofce, 2013tatistisches Jahrbuch. (accessed 13.07.14).Fischler, C., 1988. Food, self and identity. Soc. Sci. Inform. 27 (2), 275292.Frewer, L.J., Howard, J.C., Hedderley, D., Shepherd, R., 1996. What determines trustininformation aboutfood-relatedrisks? Underlying psychologicalconstructs.Risk Anal. 16 (4), 473486.Gambetta, D., 1988. Can we trust trust? In: Gambetta, D. (Ed.), Trust: Making andBreaking Co-operative Relations. Basil Blackwell, Oxford, pp. 213237.Giddens, A., 1990. The Consequences of Modernity. Polity, Cambridge.Glaeser, E.L., Laibson, D.I., Scheinkman, J.A., Soutter, C.L., 2000. Measuringtrust.Quart. J. Econ. 115 (3), 811864.Government of Canada, 2013. Nanoportal. (accessed 08.09.13).Hanemann, W.M., Loomis, J., Kanninen, B., 1991. Statistical efciencyof doubleboundeddichotomous choice contingent valuation. Am. J. Agric. Econ. 73,12551263.Hudson, J., 2006. Institutional trust and subjective well-being across the EU. Kyklos59 (1), 4362.Kasperson, R.E., Golding, D., Tuler, S., 1992. Social distrust asafactor insitinghazardous facilities and communicating risks. J. Soc. Issues 48 (4), 161187.Kjrnes, U., Harvey, M., Warde, A., 2007. Trust in food. A comparative andinstitutional analysis. Palgrave Macmillan, New York.Luhmann, N., 1993. Risk: a sociological theory. A. de Gruyter, New York.Luhmann, N., 2000. Vertrauen, UTB, Englishtranslationavailableas Trust andPower, 4th ed. Wiley, Chichester, Stuttgart, 1979 (First published 1968).Matin, A.H., Goddard, E., Vandermoere, F., Blanchemanche, S., Bieberstein, A.,Marette, S., Roosen, J., 2012. Do environmental attitudes and food technologyneophobia affect perceptions of the benets of nanotechnology? Int. J. Consum.Stud. 36, 149157.McEvily, B., Radzevick, J.R., Weber, R.A., 2012. Whom do you distrust and how muchdoes it cost? An experiment on the measurement of trust. Games Econ.Behavior 74, 285298.Meyer, R., Liebe, U., 2010. Are the afuent prepared to pay for the planet?Explaining willingness to pay for public and quasi-private environmental goodsin Switzerland. Popul. Environ. 32 (1), 4265. http://dx.doi.org/10.1007/s11111-010-0116-y.Nocella, G., Hubbard, L., Scarpa, R., 2010. Farmanimal welfare, consumerwillingnesstopay, andtrust:resultsof across-national survey. Appl. Econ.Perspect. Policy 32 (2), 275297. http://dx.doi.org/10.1093/aepp/ppp009.Nooteboom, B., 1996. Trust, opportunismandgovernance:aprocessandcontrolmodel. Organ. Stud. 17 (6), 9851010.Oberdrster, G., Oberdrster, E., Oberdrster, J., 2005. Nanotoxicology: an emergingdiscipline from studies of ultrane particles. Environ. Health Perspect. 113 (7),823839.Oh, H., Hong,J.,2012. Citizens trust in government and their willingness-to-pay.Econ. Lett. 115 (3), 345347. http://dx.doi.org/10.1016/j.econlet.2011.12.010.Pelley, J., Saner, M., 2009. International approaches to the regulatory governance ofnanotechnology. Regulation papers, Carleton University, School of Public Policyand Administration. (access, 14.09.13).Renn, O., Rohrmann, B., 2000. Cross-cultural risk perception. A survey of empiricalstudies. Kluwer Academic Publishers, Dordrecht, Boston and London.Decima Research, 2006. Emerging Technologies Tracking Research, report preparedfor Industry Canada, June, 50 pages. (accessed30.05.13).Rothschild, M., Stiglitz, J., 1970. Increasing risk: I. A denition. J. Econ. Theory 2 (3),225243.Rotter, J.B., 1967. A new scale for the measurement of interpersonal trust. J. Pers. 35,651665.Rousseau, D.M., Sitkin, S.B., Burt, R.S., Camerer, C., 1998. Not so different after all: across-discipline view of trust. Acad. Manag. Rev. 23, 393404.Sassatelli, R., Scott, A., 2001. Novel food, new markets and trust regimes: responsesto the erosion of consumerscondence in Australia, Italy and UK. Eur. Soc. 3 (2),231244.Siegrist, M., Cvetkovivh, G., Roth, C., 2000. Salient value similarity, social trust, andrisk/benet perception. Risk Anal. 20 (2), 353362.Siegrist, M., Cousin, M., Kastenholz, H., Wiek, A., 2007. Public acceptance ofnanotechnologyfoodsandfoodpackaging:theinuenceof affectandtrust.Appetite 49, 459466.Slovic, P., 1999. Trust, emotion, sex, politics, andscience: surveying the risk-assessment battleeld. Risk Anal. 19 (4), 689700.Steptoe, A., Pollard, T.M., Wardle, J., 1995. Development of a measure of the motivesunderlyingtheselectionoffood:thefoodchoicequestionnaire. Appetite25,267284.Vandermoere, F., Blanchemanche, S., Bieberstein, A., Marette, S., Roosen, J., 2010.The morality of attitudes toward nanotechnology: about god, techno-scienticprogress, and interfering with nature. J. Nanopart. Res. 12 (2), 373381.Visschers, H.M., Meertens, R.M., Passchier, W.F., deVries, N.K., 2007. How does thegeneral public evaluate risk information? The impact of associations with otherrisks. Risk Anal. 27 (3), 715727.Wang, B., Feng, W.-Y., Wang, T.-C., Jia, G., Wang, M., Shi, J.-W., Zhang, F., Zhao, Y.-L.,Chai, Z.-F., 2006. Acute toxicity of nano- and micro-scale zinc powder in healthyadult mice. Toxicol. Lett. 16, 115123.Zhou, G., 2013. NanotechnologyintheFoodSystem: ConsumerAcceptanceandWillingness to Pay. Theses and Dissertation, Agricultural Economics, Paper 10. (accessed 10.09.13).Zimmer, R., Hertel, R., Bl, G.-F., 2008. Wahrnehmung der Nanotechnologie in derBevlkerung. Berlin: Bundesinstitut fr Risikobewertung. (accessed 21.05.13).J. Roosen et al. / Food Policy 52 (2015) 7583 83