Elements of public trust in the American food system: Experts, organizations, and genetically modified food

Food Policy 41 (2013) 145–154

Contents lists available at SciVerse ScienceDirect

Food Policy journal homepage: www.elsevier.com/locate/foodpol

Elements of public trust in the American food system: Experts, organizations, and genetically modified food John T. Lang ⇑ Department of Sociology, M-26, Occidental College, Los Angeles, CA 90041, USA

a r t i c l e

i n f o

Article history: Received 29 September 2012 Received in revised form 20 December 2012 Accepted 20 May 2013 Available online 15 June 2013 Keywords: Consumers and consumption Institutions Environment and technology Genetically modified food Risk Trust

a b s t r a c t The increasing scientific and technical complexity in the American system of food production, exemplified in this article by genetically modified food (GMF), provides a useful case with which to understand public trust in experts and organizations involved in an emerging technology. Stating that the public will judge claims about GMF based on trust in their sources brings about the question of how the public decides to trust particular sources. I use data from a mail survey to evaluate the elements of trust for a range of groups connected to the U.S. food industry. The results point to organizationally variable and dependent perceptions of trust rather than a stable set of elements, suggesting that scholars might focus productively on the ways elements of trust are distributed. Ó 2013 Elsevier Ltd. All rights reserved.

Introduction Reliance on sophisticated technologies in our daily lives forces us to come to terms with problems of institutional legitimacy and trust in myriad organizations and experts. Social scientists have repeatedly recognized that the uncertainty and vulnerability of social interactions becomes paramount without trust between social actors (e.g., Lewis and Weigert, 1985; Mollering, 2006; Seligman, 2000). Moreover, trust becomes more urgent in ‘‘contingent, uncertain and global conditions’’ (Misztal, 1996, p. 9) because people rely on trusted relationships to deal with ‘‘uncertainty and vulnerability’’ (Heimer and Cook, 2001). The increasing scientific and technical complexity in the American system of food production, exemplified in this article by genetically modified food (GMF), amplifies public reliance on the trustworthiness of a complex set of institutional actors. Scientific and technological developments produce benefits, but they also produce uncertainty, potentially disastrous failures, and harmful side effects (Erikson, 1994). While trust may increase tolerance for ambiguity and open new possibilities for cooperative interactions (Luhmann, 1979), the more that industry produces using advanced technologies, the higher the potential for institutional failure resulting from lack of competence or lack of social and fiduciary responsibility (Clarke and Short, 1993; Freudenburg, 1993; Sapp et al., 2009). Indeed, failures have resulted in severe

⇑ Tel.: +1 323 259 1310; fax: +1 323 341 4977. E-mail address: [email protected] 0306-9192/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodpol.2013.05.008

consequences even when experts and organizations take great care (Clarke, 2006; Weick and Sutcliffe, 2001). Because key institutional actors occupy a crucial link between trust and risk, and between acceptance and uncertainty, they shape how science proceeds. Furthermore, the increasingly global scope and complexity of institutions, organizations, and technological systems make them impenetrable to experts as well as ordinary people (Perrow, 1999; Sztompka, 1999). As a result, trust in ‘‘expert systems’’ is an ever-present requirement of modern societies (Giddens, 1990). It is, however, particularly tricky for novel or emerging technologies because expectations about the future performance of institutional actors often cannot be based on prior experience. This form of trust requires the public to trust experts and organizations using uncertain criteria, not based on interpersonal experience or direct knowledge. From the public’s point of view, emerging technologies add uncertainty. For example, people wonder if the innovation process really reflects their values and interests (Earle and Cvetkovich, 1995; Siegrist et al., 2000). People also become concerned that the technology developers, users, and regulators might not be competent enough to make the right decisions and that the wrong decisions will cause harm (Sjoberg, 1999). Moreover, key institutional actors might communicate overly biased information and sway public opinion (Freudenburg, 1992; Rousu et al., 2007). For the general population, therefore, part of the hesitation related to innovation may be uncertainty about the behavior of the social systems, organizations, and experts involved with the technology. Moreover, debates about appropriate public engagement strategies and the role of moral, ethical, and social considerations in

146

J.T. Lang / Food Policy 41 (2013) 145–154

scientific decision-making bring forward a wide-range of actors. In this sense, where there are competing claims, people must decide whose scientific claims to accept and whose claims to reject. For example, scientists devise and evaluate the technology used to help farmers grow crops. Food manufacturers use these crops to produce marketable food products. Grocers and grocery stores sell these products to the public. Myriad, and often disjointed, government agencies attempt to regulate the quality and safety of food. Environmental organizations act as watchdogs trying protect public health and the environment and are likely initial sources of information for many. Because of the multitude of institutional actors with a claim about GMF, how and why people trust those involved with this emerging technology is an important and interesting avenue of exploration. All of this raises the question of how the public will judge genetically modified food, particularly when there are contradictory claims about the technology. One plausible answer is that they will judge such claims, at least in part, by the extent to which they trust their sources (Freudenburg, 1993). So, what are the key elements of public trust in the American food system? And, are the key elements of public trust in the American food system the same for all institutional actors? In this article I attempt to answer these questions by reporting which of six organizational actors the public trusts regarding GMF, examining how public judgments of trust vary across each of the six actors. Second, I explore if and how elements of trust vary by actor. This article makes a contribution by testing potential variation in the elements of public trust in the U.S. food system. The uneven findings across actors suggest that the relationship between trust elements and social actors is more complicated than previously thought.

Background: genetically modified food The genetic modification of food has rapidly progressed from experimental crop science to a commonplace foodstuff. In 1996, the first year of commercialization, six countries grew genetically modified food crops; by 2010, twenty-nine countries grew genetically modified crops (James, 2010). Worldwide, more than threequarters of all soybeans, more than half of cotton, more than a quarter of maize (corn), and one-quarter of all canola grown is genetically modified (James, 2010). U.S. farmers have embraced genetically modified crops at an even higher rate. Roughly 90% of all soybeans, more than three-quarters of cotton, and more than 80% of maize planted in the United States is genetically modified (United States Department of Agriculture, 2011). Food manufacturers use these commodity crops and their derivatives – such as high-fructose corn syrup, cornstarch, soy lecithin, as well as canola, soybean, and cottonseed oils – as ingredients in a vast array of processed foods. As a result, the Pew Initiative on Food and Biotechnology (2005) estimated that three-fourths of all processed foods in the U.S. contained a GM ingredient. Given that the genetic modification of key commodity crops used in the majority of processed foods has greatly increased since that report, the three-fourths estimate is likely conservative. Because the United States does not require labeling genetically modified food as such, estimates that are more recent and more accurate are not available. Moreover, the lack of labeling makes GMF ‘‘invisible ‘‘ leaving the American public unaware or unsure whether they are consuming GMF (Pew, 2005; Einsiedel, 2009). It should not be surprising that the American public remains unaware of the presence of genetically modified ingredients in their food. In general, most Americans know little about how many or what kinds of organizational actors produce, process, transport, or prepare their food for sale (Vileisis, 2007). That is not to say that

the public is entirely ignorant. Rather, knowledge is partial and imperfect because, given their displacement from the food system, the public relies on numerous institutional actors for their daily sustenance. The public has little choice but to trust the system of food production that involves ‘‘whole armies of specialists, most of whom have areas of expertise that we may not be competent to judge, and many of whom we will never even meet, let alone have the ability to control’’ (Alario and Freudenburg, 2003, p. 200). The lack of a unified, hierarchical command structure to delineate authority and power among these specialist organizations likely also means that the public will sometimes perceive some groups as competing actors; other times the public may see them as coordinating. For example, the United States Department of Agriculture (USDA) has a dual mandate. Its first duty is to support the agricultural community and promote their products in the United States and abroad. Its second duty is to ensure the safety of the American food supply. Having to promote an industry while also policing it is a precarious position; the USDA is sometimes aligned with food producers and is sometimes placed in opposition to them. Assumptions that organizations influence trust judgments in isolation from one another appear overly deterministic; so does the assumption that the public should trust institutional actors based on identical criteria. For example, we may trust scientific knowledge about genetically modified food, but might believe that scientists value scientific knowledge and novelty differently than the public. Or, we may believe that farmers are honest, but they might not have the competence to predict potentially negative effects of GMF. Despite a relatively muted response in the United States, there has been some consumer concern and backlash, particularly in Europe, surrounding GMF (Schurman and Munro, 2010). Although typically presented as a scientific issue, GMF tends to inspire emotional reactions that speak to global issues of economic and cultural power (Annear, 2004; Falkner, 2007). Social activism and political–cultural context may help explain the efficacy of antiGMF strategies in some locations (Schurman, 2004) but there is more to it. The controversies surrounding GMF are, in many ways, a proxy debate for broader issues of social and political power, democratic practice, and corporate responsibility (Jasanoff, 2005). In many ways, the issues surrounding genetically modified food presents a useful case study of public trust in experts and organizations involved in an emerging technology.

Experts and organizations Given the complex range of scientific, structural, and organizational barriers to understanding GMF, it is nearly impossible for an ideal–typical rational person to decide whom to trust. As such, the risk involved in genetically modified food is related to the social and institutional organization surrounding its production (Allum, 2007). This is consistent with the view that new technologies are more than more than a collection of scientific and technical advances. In many ways, new technologies are represented more completely as sociotechnical systems incorporating social institutions that develop, implement, operate, monitor and regulate these systems (Bijker et al., 1989). These systems, even for something as mundane as breakfast cereal, are spread out across global commodity chains involving an extensive network of actors and institutional actors (Schurman and Munro, 2009). In contrast to the non-expert public, experts and organizations maintain a crucial position in these global arrangements as the link between trust and risk, where they frame social problems, defining potential risks and imagined responses (Clarke, 1999). These actors help decide which research questions to ask and answer. In their risk communication messages, the organizations take on social

J.T. Lang / Food Policy 41 (2013) 145–154

and ethical responsibilities by acting as experts, creating and circulating knowledge about technology. Trust allows these groups to maintain their advisory positions in society. The public accepts the risks of unknown technologies because these trusted experts endorse, promote, and regulate them. Social relations are mainly responsible for the production of trust in economic life and people generate trust when agreements are embedded within a larger structure of personal relations and social networks (Granovetter, 1985; Mizruchi et al., 2006). Social structure is important not only for the formation of social capital (e.g., Fukuyama, 1995; Putnam, 1993) but also for the generation of trust itself. It allows for more rapid proliferation of obligations and expectations, imposes sanctions on the failure to meet an obligation, and helps to generate reputation (Coleman, 1990). In addition, familiar and stable relationships can relieve members of a given social structure of the uncertainty about other people’s motivations and anxiety about others’ actions not meeting their expectations. Moreover, because emerging technologies are embedded in an economic and social context, an assessment of that technology must include an assessment of the technology’s producers as well as the various institutional actors involved with the technology (Wynne, 1980). As a result, the public understanding of novel risks is often contingent on the trust granted to the institutional actors involved with the technology. Trust Given the robust effects of trust, social scientists continue to debate its nature and role in modern society (Delhey et al., 2011; Earle et al., 2010; Hardin et al., 2009; Misztal, 2011; Paxton, 2007; Yamagishi, 2011). Scholars generally believe that trust is subject to specific situations and the actors involved (Barber, 1983; Gambetta, 1988; Hardin, 2002). Researchers who study technological and environmental hazards typically attempt to determine the dimensions of trust (e.g., Allum, 2007; Johnson, 1999; Renn and Levine, 1991), measure increasing or decreasing trust (e.g., James and Marks, 2008; White and Eiser, 2005), or explore the proposition that trust is easier to destroy than create (e.g., Cvetkovich et al., 2008; Poortinga and Pidgeon, 2004; Slovic, 1993). Accounts frequently discuss several elements of trust that reflect competence and caring, including honesty, shared values, the ability to predict effects, knowledge, and the ability to determine importance (e.g., Barber, 1983; Earle and Cvetkovich, 1995; Frewer et al., 1996; Metlay, 1999; Poortinga and Pidgeon, 2003; Renn and Levine, 1991; White and Eiser, 2005). Although there is overlap in the descriptions and apparent equivalence to some degree, studies rarely, if ever, use the same measures. These accounts of trust often center on individuals (e.g., Cook, 2001; Hardin, 2002; Seligman, 2000) with special reference to the conditions or processes that induce people to trust certain others (Tilly, 2005). Within the evolving literature of American reactions to GMF there has been relatively little focused research on how people answer trust questions for individual institutional actors. Most often, scholars treat trust measures as independent rather than dependent variables. But it is not reasonable to assume that the logic underlying respondents’ answers to questions about trust in GMF institutional actor groups is uniform. Different institutional actors may attract systematically higher or lower levels of trust. Conceptually and methodologically, it is unclear whether multiple criteria or elements are important for assessing trust in the institutional actors of emerging technologies. Some argue that people lack the motivation or ability to distinguish among potentially diverse elements of trust (Earle and Cvetkovich, 1995; Meijnders et al., 2009; Midden and Huijts, 2009). In contrast, Johnson and White (2010) recently found support for people being able make relatively fine-grained distinctions across elements of trust and

147

social actors. Given the exploratory nature of their research on the redevelopment of contaminated land and drinking water standards, they caution that their results may not apply for all potential hazards. Still, there are good reasons to assess multiple trust elements. Similar to the belief that context matters in trusting relationships, the multiplexity of social roles means that people must often cope with divergent and contradictory judgments about individuals (Lewicki et al., 1998; Verbrugge, 1979). For example, if a person makes a mistake, but then publicly admits the failure, others might rate them as less competent and more honest. Given that competence and honesty are both considered elements of trust, this presents a quandary if we assume trust is monolithic. Furthermore, given evidence that people use different trust repertoires in different encounters (Mizrachi et al., 2007), it is entirely plausible that ceteris paribus, the public relies on specific social actors for a specific element of trust for a given hazard. Trust in industry and activist groups I test the premise that the public uses different criteria for assessing trustworthiness depending on the organization under scrutiny, using the technology of genetically modified food (GMF) as a proxy debate for broader issues of social trust in the institutional actors of emerging technologies. For more than 30 years scholars have focused on the role of trust in explaining risk perceptions across a range of technological hazards. Wynne (1980) was one of the first to make the link between differences in lay and expert perceptions of risk and differences in the extent of trust in regulatory and scientific institutions. Since then, the relationship between trust, confidence, and risk perception has been widely investigated (e.g., Frewer et al., 1996; Peters et al., 1997; Renn and Levine, 1991). Most results show that people who trust the people who manage risk believe technology poses a lower risk; people who express no trust in the managers of risk believe technology poses a greater risk (Johnson, 1999; Metlay, 1999). Further, trust enables societies to tolerate increasing uncertainty due to progressive technological and organizational complexity by allowing specific, rather than arbitrary, assumptions about future behavior (Barber, 1983; Luhmann, 1979). As a result, for some theorists (i.e., Giddens, 1990; Seligman, 2000), trust is fundamental to the emergence and prominence of organizations and institutions in daily life. Moreover, the potential for recreancy, institutional failure resulting from lack of competence or lack of social and fiduciary responsibility, is exactly the condition that increases our reliance on trust (Freudenburg, 1993). Recent work (Sapp et al., 2009) found that the effects of fiduciary responsibility on trust consistently far outweighed those for competence across a range of actors in several parts of the U.S. food system. My expectation in this article, well supported in previous research (e.g., Huffman et al., 2004; James and Marks, 2008; Lang and Hallman, 2005), is that the public will express different levels of overall trust in each institutional actor involved with GMF. I do not, however, expect to find that respondents consistently rate elements of trust as equally relevant for all institutional actors. Schurman and Munro (2010) argue for two distinct lifeworlds – industry and activist – that help explain conflicting viewpoints about GMF. They contend that groups involved with GMF have ‘‘culturally and cognitively distinct’’ worldviews that provide distinct frames of reference, leading to divergent values (Schurman and Munro, 2010). That there might be a parallel logic where people who trust industry groups do so in ways that are distinct from those who trust activist groups is worthy of exploration. In this article, I use environmental organizations to represent activist groups and use food manufacturers, farmers, and grocery stores to represent industry. Because the industry lifeworld

148

J.T. Lang / Food Policy 41 (2013) 145–154

privileges scientific knowledge and novelty (Schurman and Munro, 2010), I hypothesize that assessments of knowledge would be most relevant to trusting industry actors. Similarly, because the activist lifeworld privileges a normative sensibility that values the environment (Schurman and Munro, 2010), I hypothesize that assessments of shared values would be most relevant to activist actors. The classification of two actors, government agencies and university scientists, is unclear as they function in many roles that span the activist and industry lifeworlds. Survey procedures and sampling I develop my account using telephone and mail survey data collected from 363 people who participated in a mail survey followup to a 1201 person, nationally representative telephone survey of Americans’ knowledge and thoughts about agricultural biotechnology conducted via random digit dialing conducted in May and June 2004 (Hallman et al., 2004). After completing the telephone survey, interviewers asked the 1201 respondents if they would be willing to complete a mail questionnaire for a $5.00 incentive. Slightly less than half (47%, 559) of the respondents agreed and gave a valid mailing address. For the 440 respondents who did not complete and return a questionnaire within 14 days, interviewers sent a second questionnaire without the gratuity. Of the 559 who originally agreed, 363 (65%) returned a completed mail survey by August. All returned mail questionnaires were keypunched with 100% verification and compiled into an SPSS dataset. The final sample size of 363 from a population of 1201 yields a 30.2% response rate according to the AAPOR RR2 definition and allows a sampling error rate of ±5.5%. Description of survey instrument Among other questions about food preferences, the self-administered mail questionnaire contained questions about trust, with closed-ended response alternatives of mainly Likert-type scales. Although the wording for each question in the survey remained constant, to minimize response-order effects respondents were randomly assigned one of six versions of the trust questionnaire. I present an outline of the topics below. Trust in organizational actors The survey instructed respondents to rate their trust in six organizational actors concerning GMF. Specifically, the instructions asked respondents ‘‘How much do you trust the following groups to make appropriate decisions about genetically modified food? (from 1 no trust to 7 complete trust)’’. These groups – environmental organizations, farmers, food manufacturers, government agencies, grocers and grocery stores, and university scientists – were selected because of their importance in the organizational field, and because surveys such as the Eurobarometer have also included basic trust measures for some of these groups, permitting future cross-cultural comparisons. Elements of trust Rather than treating trust measures as components of an aggregated trust measure, they are the focus of my analysis. Modeling the questionnaire on existing trust scholarship (e.g., Barber, 1983; Frewer et al., 1996; Metlay, 1999; Peters et al., 1997; Renn and Levine, 1991; White and Eiser, 2006) I designed five questions that evaluate the intended elements – honesty, knowledge, prediction, shared values, and importance – and tested them in a series of preliminary studies. This operationalization of trust is based on a definition of trust shared by several scholars that views trust as a belief

that the trustee is honest, posses skills and expertise within a specific domain, and shares similar salient values with the trustor. By including these items, I intend to capture much of the theoretical meaning attributed to trust in the social science literature. Although following convention, I ideally would have included several questions for each of the elements to allow for more robust measurement. However, a need to minimize respondent’s time spent completing the questionnaire precluded that possibility. Explicit ratings were measured on a scale that ranged from 1 (not at all) to 7 (completely); respondents were also given an explicit ‘‘unsure’’ option. Specifically, respondents were asked the following block of questions: ‘‘When thinking about genetically modified food, how would you rate [environmental organizations] on each of these items?’’ 1. 2. 3. 4. 5.

How honest they are. (honesty) How knowledgeable they are. (knowledge) How well they can predict potential effects. (prediction) How much they share my values. (shared values) How well they can tell which potential effects are important. (importance)

After responding to this block of questions about one actor, the survey instructions asked respondents to answer the same questions about five other organizational actors. The survey repeated this sequence until respondents had rated all six actors on all five measures. Survey respondents To obtain demographic information, I merged the mail and telephone survey data via a respondent identification number. I coded sex with female equal to 0 and male equal to 1 (36.4%). I report age as an ordinal level measure with 6 categories that range from 18 to 24 years (coded 1) to 65 and older (coded 6) (mode = 4, 45– 54 years old). I coded education as an ordinal level measure with 4 categories that ranged from less than a high school degree (coded 1) to more than a 4-year college degree (coded 4) (mode = 4, at least a 4-year college degree). I report household income (in dollars) as an ordinal level measure with 7 categories that range from less than $25,000 (coded 1) to more than $125,000 (coded 7) (mode = 4, $50,000–$74,999 yearly household income). Finally, I measure religiosity dichotomously, with church attendance of less than once per month equal to 0 and attendance at least once per month equal to 1 (58.1%). Potential respondents to the original telephone survey were selected using national random digit dialing across all 50 states. U.S. Census Bureau population estimates determined the distribution necessary for proportionate geographic coverage. The mail survey respondents were distributed throughout 44 states; there were no mail survey respondents from Alaska, Montana, Nevada, New Hampshire, North Dakota, or Vermont. The largest numbers of responses were from the most populous states including California, Texas, and New York. The socio-demographic characteristics of the non-weighted mail and telephone surveys are similar and largely reflect typical survey response biases (Krosnick, 1999). I cannot, however, discount the possibility of self-selection bias from the original population. These limitations affect the generalizability of the findings regarding levels of overall trust and trust in specific institutional representatives. I present detailed socio-demographic characteristics of respondents in Table 1. Results Previous research does not explore how responses to survey based measures of trust elements vary across institutional actors

149

J.T. Lang / Food Policy 41 (2013) 145–154

of GMF. Therefore, this article contributes to the literature by testing variations in trust judgments for six institutional actors. The survey evidence reveals that respondents exhibit a hierarchical view of institutional actors and that respondents evaluate each institutional actor using differing criteria. In general, the results illustrate that a perception of honesty is the key element of expressed trust in each of the institutional actors involved with GMF. To a lesser extent, perceived knowledge was also important for two actors – university scientists and environmental organizations. The ability to predict the potential effects of GMF, the perception of shared values, and the ability to determine important effects of GMF each were significant elements of trust for only one actor. These results are important because they show that elements of public trust are organizationally variable and dependent. The results point to a range of relationships between trust and specific sources. Perceptions of trust My first objective was to examine public perceptions of trust toward six institutional actors involved in GMF. As such, I computed mean scores and standard deviations for the explicit trust measure. Table 2 shows that trust is highest in university scientists and farmers and lowest in grocers and grocery stores and food manufacturers. It is moderate for environmental organizations and government agencies. The mean differences between each organization are statistically significant, though the overall range was narrow. The highest trust rating (4.73) is a bit better than neutral (4.0) and the lowest rating (3.21) is a bit worse than neutral. Three groups have mean trust scores above a neutral rating: respondents rated university scientists, farmers, and environmental organizations as relatively trusted regarding genetically modified food. Three groups have mean trust scores below a neutral rating: respondents rated government agencies, grocers and grocery stores, and food manufacturers as relatively untrustworthy regarding genetically modified foods. These results have implications for how skeptics and propopents of GMF might try to manage trust. These ratings are roughly

Table 1 Demographic variables. Variable

Variable label (Code)

%

Sex

Female (0) Male (1)

63.6 36.4

Age group (in years)

18–24 (1) 25–34 (2) 35–44 (3) 45–54 (4) 55–64 (5) 65 and older (6)

8.5 13.8 20.4 24.5 14.9 17.9

Level of education

= 4-Yr College Degree (4) No Answer (missing)

6.9 26.2 31.4 35.3 0.3

Household income ($)

<$25,000 (1) $25,000–$34,999 (2) $35,000–$49,999 (3) $50,000–$74,999 (4) $75,000–$99,999 (5) $100,000–$124,999 (6) >$125,000 (7) No Answer (missing)

17.9 13.8 15.2 22.3 14.0 5.2 6.1 5.5

Attend church


41.6 58.1 0.3

Note: N = 363.

Table 2 Mean and standard deviation of trust ratings.

University scientists Farmers Environmental organizations Government agencies Grocers and grocery stores Food manufacturers

N

Mean

Standard deviation

355 355 355 356 351 352

4.73 4.37 4.19 3.36 3.27 3.21

1.68 1.63 1.77 1.81 1.53 1.57

Note: Scales range from ‘‘no trust’’ (1) to ‘‘complete trust’’ (7).

consistent with results reported from other American (Lang and Hallman, 2005), European (Gaskell et al., 2003), and comparative (Priest et al., 2003) quantitative surveys. The primary institutional actors in GM food (food manufacturers and government agencies) as well as those that are most likely to have public contact (grocers and grocery stores) are less trusted than other actors. In this case, somewhat paradoxically, consumers are less likely to subject more distant actors to careful public scrutiny. Although respondents trusted some more than others, no group was overwhelmingly trusted. If the conventional wisdom that says that effective communication should come from trusted experts is true, these industry actors may not be best positioned to alleviate the public’s uncertainty about GMF. Moreover, given evidence for Bayesian updating of beliefs about GMF, where new information and prior belief are combined together to reach a new state of belief (Huffman et al., 2007), these groups are relatively disadvantaged regarding the public’s trust. Perceptions of trust elements Looking at the individual elements of helps to determine the extent to which public judgments of the trust elements vary for each of the actors. As such, I computed mean scores and standard deviations for the five elements for each organization. Except for shared values, university scientists rank highest in these five categories. Food manufacturers as well as grocers and grocery stores rate lowest in these five categories (see Table 3). Considering the mean values for the five trust elements for each organizational actor, the general hierarchy of trusted organizations is not entirely consistent with Table 2. For ease of comparison, I distinguish between those that fall above and those that fall below neutral (4.0) for each trust element. For the measure of shared values, predicting effects, and determining importance, three actors have scores above the element’s neutral rating; respondents rate university scientists, farmers, and environmental organizations relatively high on these elements. Three groups have scores below neutral; respondents rate government agencies, grocers and grocery stores, and food manufacturers as relatively low on these elements. This pattern for these three measures is identical to the general measure of trust. For the measure of honesty and knowledge, four actors have ratings above neutral. In addition to university scientists, farmers, and environmental organizations, respondents also rate grocers and grocery stores above neutral for honesty. Regarding knowledge, respondents only rate grocers and grocery stores relatively low. While respondents did not rate any actor below neutral for every element, they only rate government agencies, grocers and grocery stores as well as food manufacturers above neutral on one element each. To say that members of the public will judge scientific claims based on trust in their sources brings about the question of how the public will decide to trust particular sources. These results complicate that answer. Rather than mirroring results for the general trust measures, individual ratings for elements of trust display

150

J.T. Lang / Food Policy 41 (2013) 145–154

Table 3 Mean and standard deviation of element ratings.

University scientists Farmers Environmental orgs. Government agencies Grocers and stores Food manufacturers Averages

N

Honesty

Shared values

Prediction

Knowledge

Importance

284 279 255 287 279 285 278

5.39 5.10 4.81 3.69 4.38 3.74 4.52

4.26 4.43 4.20 3.29 3.52 3.24 3.81

5.29 3.93 4.41 3.89 3.20 3.89 4.07

5.89 4.96 5.18 4.39 3.92 4.61 4.86

5.31 4.10 4.47 3.91 3.27 3.88 4.14

(1.52) (1.52) (1.80) (1.88) (1.67) (1.85) (1.32)

(1.75) (1.67) (1.87) (1.79) (1.68) (1.75) (1.42)

(1.53) (1.89) (1.84) (1.89) (1.77) (1.84) (1.44)

(1.26) (1.50) (1.56) (1.79) (1.66) (1.70) (1.14)

(1.56) (1.87) (1.92) (1.89) (1.85) (1.85) (1.45)

Note: Scale ranged from 1 (not at all) to 7 (completely).

slightly divergent patterns. Three trust element measures display the same patterns of trustworthiness and untrustworthiness as the general measures of trust. Respondents rate organizations differently, however, for three of the measures. I should point out a few similarities across actors. University scientists, farmers and environmental organizations rate highly on all measures. In contrast, government agencies, grocers and grocery stores as well as food manufacturers rate low on four of five measures.

of education is directly related to trusting environmental organizations; age group is inversely related. Religiosity does not significantly help explain trust for any group. Because I am ultimately interested in understanding if people make relatively fine-grained distinctions across elements of trust and social actors, I calculated the simple linear regression effects of the five trust elements on explicit trust judgments. This helps determine how much the public relies on specific social actors for a specific element of trust for a given hazard. I used measures of sex, age group, educational level, household income, and religiosity to control the possible mediating influences of socio-demographic factors. Because the default expectation is that all elements of trust are relevant for all actors, I used a direct method of entry for the predictor variables. Taken as a whole, the predictive ability of the model using individual items was slightly more robust than observed in the model using overall trust. As noted in Table 6, at worst, this simple model accounts for less than one-third (30.0%) of the variance in trust in food manufacturers; at best it accounts for slightly more than half (53.2%) of the variance in trust in environmental organizations. The predictive ability of the model ranges between those two extremes for the remaining organizations. I observe distinct patterns of significance for the elements in each model. For all actors, an increase in honesty results in a larger increase in trust than any other variable. The results for the other elements are mixed. That is, the set of elements that predict trust depends on the actor. Knowledge about GM food is a significant element of trust in environmental organizations and university scientists, but not for the other groups. Perceptions of shared values as well as the ability to predict the effects of GMF are important elements of trust in grocers and grocery stores, but neither element was a significant predictor of trust for other groups. The ability to tell which potential effects are important is a significant predictor for trusting university scientists, but is not significant for other groups. The results for demographic variables are similarly mixed. Prior research (e.g., Costa-Font et al., 2008; Huffman et al., 2004) indicates that demographic variables can have significant impact of consumer acceptance or willingness to purchase GMF. In the present study, demographic variables are not significant predictors of trust for food manufacturers, farmers or grocers and grocery stores. Age is inversely related to trust for government agencies. Household income is directly related to trusting university scientists.

Predicting trust in experts and organizations The measures of general trust for each of the groups were moderately and positively correlated (p 6 0.01), the Kaiser–Meyer–Olking (KMO) statistic ranged from a low of .789 for University Scientists to a high of .994 for Environmental Organizations, and Bartlett’s test was significant (p < .001) for all groups (analyses not presented here). Given that, and the likelihood that the five elements used do not account for all of the variability possible in a measure of trust, I performed an exploratory factor analysis for each organization (see Table 4). The five elements – honesty, knowledge, prediction, shared values, importance – all load onto a single factor, suggesting that these items measure a single concept that I labeled overall trust. Using the principal axis method of extraction a one-factor solution that explains a considerable portion of the variance – ranging from a low of 59.17% for university scientists to a high of 71.56% for environmental organizations – and statistically substantiates the use of a single concept to describe the five elements. By saving computed Bartlett factor scores for overall trust, I was able to enter them into a simple linear regression model. In this model I attempted to determine how well overall trust could predict public judgments of trust in each actor. I used measures of sex, age group, educational level, household income, and religiosity to control the possible mediating influences of socio-demographic factors. This simple model accounts for less than one-quarter (24.2%) of the variance in trust in food manufacturers; at best it accounts for almost half (48.2%) of the variance in trust in environmental organizations. Overall trust was significant for all actors. Demographic variables are not significant predictors of trust for government agencies, food manufacturers, or grocers and grocery stores. Household income is directly related to trusting university scientists. Sex significantly helps explain trust in farmers. Level Table 4 Overall trust as a single factor for each of six organizations.a

University scientists Farmers Environmental orgs. Government agencies Grocers and stores Food manufacturers

Honesty

Shared values

Prediction

Knowledge

Importance

Percent variance

0.73 0.74 0.88 0.73 0.71 0.73

0.61 0.70 0.73 0.61 0.74 0.61

0.82 0.88 0.86 0.82 0.92 0.82

0.80 0.78 0.83 0.80 0.85 0.80

0.86 0.85 0.92 0.86 0.88 0.86

59.17 63.29 71.56 71.39 67.68 68.23

a Notes: Each row represents an independent factor analysis. Only one factor was extracted in each of the six factor analyses. Extraction Method: Principal Axis Factoring. The scales were coded to range from 1: ‘‘totally disagree’’ to 5: ‘‘totally agree’’.

151

J.T. Lang / Food Policy 41 (2013) 145–154 Table 5 Standardized coefficients (Betas) predicting the effects of overall trust and demographics on explicit trust judgments. University scientists (df = 278) Overall trusta Sex (M = 1; F = 0) Level of education Age group Household income Attend church Adjusted R2

.556*** .014 .004 .015 .158** .041 .300

Farmers (df = 272) .535*** .111* .087 .024 .058 .052 .293

Environmental organizations (df = 249) .678*** .010 .136** .154*** .005 .027 .482

Government agencies (df = 282) .520*** .050 .026 .075 .087 .048 .278

Grocers and stores (df = 269) .507*** .044 .024 .050 .034 .101 .270

Food manufacturers (df = 276) .477*** .009 .070 .058 .031 .058 .242

Note: a Computed Bartlett factor score that combines five trust elements. * Significance level: p 6 0.05. ** Significance level: p 6 0.01. *** Significance level: p 6 0.001.

Table 6 Standardized coefficients (Betas) predicting the effects of five trust elements on explicit trust judgments. University scientists (df = 278) Honesty Knowledge Prediction Shared values Importance Sex (M = 1; F = 0) Level of education Age group Household income Attend church Adjusted R2

.415*** .177* .160 .049 .218* .029 .022 .029 .129* .049 .397

Farmers (df = 272) .323*** .076 .077 .108 .072 .079 .113 .016 .062 .045 .323

Environmental organizations (df = 249) .528** .177* .035 .079 .016 .002 .118* .150** .017 .030 .532

Government agencies (df = 282) .557*** .035 .037 .044 .115 .012 .041 .147** .062 .019 .390

Grocers and stores (df = 269) .264*** .087 .238** .217* .159 .036 .082 .054 .002 .083 .325

Food manufacturers (df = 276) .389*** .007 .094 .123 .006 .003 .034 .074 .043 .029 .300

Note: Significance level: p 6 0.05. Significance level: p 6 0.01. *** Significance level: p 6 0.001. *

**

Level of education is directly related to trusting environmental organizations; age group is inversely related. Two demographic variables, sex and religiosity, do not significantly help explain trust for any group. Overall, the relatively simply model of trust is effective. However, the differences in which elements are most important are fascinating. There are two key differences when compared to the earlier regression analysis presented in Table 5. First, there was an inverse and significant relationship between age and trust in government agencies in this analysis but demographic variables were not significant predictors of trust in the earlier model. Second, though sex significantly helped explain trust in farmers in the earlier model, sex is not significant for any group in this model. Discussion and conclusions I pursued two empirical objectives in this article. First, I explored which actors the public trusts regarding GMF. Levels of trust varied considerably across sources. Respondents considered a mix of activist and industry groups, including university scientists, farmers, and environmental organizations, relatively trustworthy. Respondents considered government agencies, grocers and grocery stores, and food manufacturers less trustworthy. Looking at the specific elements, the results suggest that the public entrusts experts and organizations within a narrow range. In this context, the public generally trusts university scientists and generally distrusts food manufacturers. Government agencies are (dis)trusted more narrowly. Perhaps the variation lies in social role expectations and public beliefs about the scope of organizational authority. These general measures of trust, however, do not

capture the ways that public judgments of an actor vacillate from trustworthy to untrustworthy depending on the element measured. My analysis suggests instead that trust itself is an analytical artifact that depends as much on the institutional actor referent as the trust measure. This lends credence to the idea that the measurement of trust may be organizationally variable and dependent. Recognizing that the elements of trust for one group is different than the elements of trust for another group is reminiscent of Wynne’s (1992, p. 299) assertion that trust represents ‘‘underlying tacit processes of social identity negotiation, involving senses of involuntary dependency on some groups, and provisional or conditional identification with others in an endemically fluid and incomplete historical process’’. Thus, individuals’ beliefs about various experts and organizations are rich insights for scholars to explore, beyond the reductionism of typical trust scholarship. Because organizational actors are not interchangeable institutional actors in debates about trust, they should be essential subjects of critical evaluation. As a general commentary, it is important to note the different capacities that actors have to harness public relations tools to promote positive portrayals and counter (or distort) negative news. In this sense, not all actors are equal. Corporate organizations typically have extensive resources in this respect, while, for instance, university scientists are not a cohesive grouping and not generally able to mobilize around salient issues the same way. This asymmetry is significant, especially in context of discussions about trust. My second objective was to establish if, and how, each element of trust varies by actor. I find limited support for the idea that people who trust industry groups do so in ways that are distinct from

152

J.T. Lang / Food Policy 41 (2013) 145–154

those who trust activist groups. I hypothesized that, because the industry lifeworld privileges scientific knowledge and novelty, assessments of knowledge would be most relevant to trusting industry actors. The results, however, indicate that knowledge is important element of trust for activists, but not so for industry. I also hypothesized that assessments of shared values would be most relevant to activist actors, because the activist lifeworld privileges a normative sensibility that values the environment. The results, however, indicate that shared values are sometimes an important element for industry groups, but not so for activists. Befitting their roles, it remains unclear whether the public views government agencies and university scientists similarly to activists or industry actors. Despite the rejection of my specific hypotheses, the results strongly support the general premise that people trust specific organizations in distinct ways. From a statistical point of view, I was able to substantiate the use of a single factor to describe the five trust elements. There are tradeoffs between the parsimony of the model and the explanatory power. The most important tradeoff, however, seems to be that the striking variation in the importance of trust elements across actors is lost when researchers combine trust elements into a singular, composite measure. The exploration of trust related elements suggests a novel and useful set of relationships for us to explore. For example, I find that perceived honesty is important for predicting trust across groups. Yet the other elements of trust as not universally valuable for predicting trust. The other elements are each only significant predictors of trust for one or two groups. The strength, significance, and variance explained by each element differed depending on the actor. Elements do not uniformly predict trust judgments across actors. Instead, elements vary in the amount they contribute to trust judgments depending on the actor. Wynne (1992, 1996) reminds us that trust is mutable and discursively contested. Thus, trust emerges as the public’s mixed interpretations of various elements of trust as they related to specific organizations and experts. Recognition of this multifaceted character of trust allows a more accurate perspective on public responses to emerging technologies like GMF. Rather than making inappropriate generalizations from these results, my main point is to suggest that overall trust can be a partial function of several elements and there may be differences across actors in the ways that the public applies these elements. In many ways, the analysis presented here is consistent with Levi and Stoker’s (2000) call for scholars grapple with alternative approaches to conceptualizing political trust and trustworthiness. However, this survey instrument does not allow me to assess additional variables and actors that deserve further scrutiny. As such, I have not exhausted the explanatory potential of trust elements in this exploratory analysis. While I controlled for several sociodemographic variables, measures of political orientation may also be salient (Nisbet and Lewenstein, 2002). Interpersonal networks of friends and family might also be an important source of trust. Furthermore, although I present results about consumer trust toward GMF, the relatively simplistic models used are designed to emphasize trust elements. In this, they are consistent with other recent work concerning trust in food systems (e.g., Chen, 2013; Sapp et al., 2009). More detailed statistical models that include variables about awareness, knowledge, and attitudes toward GMF, would be welcome but were not possible with the existing data. Given the sampling and methodology, the results should be interpreted narrowly; the results do not provide insight into consumer attitudes, knowledge, purchasing intentions, or regulatory awareness regarding GMF. However, these limitations do not affect the important contributions to a conceptual understanding of consumer trust in the U.S. food system, which is the primary focus of the research.

To demonstrate that individuals base their trust judgments on varying sets of organizations and experts, scholars must continue to innovate and devise new ways to measure trust reliably across groups of experts and organizations. Future analysis of trust elements using nationally representative data and based on an analysis of the semantic meanings and importance that respondents attribute to measures of trust would provide a more complete explanation of why people trust some actors and do not trust others. More comparative research, across more risk issues and other publics, would help to further evaluate the generality of the findings to theories of hazard-related trust. As Johnson and White (2010) indicated in their work, whether variations across elements turn out to be major or minor factors in future trust research, this proposed research on multiple trust elements should improve our understanding of trust itself. While the respondents are able to assign to each of the actors distinct ratings on each element of trust, and a hierarchy is evident, exactly how to interpret or make use of these results is not immediately obvious. Further research is needed to explore these differences systematically. The more specific and differentiated the subject actors, the more likely we are to come to a nuanced understanding of the social relationships (Poortinga and Pidgeon, 2003). I am hopeful that these results will lead to a theoretically considered exploration of why people may emphasize one element over another. This, along with further empirical analyses of the degree to which trust might differ, or be limited in variation, across elements and actors, seems a worthwhile endeavor based on the present results. These findings have specific lessons for theories of trust. While scholars excel at illustrating the importance of trust for emerging technologies, the variability in which elements of trust are important suggests that scholars might focus less on trust as a concept and more on how elements of trust are distributed. This requires an empirical turn that pushes the standards of evidence beyond macro-measures of trust. Such a turn will help us look at an alltoo-often homogenized trust with a more refined lens, one that brings the details into focus. By focusing on the inconsistent importance of elements of trust, this promising line of research suggests several new questions. Why do elements of trust vary based on the organization? Do the elements of trust vary based on the situation (technological hazard)? How does the public decide to trust particular sources? By directing attention to the elements of trust, rather than trust as an end in and of itself, scholars can begin to focus on the institutional and organizational contexts and features that allow for trust.

Acknowledgments The U.S. Department of Agriculture Initiative for the Future of Agricultural Food Systems grant #2002-52100-11203 supported research for this paper. An earlier version of this paper was presented at the 2009 Annual Meeting of the American Sociological Association in San Francisco, CA. I thank Lee Clarke, Dmitry Khodyakov, and Lisa Wade for their valuable comments on earlier drafts.

References Alario, M., Freudenburg, W., 2003. The paradoxes of modernity: scientific advances, environmental problems, and risks to the social fabric? Sociological Forum 18 (2), 193–214. Allum, N., 2007. An empirical test of competing theories of hazard-related trust: the case of GM food. Risk Analysis 27 (4), 935–946. Annear, C.M., 2004. GM or death: food and choice in Zambia. Gastronomica 4 (2), 16–23. Barber, B., 1983. The Logic and Limits of Trust. Rutgers University Press, New Brunswick, NJ.

J.T. Lang / Food Policy 41 (2013) 145–154 Bijker, W., Hughes, T.P., Pinch, T., 1989. The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. MIT Press, Cambridge, MA. Chen, W., 2013. The effects of different types of trust on consumer perceptions of food safety: an empirical study of consumers in Beijing Municipality, China. China Agricultural Economic Review 5 (1). Clarke, L., 1999. Mission Improbable: Using Fantasy Documents to Tame Disaster. University of Chicago Press, Chicago. Clarke, L., 2006. Worst Cases: Terror and Catastrophe in the Popular Imagination. University of Chicago Press, Chicago. Clarke, L., Short, J.F., 1993. Social organization and risk: some current controversies. Annual Review of Sociology 19, 375–399. Coleman, J.S., 1990. Foundations of Social Theory. Belknap Press of Harvard University Press, Cambridge, MA. Cook, Karen S., 2001. Trust in Society. Russell Sage Foundation, New York. Costa-Font, M., Gil, J.M., Traill, W.B., 2008. Consumer acceptance, valuation of and attitudes towards genetically modified food: review and implications for food policy. Food Policy 33 (2), 99–111. Cvetkovich, G.T., Siegrist, M., Murray, R., Tragesser, S., 2008. New information and social trust: asymmetry and perseverance of attributions about hazard managers. Risk Analysis 22 (2), 359–367. Delhey, J., Newton, K., Welzel, C., 2011. How general is trust in ‘most people’? Solving the radius of trust problem. American Sociological Review 76 (5), 786– 807. Earle, T.C., Cvetkovich, G.T., 1995. Social Trust: Towards a Cosmopolitan Society. Praeger, Westport, CT. Earle, T.C., Siegrist, M., Gutscher, H., 2010. Trust in Risk Management: Uncertainty and Scepticism in the Public Mind. Routledge, London. Einsiedel, E.F., 2009. Emerging Technologies: From Hindsight to Foresight. UBC Press, Vancouver. Erikson, K., 1994. A New Species of Trouble. W.W. Norton, New York. Falkner, R., 2007. The International Politics of Genetically Modified Food: Diplomacy, Trade and Law. Palgrave Macmillan, Hampshire, England. Freudenburg, W.R., 1992. Heuristics, biases, and the not-so-general publics: expertise and error in the assessment of risks. In: Krimsky, S., Golding, D. (Eds.), Social Theories of Risk. Praeger, Westport, CT, pp. 299–350. Freudenburg, W.R., 1993. Risk and recreancy: Weber, the division of labor, and the rationality of risk perceptions. Social Forces 71 (4), 909–932. Frewer, L.J., Howard, C., Hedderley, D., Shepherd, R., 1996. What determines trust in information about food-related risks? Underlying psychological constructs. Risk Analysis 16 (4), 473–486. Fukuyama, F., 1995. Trust: The Social Virtues and the Creation of Prosperity. Free Press, New York. Gambetta, D., 1988. Trust: Making and Breaking Cooperative Relations. Blackwell Publishers, Oxford. Gaskell, G., Allum, N.C, Stares, S.R., 2003. Europeans and Biotechnology in 2002: Eurobarometer 58.0. European Commission, Brussels. Giddens, A., 1990. The Consequences of Modernity. Stanford University Press, Stanford, CA. Granovetter, M., 1985. Economic action and social structure: the problem of embeddedness. American Journal of Sociology 91 (3), 481–510. Hallman, W.K, Hebden, W.C., Aquino, H.L, Cuite, C.L, Lang, J.T, 2004. Americans and GM Food: Knowledge, Opinion and Interest in 2004. Food Policy Institute, Cook College, Rutgers, The State University of New Jersey, New Brunswick, NJ, . Hardin, R., 2002. Trust and Trustworthiness. Russell Sage Foundation, New York. Hardin, R., Levi, M., Cook, K.S., 2009. Whom Can We Trust? How Groups, Networks, and Institutions Make Trust Possible. Russell Sage Foundation, New York. Heimer, C.A., Cook, K.S., 2001. Solving the problem of trust. In: Cook, K.S. (Ed.), Trust in Society. Russell Sage Foundation, New York, pp. 40–88. Huffman, W.E., Rousu, M., Shogren, J.F., Tegene, A., 2004. Who do consumers trust for information: the case of genetically modified foods? American Journal of Agricultural Economics 86 (5), 1222–1229. Huffman, W.E., Rousu, M., Shogren, J.F., Tegene, A., 2007. The effects of prior beliefs and learning on consumers’ acceptance of genetically modified foods. Journal of Economic Behavior & Organization 63 (1), 193–206. James, C., 2010. Global Status of Commercialized Biotech/GM Crops: 2010. ISAAA Brief 42. Ithaca, NY. . James, H.S., Marks, L.A., 2008. Trust and distrust in biotechnology risk managers: insights from the united kingdom, 1996–2002. AgBioForum 11 (2), 93–105. Jasanoff, S., 2005. Designs on Nature: Science and Democracy in Europe and the United States. Princeton University Press, Princeton, NJ. Johnson, B.B., 1999. Exploring dimensionality in the origins of hazard related trust. Journal of Risk Research 2 (4), 325–354. Johnson, B.B., White, M.P., 2010. The importance of multiple performance criteria for understanding trust in risk managers. Risk Analysis 30 (7), 1099–1115. Krosnick, J.A., 1999. Survey research. Annual Review of Psychology 50, 537–567. Lang, J.T., Hallman, W.K., 2005. Who does the public trust? The case of genetically modified food in the United States. Risk Analysis 25 (5), 1241–1252. Levi, M., Stoker, L., 2000. Political trust and trustworthiness. Annual Review of Political Science 3 (1), 475–507. Lewicki, R.J., McAllister, D.J., Bies, R.J., 1998. Trust and distrust: new relationships and realities. The Academy of Management Review 23 (3), 438–458. Lewis, J.D., Weigert, A., 1985. Trust as a social reality. Social Forces 63 (4), 967–985. Luhmann, N., 1979. Trust and Power. Wiley, Chichester, NY.

153

Meijnders, A., Midden, C., Olofsson, A., Öhman, S., Matthes, J., Bondarenko, O., Gutteling, J., Rusanen, M., 2009. The role of similarity cues in the development of trust in sources of information about GM food. Risk Analysis 29 (8), 1116– 1128. Metlay, D., 1999. Institutional trust and confidence: a journey into a conceptual quagmire. In: Cvetkovich, G.T., Lofstedt, R.E. (Eds.), Social Trust and the Management of Risk. Earthscan, London, pp. 110–116. Midden, C.J.H., Huijts, N.M.A., 2009. The role of trust in the affective evaluation of novel risks: the case of CO2 storage. Risk Analysis 29 (5), 743–751. Misztal, B.A., 1996. Trust in Modern Societies: The Search for the Bases of Social Order. Polity Press, Cambridge, MA. Misztal, B.A., 2011. Trust: acceptance of, precaution against and cause of vulnerability. Comparative Sociology 10 (3), 358–379. Mizrachi, N., Drori, I., Anspach, R.R., 2007. Repertoires of trust: the practice of trust in a multinational organization amid political conflict. American Sociological Review 72 (1), 143–165. Mizruchi, M.S., Stearns, L.B., Marquis, C., 2006. The conditional nature of embeddedness: a study of borrowing by large U.S. firms, 1973–1994. American Sociological Review 71 (2), 310–333. Mollering, G., 2006. Trust: Reason, Routine, Reflexivity. Elsevier Science, London. Nisbet, M.C., Lewenstein, B.V., 2002. Biotechnology and the American media: the policy process and the elite press, 1970 to 1999. Science Communication 23 (4), 359–391. Paxton, P., 2007. Association memberships and generalized trust: a multilevel model across 31 countries. Social Forces 86 (1), 47–76. Perrow, C., 1999. Normal Accidents: Living with High Risk Technologies. Princeton University Press, Princeton, NJ. Peters, R.G., Covello, V.T., McCallum, D.B., 1997. The determinants of trust and credibility in environmental risk communication: an empirical study. Risk Analysis 17 (1), 43–54. Pew Initiative on Food and Biotechnology, 2005. U.S. vs. EU: An Examination of the Trade Issues Surrounding Genetically Modified Food. Pew Charitable Trusts, Washington, DC, . Poortinga, W., Pidgeon, N.F., 2003. Exploring the dimensionality of trust in risk regulation. Risk Analysis 23 (5), 961–972. Poortinga, W., Pidgeon, N.F., 2004. Trust, the asymmetry principle, and the role of prior beliefs. Risk Analysis 24 (6), 1475–1583. Priest, S.H., Bonfadelli, H., Rusanen, M., 2003. The ‘‘trust gap’’ hypothesis: predicting support for biotechnology across national cultures as a function of trust in actors. Risk Analysis 23 (4), 751–766. Putnam, R.D., 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton University Press, Princeton, NJ. Renn, O., Levine, D., 1991. Credibility and trust communication. In: Kasperson, R.E., Stallen, P.J.M. (Eds.), Communicating Risks to the Public. Kluwer Academic, Dordrecht, pp. 287–324. Rousu, M., Huffman, W.E., Shogren, J.F., Tegene, A., 2007. Effects and value of verifiable information in a controversial market: evidence from lab auctions of genetically modified food. Economic Inquiry 45 (3), 409–432. Sapp, S.G., Arnot, C., Fallon, J., Fleck, T., Soorholtz, D., Sutton-Vermeulen, M., Wilson, J.J.H., 2009. Consumer trust in the U.S. food system: an examination of the recreancy theorem. Rural Sociology 74 (4), 525–545. Schurman, R., 2004. Fighting ‘frankenfoods’: industry structures and the efficacy of the anti-biotech movement in western Europe. Social Problems 51 (2), 243– 268. Schurman, R., Munro, W., 2009. Targeting capital: a cultural economy approach to understanding the efficacy of two anti-genetic engineering movements. American Journal of Sociology 115 (1), 155–202. Schurman, R., Munro, W., 2010. Fighting for the Future of Food: Activists Versus Agribusiness in the Struggle over Biotechnology. University of Minnesota Press, Minneapolis, MN. Seligman, A.B., 2000. The Problem of Trust. Princeton University Press, Princeton, NJ. Siegrist, M., Cvetkovich, G., Roth, C., 2000. Salient value similarity, social trust, and risk/benefit perception. Risk Analysis 20 (3), 353–362. Sjoberg, L., 1999. Perceived competence and motivation in industry and government as factors in risk perception. In: Cvetkovich, G., Lofstedt, R.E. (Eds.), Social Trust and the Management of Risk. Earthscan Publications, London, pp. 89–99. Slovic, P., 1993. Perceived risk, trust, and democracy. Risk Analysis 13 (6), 675–682. Sztompka, P., 1999. Trust: A Sociological Theory. Cambridge University Press, Cambridge, UK. Tilly, C., 2005. Trust and Rule. Cambridge University Press, New York. United States Department of Agriculture, 2011. The NASS June Agricultural Survey. National Agricultural Statistics Service, Washington, DC. Verbrugge, L.M., 1979. Multiplexity in adult friendships. Social Forces 57 (4), 1286– 1309. Vileisis, A., 2007. Kitchen Literacy: How We Lost Knowledge of Where Food Comes from and Why We Need to Get it Back. Island Press, Washington, DC. Weick, K.E., Sutcliffe, K.M., 2001. Managing the Unexpected: Assuring High Performance in an Age of Complexity. University of Michigan Business School Management Series. Jossey-Bass, San Francisco. White, M.P., Eiser, J.R., 2005. Information specificity and hazard risk potential as moderators of trust asymmetry. Risk Analysis 25 (5), 1187–1198. White, M.P., Eiser, J.R., 2006. Marginal trust in risk managers: building and losing trust following decisions under uncertainty. Risk Analysis 26 (5), 1187– 1203.

154

J.T. Lang / Food Policy 41 (2013) 145–154

Wynne, B., 1980. Technology, risk and participation: on the social treatment of uncertainty. In: Conrad, J. (Ed.), Society, Technology and Risk Assessment. Academic Press, New York, pp. 167–202. Wynne, B., 1992. Misunderstood misunderstanding: social identities and public uptake of science. Public Understanding of Science 1 (3), 281–304.

Wynne, B., 1996. May the sheep safely graze? A reflexive view of the expert-lay knowledge divide. In: Lash, S., Takahashi, N., Wynne, B. (Eds.), Risk, Environment and Modernity: Towards a New Ecology. Sage, London, pp. 44–83. Yamagishi, T., 2011. Trust: The Evolutionary Game of Mind and Society. Springer, Tokyo.