Vulnerability and Social Justice as Factors in Emergent U.S. Nanotechnology Risk Perceptions
Abstract
As an emerging domain of risk research, nanotechnologies engender novel research questions, including how new technologies are encountered given different framing and contextual detail. Using data from a recent U.S. national survey of perceived risks (N= 1,100), risk versus benefit framings and the specific social positions from which people encounter or perceive new technologies are explored. Results indicate that vulnerability and attitudes toward environmental justice significantly influenced risk perceptions of nanotechnology as a broad class, while controlling for demographic and affective factors. Comparative analyses of different examples of nanotechnology applications demonstrated heightened ambivalence across acceptability when risk versus benefit information was provided with application descriptions (described in short vignettes as compared to the general category “nanotechnology,” absent of risk or benefit information). The acceptability of these nano-specific vignettes varied significantly in only some cases given indexes of vulnerability and attitudes toward environmental justice. However, experimental narrative analyses, using longer, more comprehensive descriptive passages, show how assessments of risks and benefits are tied to the systematically manipulated psychometric qualities of the application (its invasiveness and controllability), risk messaging from scientists, and the social implications of the technology with regard to justice. The article concludes with discussion of these findings for risk perception research and public policy related to nanotechnology and possibly other emerging technologies.
1. INTRODUCTION
Nanotechnologies are manufactured materials (e.g., quantum dots, carbon nanotubes) and enabling platforms built at the nanoscale. They promise many novel applications such as molecular-level targeted drug delivery systems, improved sensors and environmental cleanup technologies, and lighter more efficient materials. However, their small size and unique properties such as surface reactivity mean that they may carry substantial health risks.(1-6) Given the newness of nanotechnologies, their intriguing novel capacities, and yet some potential for harm, social scientists have begun investigating how different publics might perceive these technologies, absent widespread familiarity and well-established toxicological understanding. The question generally posed by those interested in perceived risk is: How might current growth in nanotechnology research and development be viewed by different publics and will the products and capacities derived from nanomaterials be met with optimism or aversion? In particular, will nanotechnologies be the subject of controversy, as has been the case with some other new technologies, such as genetically modified organisms or cloning research? Or, will benefit appreciation prevail over risk aversion and, for either, why?
This article explores these possibilities by examining how information about these new technologies is encountered and perceived given variations in the attributes of information, and given the attitudes and social experience of perceivers (whether they see the world as just, or not; and whether they experience themselves as vulnerable in the world, or not). Using data from a recent national risk perception survey of American adults, we demonstrate that when the distribution of risks and benefits from nanotechnologies is perceived as unfair, concomitant concerns for social justice lead to heightened perception of the nanotechnologies as risky. Similarly, we show that experiences of vulnerability amplify perceptions of risk associated with nanotechnologies. Further, we explore variation in risk perception between application domains (energy, food, and medicine).
Because many remain unfamiliar with this new and potentially ubiquitous class of technologies, exploring such questions requires the use of tutorials and more detailed information frames. The acceptability of risks and benefits is examined across one question and two experiments: first, we examine attitudes toward nanotechnology generally and in the absence of any experimental manipulations; second, we examine the same question following short descriptive vignettes; third, we use longer narrative passages (the designs of each are described below). Together these experiments indicate the close link of acceptability and risk-benefit judgments to the social implications of the technology; to the equity of their availability to less affluent citizens; as well as the technologies’ perceived bodily invasiveness and (to a lesser degree) scientifically reported uncertainty as to the behavioral properties of the materials. Recognizing the embeddedness of technological innovation in social contexts, including experiences of vulnerability and normative evaluations related to social justice, extends the conventional foci of risk perception research and demonstrates how justice deserves greater attention as a factor in perceptions of risk.
To date, survey research on public perceptions of nanotechnology has confirmed relatively low levels of familiarity with the technology. A meta-analysis of 22 surveys conducted in the United States, Canada, Europe, and Japan between 2002 and 2009 found that on average over 51% of survey respondents reported knowing “nothing at all” about nanotechnology. Approximately 20% reported knowing “some or a lot.”(7) Overall, American and European publics are optimistic about nanotechnology, with approximately three times as many people reporting that the benefits will outweigh the risks than the reverse. At the same time, respondents are also hesitant to judge the riskiness or acceptability of nanotechnologies. Approximately 44% of pooled respondents across nations indicated that they were “not sure” about the relative risks and benefits, demonstrating a striking “judgment conservatism” and possibly indicating malleability of thought as yet with regard to perceptions of risks and benefits of nanotechnologies.(7,8) These findings demonstrate the unique opportunities(9) offered risk researchers to better understand how new technologies are encountered and made sense of given different information and positions of perceivers.
Most attempts to identify the factors driving risk perceptions of nanotechnology have attended to the demographic characteristics of perceivers, levels of familiarity with these technologies, and trust in their regulation.(7) Female gender and nonwhite race are predictive of increased perceived risk.(10) Investigation of attitudinal dispositions has focused largely on attitudes about trust in science and identification with political parties. Research on religiosity and the influence of sociopolitical worldviews or environmental values and how they affect risk perceptions has been limited, but early findings suggest this as an important target for research on emerging perceptions of nanotechnologies.(11)
For instance, Vandermoere et al.,(12) in their French study, called for greater attention to “political” research on public risk perceptions of nanotechnology, and in particular, greater attention to the “(re)organisation [sic] of the relations between science, technology, and nature” (p. 10). This includes a focus on the influence of religious and moral values on the desirability of nanotechnology. While Schuefele and colleagues(11) found religiosity to be an important predictor of nanotechnology risk perceptions in the United States and Europe, Vandermoere et al.(13) did not in the case of Germany, though they did find it to be a predictor of having an opinion on nanotechnology. Brossard et al.(14) showed that, in the United States, religiosity mediates knowledge of nano and that knowledge is more strongly associated with acceptance of nano for the nonreligious.
Gaskell et al.(15) identified a techno-optimism, particularly in the United States, that is associated with perception of benefits of nanotechnology, while controlling for measures of environmental values. Kahan et al.(10,16) demonstrated how shared sociopolitical worldviews (aka cultural cognitions) in general and in reference to key American political topics such as climate change and same-sex marriage help shape the social meanings associated with new technologies or other potentially risky activities, and underlie the cognitive processes by which individuals encounter and make sense of new information.(16) Thus we might expect, and others have found, that cultural cognition is a good indicator of just how risk perceptions related to nanotechnology might unfold. Specifically, while most people rely on cultural cognition to assess new information, those with uninformed views about nanotechnologies tend to more fully “use” their cultural and political worldviews as heuristics for making sense of and evaluating this new class of technologies and any risks they may engender. We build on these findings by asking whether attitudes about justice similarly operate as value-driven cognitions informing risks perceptions of nanotechnologies.
While research to date has revealed that novel moral, demographic, and political worldviews, as well as affect,(10,17) underpin high versus low perceived risk, it has also tended to treat “nanotechnology” as a discrete, singular risk object or class of risk. Yet, scholars examining specific nano-applications have found a greater diversity of risk perceptions than research on “nanotechnology” alone anticipated.(12,18,19) In particular, public risk perceptions of nanotechnology enhancements of food and food packaging have been found to be less than optimistic. For instance, studies on risk perceptions of nano-related applications in food and food packaging in Switzerland found that people expressed reluctance to purchase food products that incorporated nanotechnology, but more willingness to buy food packaging that did.(20,21) Similarly, Vandermoere and colleagues(12) demonstrated that most people in their French study were either skeptical about the risks and benefits of nanofood applications or pessimistic, viewing the risks of nanotechnology applications in food and food packaging as greater than the benefits. Pidgeon et al.(22) also identified differences in U.S. and U.K. publics in the perceived risks and benefits of nanotechnologies for specific application domains. Energy applications were viewed more positively than applications in health or human enhancement, but across all of them, concerns about the equity of distribution of benefits and risks across society were more prominent than concerns for technological features of the risk objects.
While affirming the importance of these early findings, a more precise rendering of how nanotechnologies will be confronted by those occupying different social positions and in response to diverse descriptors remains comparatively absent. Building on earlier findings, we focus on nanotechnology applications in the domains of food, health, and energy; we expect the U.S. public to have different responses to them. In particular, we expect the public to be positive toward energy applications but skeptical or even opposed to applications in the domain of food. Moreover, we situate these applications in greater contextual detail that more closely approximates how people may encounter information about nanotechnologies, and whether or not they regard themselves as well positioned to realize their benefits.
Many social studies of risk recognize that the distribution of new technologies and any hazards they may engender necessitates questions of social or environmental justice. For instance, Burningham and Thrush(23) argue that concerns about the quality of the environment are caught up in concerns about local life and the effectiveness of public management of the environment. By extension, the meanings associated with technologies can be understood to be perceived not just as a function of their attributes (the technology is safe or not, invasive or not, charismatic or not, new, or bright in color, and so on), but also as a function of how they are nested in social contexts, including attributions of who in particular will face any risks introduced by the new technology and who will garner its benefits. In tracking risk perceptions about unfamiliar risk objects, individuals might be expected to weigh the implications of new technologies for fairness and justice,(24) an estimation that might extend to their judgments of how the risks and benefits of that object may be distributed across and between social groups as well as concerns for procedural fairness that governs that distribution.(25,26)
Justice and experiences of marginality have been closely examined by scholars of environmental injustice and racism, who have found that poor and minority communities tend to bear a disproportionate share of environmental hazards.(27-33) While the environmental justice hypothesis—that minority communities are more likely to be burdened with environmental hazards and other negative externalities of industrialization—has been underresearched in the domain of risk perception survey work, in one study experiences of vulnerability and attitudes toward environmental justice are predictive of perceived risk.(34) These factors were shown to be partially explanatory of persistent demographic differences in risk perception, wherein a strong “[elite] white male effect,” or risk tolerance among this demographic group, is in part seen to be produced by variables other than race and gender.(35) Thus, awareness of and experience with how new technologies are taken up by social groups,(36) including those historically marginalized, affect how those technologies are understood. In this regard, nanotechnologies inspire questions about social life, including its inequalities. By this logic, perceptions of nanotechnologies are likely to be entangled in assessments of fairness and justice, and thus we expect environmental justice and vulnerability to be predictive of risk perceptions. More specifically, we expect those affirming vulnerability to be more sensitive to risk information and thus more risk adverse in response to shorter (vignette) passages. Similarly, when the social conditions or the availability or usability of new technologies is/are made more salient, we expect concerns for justice to be associated with heightened risk aversion.
In the remainder of the article, we examine the following hypotheses:
- 1
Affirmation of environmental justice and experiences of vulnerability will stand up as logically consistent and useable constructs, associated with heightened perceptions of nanotechnology risks while controlling for known demographic predictors of risk; but will not override affect as predictors of risk judgments (given the predictive power of affect in earlier studies).
- 2
Judgments about environmental justice and experiences of vulnerability will be associated with amplified risk perceptions in the context of risk information but have only negligible effects when nanotechnology applications are framed as beneficial or safe.
- 3
Justice and vulnerability will prevail against known psychometric variables (controllability, dread-like qualities, and whether or not technologies are understood by scientists), under conditions of detailed and systematically manipulated information.
In investigating these hypotheses, we first assess the role of justice and the related concept of vulnerability through one test and two experiments utilizing risk perception survey data. We describe the construction of scales for environmental justice and vulnerability and then test them against two well-known risk perception factors: affect (the positive/negative valences assigned to objects and key to rapid evaluation of information) and gender.(37) We then assess the sensitivity of vulnerability and environmental justice to risk versus safety information provided in brief vignettes across energy, environmental, and medical applications. Finally, we evaluate social justice against three well-established predictors of risk perceptions in the context of extended narrative passages containing four manipulated variables. The article concludes with a discussion of the implications of these findings for risk research and the future of nanotechnologies.
2. DATA AND DESIGN
The data reported here were collected through a U.S. national telephone survey (N= 1,100) designed to assess public perceptions of nanotechnology. The survey was administered by telephone in the fall of 2008. Significant effort was made to contact participants. Numbers were dialed a minimum of seven times until either an interview was conducted or the potential respondent refused to participate. Every reasonable effort was made to encourage participation. Of those numbers that were deemed to be eligible to participate in the survey (residential numbers only) the response rate was 51.9% and the refusal rate was 38.8%, based on actual contacts. The mean age was 57, with 57% women and 43% men. Eighty-two percent of participants were white, 7% black, and 3% Hispanic.4
As described, while most surveys about nanotechnology to date have found that different U.S. publics view the benefits of nanotechnology as outweighing the risks, the evidence also suggests that these views have not stabilized, and information provided is an important variable.(7,8,38-40) We tested this assertion in our research design via three relevant question sets, which included (1) self-assessments of experiences of vulnerability and attitudes about environmental justice tested against other measures of risk perception; (2) short, paired vignettes designed to pit risk and safety information against one another to better understand the effect of each; and (3) the above noted narrative passages, each of which systematically manipulated four dimensions of specific nano-enabled applications, including information on risk and safety from experts, the controllability of material once released, the bodily invasiveness of material, and the equity or not of access to these new technologies by the rich or poor.
3. ANALYSIS AND RESULTS
3.1. Vulnerability and Environmental Justice Scales
The survey items used to elicit attitudes toward vulnerability and environmental justice were drawn from Satterfield et al.,(34) who demonstrated that they are reliable proxies for these concepts. The items used in creating the environmental justice scale were as follows:
- 1
I think hazardous facilities are more common in minority communities.
- 2
Minority communities lack the political clout to stop hazardous facilities from being located near them.
- 3
The government should restrict the placing of hazardous facilities in minority communities.
Respondents were asked to report on a four-point scale how strongly they agreed or disagreed with those statements. The reliability is acceptable (Cronbach alpha = 0.66). Two-thirds of survey participants either agreed or strongly agreed with the environmental justice hypothesis. Of those who agreed, 75% were age 45 or older, nearly 58% had achieved a postsecondary degree, and 65% earned $75K or less per year. By a ratio of almost 2 to 1, more women agreed with the environmental justice claims than men (X2 (df) = 27.49 (1), p= 0.000).5
Theoretically, environmental justice is primarily an assessment of the distribution of environmental hazards within society.6 It also suggests a greater perceived vulnerability to hazards. We thus used three items developed and tested in earlier studies(34) to elicit experiences of vulnerability though we did not ask, specifically, about vulnerability to environmental injustices:7
- 1
I often feel discriminated against.
- 2
I have very little control over risks to my health.
- 3
How would you rate the quality of medical care that is available to you and your family?
As with attitudes about environmental justice, respondents were asked to report on a four-point scale how strongly they agreed or disagreed with these statements. For item three only, the response categories ranged from poor to excellent.8 The reliability of this scale is low (Cronbach alpha = 0.56), but acceptable and based on familiar constructs tested in other contexts. It is a likely contributor to lower power in tests reported in the next section. Vulnerability combines aspects of health inequality, particularly a lack of control over one's health and inadequate medical care, with experiences of discrimination.9
Those that reported vulnerability were disproportionately of lower income (70%, <$50K) and without a college degree (63%). While most in our sample who were reporting vulnerability were white (74%), nearly 25% of nonwhite men and women reported vulnerability compared to between 13 and 15% of white men and women.
Given the newness of nanotechnology and its expected importance in the future, it is worth considering the influence of vulnerability and environmental justice in the context of other well-established predictors of risk perceptions, including affect and the white male effect.(35,41) Affect is the positive or negative valence assigned to objects and key to rapid evaluation of their risks or benefits. Gender is a known predictor of risk perception, as expressed by the white male effect: the finding that politically conservative, authoritarian, high-income earning, and highly educated white males are very risk tolerant, relative to all others. The vulnerability and environmental justice scales are associated with distinctive demographic characteristics, correlating with race and gender, that suggest a relation with the white male effect. For instance, reports of experiences of vulnerability are negatively associated with white respondents (rho =−0.01, p= 0.000; nonwhite males: rho = 0.09, p= 0.005; nonwhite females: rho = 0.15, p= 0.000). Reports of environmental justice are negatively associated with male (rho =−0.19, p= 0.000), particularly white males (rho =−0.17, p= 0.000; nonwhite males, p= n.s.). Among females, white females are positively associated with environmental justice (rho = 0.135, p= 0.000) as are nonwhite females (rho = 0.07, p= 0.03). While there is significant overlap between experiences of vulnerability and environmental justice with women and people of color, the relationship is not perfect. Most white women (N= 399; 85%), nonwhite women (N= 70; 75%), and nonwhite men (N= 53; 75%) report experiences of vulnerability. Where a small minority of white males report experiences of vulnerability (N= 45; 13%), a slight majority affirm environmental justice (N= 188; 55%) as do majorities of white women (N= 324; 70%), nonwhite women (N= 71; 76%), and nonwhite men (N= 37; 54%). As with earlier findings,(34) a number of white men report experiences of vulnerability and endorse the environmental justice items. The question is whether vulnerability and environmental justice can be reduced to their demographic characteristics. Or, in other words, are vulnerability and environmental justice analogs for the white male effect or even affect?
The ordered logistic regression results presented in Table I confirm how experiences with vulnerability and attitudes about environmental justice inform risk perceptions, while controlling for standard measures such as race, gender, and affect.10 In this test, an assessment of the risks and benefits of nanotechnology served as the dependent variable. The independent variables included binary indicators for race (white vs. nonwhite) and gender (male vs. female), an affective rating of nanotechnology, and the two scales for vulnerability and environmental justice.
| I | II | III | |
|---|---|---|---|
| White | 0.59*** | 0.38* | 0.43 |
| (0.16) | (0.17) | (0.18) | |
| Male | 0.86*** | 0.81*** | 0.55*** |
| (0.13) | (0.14) | (0.15) | |
| Vulnerability | −0.21*** | −0.14*** | |
| (0.04) | (0.04) | ||
| Environmental Justice | −0.09* | −0.08* | |
| Affect | 1.24*** | ||
| (1.00) | |||
| Pseudo R2 (Nagelkerke) | 0.07 | 0.13 | 0.33 |
| Model X 2 | 000 | 000 | 000 |
| −2LL | 43.04 | 846.40 | 1044.72 |
- Notes: Standard errors are in parentheses. Vulnerability and environmental justice are reverse coded; a high number indicates strong agreement.
- Significant at *p < 0.05; **p < 0.01; ***p < 0.001.
The goal of the modeling exercise is to assess the extent to which these items help explain risk perceptions of nanotechnology vis-à-vis these other well-known predictors of risk perceptions. We elicited overall risk perceptions using the following scale: 1 (the risks of nanotechnology will outweigh its benefits); 2 (… be about the same); 3 (the benefits … will outweigh its risks), whereas we elicited affective responses to nanotechnology by asking “how would you say nanotechnology made you feel?” where the scale consisted of 1 (very bad); 2 (bad); 3 (neither good nor bad); 4 (good); 5 (very good). Responses on both scales were elicited immediately after a tutorial that read:
Nanotechnology involves human made materials or devices that are extremely small, for instance 10,000 times smaller than a fine grain of sand. Because they’re so small, they can have unique physical, chemical, electrical or other properties. This may produce new health or environmental risks because these nanomaterials can get into the human body through breathing, skin contact, or eating or they may end up in the environment with unexpected or unknown long-term effects. They’re being developed because they may provide new ways to treat disease, clean up pollution, improve food, and provide cheaper energy.
Consistent with the white male effect, white and male participants perceived the benefits of nanotechnology as outweighing the risks as compared to women and nonwhites. Controlling for the effects of gender and race, experiences of vulnerability and attitudes toward environmental justice are negatively associated with benefits outweighing the risks. Also consistent with Satterfield et al.,(34) this demonstrates how vulnerability and environmental justice do not fully account for the effects of race and gender and are also distinct from generalized affect, which is significant and predictive of the perceptions of benefits outweighing the risks. Vulnerability and environmental justice constitute distinct factors affecting perceptions of the risks and benefits of nanotechnologies, but, as expected, they do not eliminate the power of affect in explaining initial reactions to nanotechnology.
3.2. Influence of Vulnerability and Environmental Justice on Acceptability of Nanotechnology Applications
While most nanotechnology risk perception surveys report strong benefit ratings, it is not entirely clear the extent to which this is a function of early-stage technological optimism, or a framing effect derived from the fact that many surveys have either not provided information tutorials or described nanotechnologies in reference to their uses, which are all beneficial to the extent that nano-applications are typically introduced and justified by their uses, for example, an energy-saving device.(22) To examine this possibility, we designed an experiment to manipulate the delivery of risk and safety information about different nanotechnology applications. We also sought to examine the influence of vulnerability and environmental justice as a factor that might influence respondent sensitivity to this information. We expected that risk statements would generate ratings of unacceptability and safety statements would produce acceptability, but those with high rating on vulnerability and environmental justice would be more sensitive to risk information.
Participants (four subsamples of approximately N= 275) were asked to provide ratings of acceptability for a range of specific nano-applications, each described in short vignette form. All six vignettes began with a statement of the purpose of the technology and its benefits; three of these were accompanied by clear safety messages, and three were followed by counterpart risk messages.11 Risk and safety messages were drawn from concurrent deliberation work.(22) For instance, safety messages included social safety in the form of promised benefits to the poor with informed consent; and scientific safety in the form of certainty of the materials’ controllability; whereas risk messages included the possibility of compromised privacy and examples of environmental harms. Respondents rated the acceptability of each vignette on a four-point scale where 1 equals “very acceptable,” 2 equals “acceptable,” 3 equals “unacceptable,” and 4 equals “very unacceptable.”
Table II presents the results of an ordinal regression assessing the effects of environmental justice and vulnerability on the acceptability ratings of the short vignettes. All significant effects are negative, indicating that agreement with vulnerability and environmental justice items is associated with unacceptable ratings of the application vignettes. For the application vignettes with risk content, some sensitivity to risk messaging is evident, across those who agree with the environmental justice and vulnerability statements, to the extent that one item (small radio transmitters for medical diagnoses) generated significant negative effects in the case of vulnerability, a second item (nanosilver bandages) in the case of environmental justice, and the third item (pollution sensors) in reference to both environmental justice and vulnerability.
| Risk Corrective | Safety Corrective | |||||
|---|---|---|---|---|---|---|
| Nanotechnology sensors can be used to detect pollutants in soil and water. However, experts worry that these tiny sensors will degrade overtime becoming toxic to fish or humans who use or drink the water. | Tiny nanosilver particles can be used in bandages to stop infections by killing bacteria. But nanosilver is considered a water pollutant; if it turns up in our rivers or oceans, fish, and other marine life may be widely harmed. | Nanotechnologies are used to create small radio transmitters that can send medical diagnoses and records from remote areas by Internet or cell phone. As the information is sent, however, leaks in privacy may be unstoppable or irreversible. | Tiny instruments known as nanorobots can be used to repair damaged cells and organs in water birds damaged by oil spills. This new technology and how to control it is now well understood by scientists. | Tiny nanoparticles are being developed for targeted delivery of medicine at the cellular level, a process that means that chemotherapy will only kill diseased cells. This new technology will be available for the poor or those with limited or no healthcare. | Nanotechnology can be used to make a very small device for easy use and medical diagnoses in remote areas of the world. They are being tested in poor countries at the request of people there in areas where few medical services exist. | |
| Vulnerability | −0.21** | −0.02 | −0.24*** | −0.19** | −0.08 | −0.23** |
| (0.07) | (0.06) | (0.06) | (0.07) | (0.06) | (0.07) | |
| Environmental | −0.21** | −0.24*** | −0.003 | 0.03 | 0.03 | −0.10 |
| Justice | (0.07) | (0.06) | (0.06) | (0.07) | (0.06) | (0.07) |
| Pseudo R2 | 0.09 | 0.07 | 0.08 | 0.03 | 0.07 | 0.06 |
| (Nagelkerke) | ||||||
| −2LL | 223.49 | 294.71 | 322.80 | 261.01 | 305.35 | 269.80 |
| Model X2 Sig. | 0.000 | 0.000 | 0.000 | 0.03 | 0.48 | 0.02 |
- *p < 0.05; **p≤ 0.01; ***p≤ 0.001.
- Notes: Vulnerability and Environmental Justice are in their original form; a high number indicates strong disagreement.
- Standard errors are in parentheses.
In the other three vignettes, which deliver information about the safe use of nanotechnologies, the environmental justice scale was nonsignificant. Vulnerability is significant for two of three safety-oriented vignettes, indicating that those reporting vulnerability also rate nanotechnology applications as unacceptable even when described as safe, such as in the case of nanorobots for environmental remediation and remote medical diagnostics for the poor.
These findings are largely consistent with our second hypothesis, namely: agreement with environmental justice positions and experiences of vulnerability are associated with heightened sensitivity to risk information. The negative coefficients suggest that those who report vulnerability and affirm environmental justice display sensitivity toward risk information and as a result tend to rate nanotechnology applications described as risky as more unacceptable. Not anticipated in our hypothesis, however, were the negative coefficients for vulnerability in two of three safety-oriented vignettes. These results indicate that experiences of vulnerability can generate negative judgments of nanotechnology even when framed entirely as beneficial and safe. Given the low reliability, and thus low power, of the scale for vulnerability, these results are likely understated. This may also be affected by wording variability in the examples and messaging itself, as we sought realistic and scientifically valid examples across domains at the expense of manipulations that were perfect mirrors of one another.
3.3. Narrative Scenarios
To further assess the influence of justice on the perception of nanotechnologies, we developed more elaborated scenarios of nano-applications wherein the effect of particular combinations of variables could be inferred. In this case we used longer narrative passages because we sought an alternative to the preponderance of short descriptions of nano-enabled products in the literature thus far, which omit the “realism” of more complex descriptions of nanotechnologies that people are likely to encounter in real-life contexts. Longer narratives were also selected as narrative communication has been shown to enhance the ability of people to comprehend complex information and consider multiple dimensions of a problem.(42-45) Specifically, we embedded in narrative passages four binary variables expressed as statements within each narrative. These variables included scientists’ risk judgments (R), controllability (C) of the technology, the potential for bodily invasion (I), and whether the nanotechnology application was produced in a socially just (J) manner. The first three variables were selected because they are proximate to known psychometric variables demonstrated to influence risk judgments,(46-48) and the fourth was chosen as an expression of social justice values, which could then be compared to these other well-examined factors. Indeed, the primary goal of this experiment, in addition to contextualizing specific nano-applications, was to assess the influence of perceived social justice alongside familiar constructs associated with risk perceptions. Given the significance of vulnerability and environmental justice scales, we expected that contextualized presentations of nano-applications would also elicit concerns for social justice in a way that may be less evident when nanotechnologies are considered as a class defined largely by its charismatic potential for great benefits.
Each of the four variables has a positive and negative form that was tested across three domains: nanofood additives (food), nanopharmaceuticals (medicine), and nanofuel additives (energy) (Table III; Appendix). For instance, the following narrative scenario about nanotechnology-enabled food begins with a statement describing the application and the rationale for developing it (a benefit statement):
| Positive Valence | Negative Valence | |
|---|---|---|
| Benefit/No Risk | Scientists are positive about the benefits of these. | Scientists are worried about the health risks of this. |
| Bodily Invasion/No Invasion | Even though these materials are small, they cannot be absorbed into the bloodstream through the intestines. | Because they are so small they may be absorbed into the bloodstream through the intestines. |
| Control/No Control | They are designed to expel quickly through sweating or urination. | We may not be able to control or retrieve these if this happens. |
| Social Justice/Social Injustice | Currently, the production of nanofood is occurring in both poor and well-off neighborhoods. | Currently, the production of nanofood is only occurring in poor and minority neighborhoods. |
Nanotechnology may be used in food to improve its taste and nutritional value. (1) Scientists are positive about the benefits of this. (2) Because nanoparticles are so small, they may be absorbed into the bloodstream through the intestines. (3) We may not be able to control or retrieve these nanoparticles if this happens. (4) Currently, the production of nanofoods is occurring in both poor and well-off neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
In this example, the first variable statement (scientists’ risk judgment) is positive. The second variable statement (bodily invasion) is negative and so indicates that nanoparticles may invade the body. The third variable provides a negative statement about the controllability of the nanomaterial. The final variable statement describes the location of production, in this case, it is equitably distributed (social justice). This example illustrates condition 4 for “Nanofood” in Table IV. Each variable statement appears 24 times overall in the experiment.
| Condition | Risk | Control | Bodily Invasion | Social Justice | Nanofood | Nanopill | Nanofuel | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Mean | SD | N | Mean | SD | N | Mean | SD | |||||
| 1 | – | – | – | – | 34 | 5.94 | 1.54 | 42 | 5.62 | 1.62 | 34 | 5.65 | 1.98 |
| 2 | + | – | + | – | 42 | 5.14 | 1.95 | 47 | 4.30 | 1.74 | 39 | 4.82 | 1.90 |
| 3 | + | + | – | – | 40 | 5.63 | 1.67 | 50 | 5.54 | 1.74 | 33 | 5.15 | 1.54 |
| 4 | + | – | – | + | 42 | 5.48 | 1.77 | 40 | 4.40 | 1.96 | 44 | 4.98 | 1.85 |
| 5 | – | + | + | – | 42 | 5.64 | 1.57 | 49 | 4.92 | 1.96 | 45 | 5.36 | 2.09 |
| 6 | – | – | + | + | 51 | 4.88 | 1.61 | 46 | 4.04 | 1.87 | 42 | 4.55 | 1.58 |
| 7 | – | + | – | + | 36 | 5.50 | 1.54 | 46 | 4.78 | 1.76 | 39 | 5.31 | 1.44 |
| 8 | + | + | + | + | 44 | 5.07 | 1.86 | 44 | 3.45 | 1.76 | 43 | 3.21 | 1.91 |
| N | 331 | 364 | 319 | ||||||||||
| Overall | 5.41 | 4.63 | 4.88 | ||||||||||
- Notes: The fractional design employed eight conditions across three domains of nanoapplications.
- The alias structure is represented by the presence (+) or or absence (–) of each factor.
An optimal “full” factorial model of four binary variables would have required 16 vignettes per domain to capture all of the combinations of each of the variables. To reduce the sample size needed to evaluate a full factorial model, we instead employed a 24 - 1 fractional, or half-factorial, design that reduced the number of variable combinations to eight conditions per domain. This is a resolution IV design in which the defining relation I = RCIJ. As depicted in Table IV, the alias structure is RI = CJ, RC = IJ, and RJ = IC. This permits testing of main effects, which was the focus of this experiment, and we assumed that higher order interactions are negligible.(49,50) As such, this design allowed for evaluation of the four variables, which are uncorrelated, without generating unnecessary information on higher order combinations, and thus permitting a smaller sample size adequate for the analysis. This design avoids problems associated with direct questioning, given the finding that study respondents do not necessarily use the attributes in a decision scenario in the way they self-report or think they have.(42,51)
Each participant received one scenario from one application domain for a sample size of between 319 and 364 per domain. After being read the requisite narrative, participants were asked to rate the acceptability of the scenario on a seven-point scale. We then used these ratings as the dependent variable against which we could assess the influence of each variable.
Responses ranged from 1 to 7, where 1 equals “strongly support,” 7 equals “strongly oppose,” and 4 equals “neutral.” Except for two conditions (condition 8 for Nanopill and Nanofuel), all conditions were on average rated neutral to strongly oppose (mean > 4.00). No overall mean score for any application scenario can be read as supportive. The two conditions that were (weakly) supported were defined by all positive attributes. In other words, they were defined by scientists’ assessment of low risk, controllability, impossibility, or low consequence of bodily invasion by nanoparticles and by a socially just production process. Except for these two instances of guarded support for applications framed as entirely beneficial, participants tended to be neutral or opposed to these nano-applications. All nanofood applications were unsupported, including the entirely positive condition 8.
Condition 8 of the narratives, where all attributes were offered in their most positive form, is akin to how people are “hearing” about nano right now. This typical description of a nanotechnology application takes the form of a description of the object plus a rationale for its usefulness in the form of a claim of benefit. Nonetheless, the acceptance of these highly positive accounts is marginal at best. These descriptive findings reveal an overall portrait of participants as risk averse that contrasts with the benefit-centricity reported by most of the other nano-related surveys to date.
At first glance, the means of the application scenario ratings (4.63/pill; 4.88/fuel to 5.41/food) do not seem that different, suggesting a lack of sensitivity to application domain. However, there are noticeable differences between experimental conditions within domains and these help define some important differences between domains. For nanopill and nanofuel, the change in mean scores between condition 8, the most positive, and condition 1, the most negative, corresponds to a change from roughly neutral/slight support to firm opposition. Compared to those two domains, there is less variation in the difference in mean scores between most and least supported conditions within the nanofood application domain (Δ 1.06 for nanofood vs. 2.17 for nanopill and 2.44 for nanofuel) and ranges roughly from minimally to strongly oppose. Thus, the nanofood application is roundly unsupported, even in its most positive versions, suggesting that the domain as a whole (unlike the other two domains where some scenarios are supported) is the subject of heightened risk aversion. This is consistent with prior findings of risk aversion in the context of nanofood and food packaging applications.(12,21)
Table V reports the results of the analysis of the two-level fractional factorial design using ANOVA. In addition to the three application domains we also examined the four factors simultaneously across all domains, referred to as the “all applications” model. Bodily invasion was a significant factor in each application domain, which is consistent with the psychometric literature on risk perceptions. Justice influenced the support for nanofuel, F(1, 314) = 17.84, p < 0.001, nanopill, F(1, 317) = 15.30, p < 0.001, nanofood, F(1, 326) = 4.14, p= 0.04, and all applications, F(1, 1003) = 39.93, p < 0.001. Scientists’ risk judgments significantly influenced the ratings of the nanopill, F(1, 314) = 17.01, p < 0.01, and all applications, F(1, 1003) = 16.17, p < 0.001, but not the ratings of nanofood or nanopill. Controllability was nonsignificant in each model.
| Domain | Dimensions | Sum of Squares | df | F | Partial n2 |
|---|---|---|---|---|---|
| Nanofood | Scientists’ risk judgments | 3.07 | 1 | 1.09 | 0.003 |
| Control | 0.77 | 1 | 0.27 | 0.001 | |
| Bodily invasion | 20.47 | 1 | 7.26** | 0.02 | |
| Justice | 11.67 | 1 | 4.14* | 0.01 | |
| Error | 908.10 | 322 | |||
| Corrected total | 944.11 | 326 | |||
| Nanopill | Scientists’ risk judgments | 7.03 | 1 | 2.16 | 0.007 |
| Control | 1.29 | 1 | 0.4 | 0.001 | |
| Bodily invasion | 45.29 | 1 | 13.92*** | 0.04 | |
| Justice | 49.77 | 1 | 15.30*** | 0.05 | |
| Error | 1018.28 | 313 | |||
| Corrected total | 1112.78 | 317 | |||
| Nanofuel | Scientists’ risk judgments | 52.49 | 1 | 17.01*** | 0.05 |
| Control | 8.86 | 1 | 2.87 | 0.009 | |
| Bodily invasion | 55.71 | 1 | 18.06*** | 0.06 | |
| Justice | 55.06 | 1 | 17.84*** | 0.05 | |
| Error | 956.57 | 310 | |||
| Corrected total | 8670.00 | 314 | |||
| All Applications | Scientists’ risk judgments | 51.41 | 1 | 16.17*** | 0.02 |
| Control | 1.06 | 1 | 0.33 | 0.000 | |
| Bodily invasion | 135.75 | 1 | 42.71*** | 0.04 | |
| Justice | 126.94 | 1 | 39.93*** | 0.04 | |
| Error | 3175.56 | 999 | |||
| Corrected total | 3480.16 | 1003 |
- *p < 0.05; **p < 0.01; ***p < 0.001.
- Notes: No risk = 1, expert concern for risk = 0; ability to control = 1, inability to control = 0; impossibility of bodily invasion = 1, bodily invasion = 0; socially just = 1, unjust = 0.
4. DISCUSSION AND CONCLUSION
This research demonstrates that justice, with some exceptions, plays an important role in the formation of risk perceptions related to novel technological risk objects. It provides mechanisms for assessing the possible implications of largely unfamiliar risk objects, based on prior experiences and awareness of inequalities in the social distribution of both the risks and benefits of technological change. We were able to differentiate respondents in meaningful ways by their attitudes toward vulnerability and environmental justice, and such judgments were comparatively predictive thereby confirming our first hypothesis. That said, the briefer risk versus safety vignettes, which intentionally provided risk-only and safety-only statements following each nanotechnology's purpose, revealed only some differentiation, thereby offering less than fully robust confirmation of the second hypothesis (that justice and vulnerability will produce significant sensitivity to risk vs. safety information), while generating the surprising finding of risk aversion related to vulnerability in the context of safety messaging. This may be a function of design, and a possible indication of the many possibilities and tradeoffs that may have a part in the overall judgment. It may also be a suspicion effect, whereby the extra information offered in the vignettes is perceived by this group as a propagandizing manipulation, which is then met with suspicion and so a negative/risk rating.(52)
A more complex and systematically manipulated set of variables was, however, accomplished with the narrative design, the results of which uphold the third hypothesis, which anticipated that, in this upstream context (i.e., perceptions are as yet emerging), justice and vulnerability would prevail against known psychometric variables (controllability, dread-like qualities, and whether or not technologies are understood by scientists), under conditions of detailed and systematically manipulated information.
Many aspects of these findings were intriguing and, overall, justice remained a powerful variable against other known and well-established psychometric predictors of risk. The general nonsignificance of controllability in our narrative analysis is particularly interesting. Perhaps this is a result of the “upstream” moment, and over time concern for controllability may grow alongside familiarity. It would also be worth eliciting trust judgments and assessing more nuanced questions, such as controllability by whom? More generally, this line of questioning places emerging perceptions in an explicit “social-benefit” context, opening further questions, raised here and elsewhere, about trust in scientific experts, government regulation, and business.
These findings also demonstrate the benefits of greater specificity of application type and domain in nanotechnology risk perception research. Some domains, such as food, appear to be perceived differently from other domains. When the focus is placed on specific nano-applications, as in this analysis, public support tends more toward neutrality and lack or withdrawal of support. This is evident in the narrative analyses: when all four variables were positive (safety/benefit oriented), then and only then, two of three applications were rated as acceptable. Together this suggests that the public's likely acceptance of nano-enabled products is reliant upon a confluence of multiple factors. Not only must scientists declare them safe, but nano-enabled products must not be unpredictable as concerns bodily uptake or response and they must be manufactured in ways perceived as socially just. And even then, as the nanofood applications demonstrate, public acceptance may still not be forthcoming. While we were careful to ensure balance across positive and negative framing in this experiment, it appears that people are more sensitive to the negative information. This is consistent with the negativity bias in risk perception work, whereby we are most sensitive and attentive to negative information.(53) Finally, in terms of nano-enabled food, the robustness of bodily invasion in our experiments indicates that [nano]-food may trigger particularly strong reactions and concerns because it is consumed intentionally, but possibly unknowingly.(20) This is likely affected by cultural constructions of food and water that place special meanings on their consumption and embodiment, and that may distinguish it from other applications, such as medicine, that are also intentionally consumed. Still, it is the specific application and the particular way in which the public anticipates encountering it that drives perceptions of risk, not inadequate rationalization of benefits or charismatic features of the technological object encountered in the abstract.
Most broadly, the findings of this research suggest that technological objects are not encountered in the abstract or only in their toxicological dimensions but are incorporated in complex social relationships that help to define what the object is, what it is good for, as well as who benefits from it and who bears the costs. Extending ideas developed by economic sociologists,(54,55) risk scholars might benefit from consideration of the sociological concept “embeddedness,” which refers to how a technological object acquires its meaning in reference to its social contexts. An embeddedness perspective requires that one considers that the fields in which technologies emerge and exist are laden with power to the degree that technological risk objects are entangled in social institutions structuring opportunities for individuals and communities to control how the benefits and risks of that technology are distributed.(56,57) This is evident in recent, cross-national, deliberative research on nanotechnologies, which has shown how participants felt greater concern for social risk, including fairness in the distribution of benefits and risks across a stratified society, than for the technological or toxicological risks of nanotechnology.(22) Social risks include the possibilities of an unequal distribution of benefits and risks across a society, assessments of the fairness of that distribution, and recognition that some social groups are more vulnerable to bearing a disproportionate share of the ill effects of technological innovation while others monopolize the benefits.(57-59) Attitudes toward justice thus provide cues for the public to understand and evaluate the full range of likely social impacts of new technologies. A given technological object or system, such as nano-enabled cosmetics or medical devices, is not only evaluated as a discrete “thing” but as part of the social contexts in which it is encountered, including assessments of procedural and social justice. As a result, our research suggests that it is not just the toxicological risks of a nano-application that will face scrutiny, but also the processes by which that application is produced, disseminated (sold), used, and disposed of, and the associated implications for equity and social justice. Public perceptions of nanotechnologies are not just about toxicology or even psychometric and attitudinal variables, but are likely to be shaped by how they are expected to affect the distribution of risks and benefits across a diverse and unequal society, including the organizational and institutional processes for managing them.
Future research could adopt a more nuanced focus both on application domain and the social contexts in which they will be encountered and understood by social groups and persons in different social locations and ask: Under what conditions are concerns for justice made salient and when are they muted or trumped by other concerns? At the very least, questions of equity, vulnerability, and distributional justice should be addressed as new nano-applications are developed as they appear critical to the success of the nanotechnology enterprise, above and beyond toxicological health and safety.
Footnotes
ACKNOWLEDGMENTS
This material is based upon work supported by the National Science Foundation under Cooperative Agreements # SES 0531184 and # SES 0938099 to the Center for Nanotechnology in Society at UCSB and by NSF and the Environmental Protection Agency under Cooperative Agreement # EF 0830117 to the UC Center for Environmental Implications of Nanotechnology. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or the Environmental Protection Agency. This work has not been subjected to EPA review and no official endorsement should be inferred. Appropriate approvals were obtained for the ethical use of human subjects.
Appendix
These examples demonstrate all positive and negative forms of the variable statements in the narrative passage experiment across three domains: nanofood, nanopills, and nanofuel.
NANFOOD1
Nanotechnology may be used in food to improve its taste and nutritional value. Yet scientists are worried about the health risks of this. Because nanoparticles are so small, they may be absorbed into the bloodstream through the intestines. We may not be able to control or retrieve these nanomaterials if this happens. Currently, the production of nanofood is only occurring in poor and minority neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
NANFOOD8
Nanotechnology may be used in food to improve its taste and nutritional value. Scientists are positive about the benefits of this. Even though these materials are small, they cannot be absorbed into the bloodstream through the intestines. They are designed to expel quickly through sweating or urination. Currently, the production of nanofoods is occurring in both poor and well-off neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
NANPILL1
Nanotechnology “pills” are being developed that can be swallowed to provide exact delivery of medicines at the cellular level. Yet scientists are worried about the health risks of this. Because nanopills are so small, they may be absorbed into the bloodstream through the intestines. We may not be able to control or retrieve these materials if this happens. Currently, the production of nanopills is only occurring in poor and minority neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
NANPILL8
Nanotechnology “pills” are being developed that can be swallowed to provide exact delivery of medicines at the cellular level. Scientists are positive about the benefits of these. Even though these nanopills are small, they cannot be absorbed into the bloodstream through the intestines. They are designed to expel quickly through sweating or urination. Currently, the production of nanopills is occurring in both poor and well-off neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
NANFUEL1
An ultrafine powder made from nanoparticles is being developed to add to gasoline making it about 30% more efficient. Yet scientists are worried about the health risks of these. Because nanoparticles are so small, they may be absorbed into the bloodstream through the skin. We may not be able to control or retrieve these materials if this happens. Currently, the production of nanofuels is only occurring in poor and minority neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
NANFUEL8
An ultrafine powder made from nanoparticles is being developed to add to gasoline making it about 30% more efficient. Scientists are positive about the benefits of this. Even though nanoparticles are small, they cannot be absorbed into the bloodstream through the skin. They are designed to expel quickly through sweating or urination. Currently, the production of nanofuels is occurring in both poor and well-off neighborhoods. How acceptable is this on a scale between 1 and 7, where 1 is strongly support it, 4 is neutral, and 7 is strongly oppose it?
