Right answers and right-wrong answers: Sources of information influencing knowledge of nuclear-related information

Socio-Economic Planning Sciences 44 (2010) 130e140

Contents lists available at ScienceDirect

Socio-Economic Planning Sciences journal homepage: www.elsevier.com/locate/seps

Right answers and right-wrong answers: Sources of information influencing knowledge of nuclear-related informationq Michael Greenberg a, *, Heather Truelove b a b

EJ Bloustein School of Planning and Public Policy, Rutgers University, 33 Livingston Avenue, New Brunswick, NJ 08901-1958, USA Department of Civil & Environmental Engineering, Vanderbilt University Station B#351831, Nashville, TN 37235-1831, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 8 April 2010

Surveys in 2008 and 2009 asked almost 6000 United States residents to indicate their knowledge about the use of nuclear and other sources of energy, and the disposition of nuclear waste. Less than 10% of respondents knew where spent commercial nuclear fuel is stored. With regard to knowledge about fuel for electrical energy, respondents overestimated solar and wind use and underestimated coal use. These responses are consistent with mass media coverage of these issues. The mass media were the source of information for the vast majority of respondents. However, the likelihood of right answers to our questions increased as reliance on the mass media decreased, and it increased with use of books, magazines, personal contacts and the web. Educated affluent white males with strong preferences for nuclear energy disproportionately were knowledgeable. These observations demonstrate the daunting challenge of providing information about subjects that are largely distant and disconnected from the public’s lives. The Department of Energy, Nuclear Regulatory Commission, state and local health, environmental and energy agencies, and facility owners and operators have huge domestic political, national security and economic stakes in improving the factual grounding of public reactions to energy production and waste management choices. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Knowledge Nuclear power Waste management

1. Introduction The objectives of this paper are to examine public knowledge of certain facts about nuclear energy and waste management, and to assess the correlates of that knowledge. These objectives arise from a long-standing ambivalence about the media’s role in shaping the policy agenda [1e6]. Indeed, some have characterized the public as uninformed and overly influenced by the media and emotions, and the media are accused of amplifying relatively small risks into major fears [1e6]. The role of the media in shaping public perceptions and preferences about energy has been highlighted during the last several years by concerns regarding the management of high-level nuclear waste and a so-called renaissance of nuclear power. With regard to waste management, the major controversy has been whether to store high-level nuclear waste from commercial nuclear power plants and the defense industry at Yucca Mountain, which is located about 90 miles north of Las Vegas, Nevada [7,8]. q The Consortium for Risk Evaluation with Stakeholder Participation at Rutgers University and Vanderbilt University. * Corresponding author. Tel.: þ1 732 932 4101x673; fax: þ1 732 932 6564. E-mail address: [email protected] (M. Greenberg). 0038-0121/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.seps.2010.04.001

During and after the presidential election of 2008, heated exchanges took place between the candidates for president, as well as the incoming head of the United States Senate who is from Nevada and is opposed to the use of Yucca Mountain [9]. With regard to energy, the major focus has been on the need to become more reliant on solar, wind, and other renewable sources and less on fossil fuels [10]. Meeting these objectives would reduce greenhouse gas emissions and dependence of the United States on foreign sources of fossil fuels. In this regard, nuclear energy has frequently been portrayed as a source of energy that does not create greenhouse gases [11,12]. It is not the purpose of this paper to argue for or against Yucca Mountain, or for or against more nuclear power plants. Rather, given the heightened publicity around these issues, this paper serves as an opportunity to determine what facts the United States public has been taking away from the debates and what perceptions, values, preferences, and sources of information are associated with their knowledge. The purposes of the research summarized in this paper were to answer three research questions: 1. Knowledge: What proportion of the U.S. population knows where high level nuclear waste from commercial power plants is

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

managed? What proportion knows how much the United States relies on nuclear, coal, wind and solar energy for electrical power generation? What proportion knows if there are nuclear power, waste management, laboratory or mining facilities in their state? 2. Sources: What is the association between energy-related knowledge and sources of information such as, mass media, internet, books and magazines, and personal contacts? 3. Correlates: What respondent attributes such as, demographic characteristics, preferences and values about energy and the environment, political, cultural and social identity, and respondent location would increase the likelihood of accurate energy-related knowledge? Behind this research is our belief that facts about energy policy are mostly unconnected and distant factoids for most people, that is, they have not likely visited Yucca Mountain, a nuclear power plant, a waste management facility, or a uranium mine. Consequently, when they read a news story, listen to a radio news story or watch a television news story, it is not surprising that they do not come away with a clear message. Accordingly in addition to the “right” or correct answers, we believe that many members of the public have learned “right-wrong” knowledge, or media enhanced knowledge, because of their reliance on the media. Their understanding is guided by media presentations, which are limited to selected information with limited time for contextual information. Distracted by the realities of life, the public, with some exceptions we hoped to discover, does not have the time to verify the reality of facts that are too distant from their lives. While the mass media surely are the main source of information available to the public about energy sources and waste management, some rely on the media less. We sought to examine the attributes of those who are more knowledgeable about energy-related issues and to determine the role of reliance on mass media in this knowledge. 2. The origin of right and wrong knowledge The environmental psychology, risk perception and communication, and more generally, the social science literatures are sources for understanding factors that influence people’s knowledge of energy-related issues. However, before summarizing elements of the literature, we note that the psychology, communications, risk and environmental literatures have examined the role of knowledge as (1) a variable to be predicted, (2) a cause of perception, and (3) as a mediating factor in perception and behavior. For example, Siegrist and Cvetkovich [13] found that people who were knowledgeable about risks did not rely on trust to assess a hazard. Those who were unfamiliar with the hazard relied heavily on their trust of the operators and mangers of the hazard. Knight [14] examined the influence of knowledge, morality, trust, and benefits on public support for biotechnology. He observed that knowledge was an indirect predictor, mediated by trust, which in turn was a weaker predictor than morality and benefits. Sjoberg [15] focused on public perception of scientists’ knowledge of risk. He observed that there is only a relatively weak association between perception and trust because people believe that scientists do not adequately understand all the impacts of technology. In another paper Sjoberg et al. [16] noted that knowledge of radiation risks was a good predictor of concern about nuclear risks among nuclear power plant workers. Vandermoere [17] examined how people perceived the need to remediate contaminated soil. He found that the public’s direct knowledge and their estimate of how much experts knew were both predictive factors. Grasmuck and Scholz [18] also examined soil pollution, finding that those who believed that they were knowledgeable did not want additional information, but overall emotional reaction was the strongest predictor.

131

The authors of this paper divided the explanatory factors for variation in knowledge into five groups: (1) demographic characteristics; (2) political, cultural and social identity; (3) preferences and values about energy and the environment; (4) respondent location-related attributes; and (5) information sources. First, we investigated the effect of demographic characteristics on energy-related knowledge. The authors anticipated a so-called “white male effect,” in other words, affluent and college educated white males were expected to be the most knowledgeable about energy-related issues. This group as a whole has had more formal education, access to information and power, and a vested interest in maintaining a grasp on factors that will strongly influence the economy of the United States. They tend to trust technology more than their counterparts and to be relatively less concerned about many hazards [19e24]. It follows that less educated and less affluent African and Latino American women and men were expected to be less knowledgeable. Second, we evaluated the role of political, cultural and social identity in explaining energy-related knowledge. Political party identification has been found to be a predictor of environmental risk perceptions and political orientation has been found to lead to attitudes about nuclear power [25e27]. As such, we expected political identity to predict energy-related knowledge. In addition to political identity, cultural identity, part of cultural orientation, has also been shown to be an important predictor of environmental risk perception and may at least partly explain the white male effect [25]. The white male effect has been challenged by proponents of an explanation that demographic characteristics are one element of cultural identity, part of cultural orientation. Specifically, Kahan et al. [25] characterize white males as tending to be individualistic (rather than communitarian), and hierarchical (rather than egalitarian). White males should be expected to be more aggressive than their counterparts about actions that will rapidly achieve their objectives. Hence, they are likely to be aware of facts that influence the direction of the economy. Anything that threatens the U.S. market system that U.S. white males have largely controlled, such as not having a place to store spent nuclear fuel or insufficient electrical energy, should be disproportionately important to individualistic and hierarchical people. The authors added indicators of political identification (self identify as Democrat, Republican, and Independent) and several questions about perceptions about discrimination, constraints that should be posed on how individuals spend their wealth, and the implications of the decline of the traditional family from Kahan et al.’s [25] research. Third, we investigated the effect of preferences and values about energy and the environment on energy-related knowledge. More accurate knowledge was expected from respondents who noted a strong commitment to environmental protection issues and those who were concerned about the overall quality of the environment in the future as those who value the environment have been known to seek out information about threats to the environment [28]. Trust of private and public authorities that manage energy facilities was also expected to be associated with more knowledge of facts about those facilities [22e24,29]. Fourth, we evaluated the effect of respondent location on energy-related knowledge. Studies show that the public focuses on hazards that affect them, their family and friends [30e32]. Consequently, many people living near nuclear waste management sites, nuclear laboratory facilities, and nuclear power plants should be expected to be motivated to learn more about facilities. Further, respondents who have worked at an energy production or waste management site or who have a friend, neighbor or other family member who has worked at a site are expected to have more energy-related knowledge. Additionally, if the site contributes to

132

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

the respondents’ income and/or the tax base of their jurisdiction, then they should have accumulated more knowledge. Fifth, we investigated the effect of information source on energy-related knowledge. The main categories are television, radio, newspapers, magazines, books, web sites, and personal contacts. Each of these has innumerable subcategories. Yet, the overall objective of this pilot study was not to make distinctions between local and national newspapers and different television channels. Rather it was to determine if the most knowledgeable members of the public rely as expected more on their own research by consulting books, magazines, the web and personal sources. Given the media’s coverage of energy, our expectation was that the media is amplifying negative messages about fossil fuels and positive messages about renewable energy sources, while increasingly offering more balanced views about nuclear fuel sources during the sampling period [33,34]. Accordingly, we expected respondents to overestimate the proportion of electricity usage from renewable energy sources and underestimate the proportion from fossil fuels. We did not have a strong a priori hypothesis about participants’ perceptions of the proportion of electricity usage from nuclear energy. However, in 2009 in balance we expected to find media amplified answers to be slightly weighted toward support for nuclear energy, but would not have been surprised if they were not. Despite these expectations, there was also a real possibility that the media might have little influence. While the literature asserts that increasingly the media not only seek to assert political influence but also built profitable audiences, which implies that the media could strongly influence public knowledge, other media studies assert that its role in influencing public opinion is not at all clear [35,36]. 3. Data and methods 3.1. Sample During the periods June 26, 2008 through August 22, 2008 and June 23, 2009 through August 14, 2009, phone surveys written by the first author were conducted under the auspices of the Center for Survey Research at the E.J. Bloustein School, Rutgers University by Abt SRBI. The first survey (2008) focused on siting of nuclear facilities and public preferences for alternative sources of energy [37], and the second survey (2009) examined public response to a new energy policy and public perceptions of nuclear, coal and renewable sources of energy [38]. These were designed to take 17e18 minutes and be conducted using land line telephones with random digit dialing (RDD) following American Association for Public Opinion Research standards. The first had 25 main questions and 76 queries and the second 26 questions and 69 queries. The first included a national sample of 600 participants and 2101 residents from 11 regions centered on nuclear power plants, major nuclear weapons and cleanup sites, and laboratories. The second included a national sample of 800 and 2400 residents of six regions that included a combination of nuclear and coal-related facilities (see below for more detail). The RDD sampling approach gives every working land line residential telephone number the same chance of being contacted for an interview. Because listed, unlisted, and not-yet-listed landline numbers are included, the process eliminates “listing” bias. Bad numbers such as, not in service and non-resident are not included. One limitation of RDD land line surveys is that they do not reach those without phones and those who use answering machines and other devices to screen callers. Also, it does not reach those with only cell phones. This limitation can reduce the sample of poor and younger people [39,40].

Sampling bias is always present, that is, samples underestimate some subpopulations and overestimate others. It was important that the regional samples be as representative of their regional populations as possible. We examined the bias in order to determine which demographic factor or factors could reduce the bias. Age and white-nonwhite were the most effective, moreso than for example, income, education and others we examined. Accordingly, the samples were weighted by age (18e44, 45e64, 65þ) and white-nonwhite. However, it is not possible to entirely correct by weighting because not all factors that influence results are weighted. Additionally, considering that response and cooperation rates for RDD surveys have been dropping in the United States from over 50% to now 20% [39,40], an eight call-back design was used for the first survey and an 11 call-back design for the second survey in order to obtain a response rate of 20% and a cooperation rate of 30%. For clarification, a response rate is the proportion of completed interviews among the number of eligible respondents. There are different versions of the definition of response rate because there are versions of respondent eligibility. Another issue is what is a “completed” survey. Completed may mean that at least 80% of the questions were answered. In other words, a 100% completed form is not required for a survey to be classified as completed. The co-operation rate is the number of completed interviews among the contacted and eligible respondents. The authors have always used the American Association for Public Opinion Research (AAPOR) definitions, but there are others with slightly different definitions. 3.2. Locations The most important point about locations was the focus on areas with major nuclear facilities. This deliberate choice was made because if new nuclear facilities are to be constructed, it is highly likely that they will be built at sites that already have nuclear facilities [37]. Hence, it is important to learn the preferences, values and perceptions of those who live near these facilities. The regions chosen for the first survey included persons who lived within 50 miles of 6 of the existing major U.S. Department of Energy facilities: Hanford (WA), Idaho National Laboratory (ID), Los Alamos (NM), Oak Ridge (TN), Savannah River (SC), and the Waste Isolation Pilot Plant (WIPP) (NM). Each of these has major nuclear waste management facilities. Los Alamos, Idaho and Oak Ridge are also major DOE research facilities. Five of the 11 sites were chosen as locations for new nuclear power plants: Calvert Cliffs (MD), Nine Mile Point (NY), North Anna (VA), Palo Verde (AZ), and South Texas (TX). These are among the locations for possible new nuclear power plants adjacent to existing nuclear power plants. The second survey included six regions each with a radius of 100 miles. Four of the six were included in the first survey: Hanford [WA], Los Alamos [NM], Oak Ridge [TN], and Savannah River [SC]. Two other sites representing the east and west coasts without large DOE waste management facilities were included in the second survey. The first was an area in west central California site-region includes the Diablo Canyon nuclear power plant at San Luis Obispo and coal plants in the 100 mile area. The second was an area in eastern Pennsylvania site region that includes parts of New Jersey, Delaware, Maryland and a small part of New York State, and includes five nuclear power plants and numerous coal facilities. Together, these two regions have six nuclear power plants and dozens of coal facilities and their waste management facilities. The choices of site-regions with radii of 100 miles in the second survey and 50 in the first were made after much deliberation. Prior studies have shown relatively limited knowledge about local energy sites beyond 20 miles [37,38]. Areas within 20 miles often

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

have people who work at the sites or have relatives and friends who do, and the facilities often have strong economic impacts on surrounding areas. If the sample population was drawn only within 20 miles of an energy site, the authors expected a bias toward more knowledge and understanding about energy issues. A 100 mile sample area, however, includes relatively few persons with local ties to local facilities. Yet, in the second survey areas with 100 mile radii were required to obtain regions with multiple types of energy production and waste management. In order not to lose an expected site-specific effect, the survey recorded county of residence, which then was used to represent a host county effect. Furthermore, respondents were asked about their familiarity with local energy facilities. Hence, a local effect should be measurable. 3.3. Questions The 2008 survey had one nuclear knowledge question. The debate about nuclear waste management was highlighted in 2008 around the Yucca Mountain controversy. How many people know that “spent” or “used” nuclear fuel from power plants is stored at existing sites versus sending it and storing it at Yucca Mountain? This question was deliberately written as open-ended without a set of options to choose from. Actual responses were recorded as six answers, including “don’t know.” Because the single question in the 2008 survey produced interesting results, the 2009 survey had three open ended questions about energy. The first of these listed energy facilities and asked the respondent to tell us if “your state has one of these facilities or locations.” The focus was on nuclear power plants, nuclear waste management and laboratory facilities, and uranium mines. Since we had access to government data bases, we could compare each respondent’s knowledge to reality. The next and key question stated: “For each of the forms of power generation I am about to read, please tell e as a percent- how much the United States relies on each to generate electricity to power homes and businesses.” No guidance was provided on the actual proportions. The numbers provided by the respondents were compared to the actual numbers for 2008 from the Department of Energy’s Energy Information Administration’s data base. If there is a media correct answer, then as noted earlier, we should find a disproportionate number of wrong answers in the direction of media coverage (“rightewrong” answers). For example, solar and wind have received so much positive emphasis that we expected many people to believe that the proportion is 10%þ, rather than about 1%. In contrast, coal has been associated with greenhouse gases, and hence we would expect respondents would be influenced toward underestimating its contribution. With regard to nuclear power, we were less certain what to expect because while nuclear seems to be treated more favorably in some media there remains concern reflected in media articles [33,34]. There is no comprehensive list of sources that can be consulted to determine media emphasis and thereby to determine what is a “right-wrong” answer. The first author chose the list of 10 stories identified daily by Joseph Davis for the Society for Environmental Journalists (SEJ) as important environmental stories and sent to SEJ members. In correspondence with the author, Davis noted that he does not consider his list to be random because it is only 10 stories a day and it tends to draw heavily from larger newspapers. Yet the other options were arguably even less random. The first author read the stories from March 2009 through June 2009. During this period of time, there was considerable negative coverage of fossil fuels and global climate change and relatively balanced coverage of nuclear power and waste management. However, we are not contenting that this represents all the coverage, it was a feasible way of monitoring media coverage.

133

Coverage can change. For example, on February 10, 2010, three negative nuclear articles were published. Matthew Wald [41] of the New York Times reported some concern on the part of several experts nominated by President Obama to the Nuclear Regulatory Commission about tritium leaks at the Vermont Yankee nuclear plant. Another story questioned President Obama’s loan guarantees for nuclear power plants [42]. And a third noted the long-term contamination of the Columbia River by the Hanford nuclear waste management facility [43]. Readers of these three different newspaper stories would be more disposed to a negative opinion of nuclear power than the coverage during the months before this survey. While the 2009 survey was being conducted, the US EPA identified 44 coal sites as hazardous [44]. Also, in December 2008, a coal disposal facility in Kingston, Tennessee failed with serious damage to the surrounding area [45]. We stopped the survey for six days and added the following question: “Within the last year, have you heard or seen any news reports about problems with coal energy production or coal waste management?” This, again, was an openended question. One additional set of questions was added to the second survey in order to answer the second research question about sources of information. Respondents were asked “where do you get most of your information about nuclear, coal, wind and solar facilities in your state?” The goal was to distinguish between respondents who indicated that newspapers, television or radio were their main sources versus those that used the internet, magazines, books, or personal contacts. As the overall objective of this pilot study was not to make distinctions between local and national newspapers and different television channels, the broader categories of television and radio, newspapers, magazines and books, internet searches, and personal contacts were used. The vast majority of the questions in both surveys were identical. Indeed, unless otherwise indicated the questions are the same in both surveys. With regard to demographic characteristics, respondents were asked to indicate their age on their last birthday, the last grade they completed, and their total annual family income in five income categories beginning with less that $25,000 and ending with $100,00 or more. They were asked to indicate if they were Latino or Hispanic origin and what race they consider themselves to be (White, Black, Asian, Native American, and Indian). With regard to political, social and cultural identify, a set of questions was included in the 2009 survey but not in the 2008 survey. Each respondent was asked if s/he thinks of him/herself as a Democrat, Independent, Republican, or another party. The 2009 survey also asked six questions about social and cultural orientation. Participants’ responded to three items from Kahan et al.’s [25] Communitarianism-Individualism scale: “Too many people expect society to do things for them that they should do for themselves”; “The government interferes too much in our everyday lives”; and “If people who are successful in business should have a right to enjoy their wealth as they see fit” and three items from Kahan et al.’s [25] Egalitarianism-Hierarchy scale: “The United States would be better off if the distribution of wealth was more equal”; “Discrimination against minorities is still a very serious problem in the U.S.”; and “A lot of problems in our society result from the decline in the traditional family, where the man works and the women stays home.” A set of questions probed preferences and values about energy and the environment. Respondents were asked if the United States should increase or decrease our reliance on coal, nuclear, solar and wind sources of energy (a “stay the same response” was permitted). The 2009 survey also asked how the respondent would allocate federal government economic stimulus money between energy

134

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

production or conservation. Perceptions and value of the environment was examined by asking our respondents to self characterize their support of the environment as active; supportive but not active; neutral; and not concerned about environmental problems. Another question asked respondents to declare if they think the environment of their state as whole will be better, the same or worse 25 years from now. The 2009 survey asked three questions about trust of owners/ operators of nuclear and coal facilities, federal and state agencies that regulate these facilities. These questions combined competence to manage health, safety, and the environment as well as to communicate information to the public. The 2008 survey had separate questions about the federal agencies (the US Environmental Protection Agency, the Nuclear Regulatory Commission, and the US Department of Energy). The final set of questions probed for a host or local effect. Respondents who lived in one of the counties within 20 miles of one of the DOE or nuclear power plant sites were labeled “host county.” Respondents were asked if they, a family member, or friend worked at one of the four sites. Finally, respondents were asked to gauge the economic impact of a nearby energy facility. Surveys about these complex subjects could easily contain at least twice the number of questions than can be included. Consequently, with a goal of 17e18 minute phone survey, each survey was pilot tested with two dozen randomly chosen respondents. The results were used to make the final choices about which questions would be retained and to fine-tune the questions. 4. Results 4.1. Preliminary The response rates to the surveys were 20.1% for the first and 23.4% for the second survey. The corresponding cooperation rates were 29.8% and 40.6% respectively. These rates are consistent with response rates for many recent landline surveys that fall below 20% [39,40]. The difference in the cooperation rates between the two surveys, we believe are partly explained by the fact that the first survey used an 8 callback design and the second an 11 callback design. The results were weighted by region-specific age and white-nonwhite ratios. A total of 15.5 million people lived in the 11 areas of the first survey (50 mile radii areas) and 29 million in the 6 areas of the second survey (100 miles area radii). These correspond to 5% and 10% of the national population, respectively. While we did not receive the 60% response rate that was customary a generation ago, we noted that our response rate is comparable to those obtained in landline surveys [39,40]. We also weighted the results to partly correct for bias. And we compared the results to check for bias by date of call and number of attempts required to reach a respondent. For example, the 2009 survey asked people to tell us if they think the United States should increase or decrease their reliance on nuclear energy. There was almost no

variation in response to key questions by how many calls were required to reach a respondent. These findings allow us to feel confident that we did not need to do an extraordinarily expensive 20 call back design. 4.2. Survey 1. Public knowledge of disposal location of fuel from nuclear power plants The proportion who knew that commercial spent fuel is managed and stored at nuclear power plant sites was slightly less than 10% (Table 1). Almost 23% incorrectly believed it is stored at Yucca Mountain. In other words, the right-wrong media answer was over twice as prevalent as the right answer. Three other answers were as follows: it is reprocessed and used as new fuel; dumped in the ocean; and reprocessed into weapons grade material. Respondents from the six nuclear waste-lab sites were more knowledgeable than their counterparts. Over 11% knew the right answer compared to only 9% from the five nuclear power plant regions and 7% of national sample. Right-wrong answers were much more evenly balanced (Table 1). So few of the remaining respondents gave any answer that they were labeled the “no answer” group. The authors examined the relationships between the right, right-wrong, and no answer groups in response to the question of where spent nuclear fuel is stored. This was done using both analysis-of-variance and multinomial regression. Given the clarity of the analysis of variance results, they are presented in Table 2. The multinomial results are not presented to conserve space. The results are similar. The correlates of the right and right-wrong answers were similar to one another and notably both were different from the much larger no answer group with regard to demographic characteristics and fuel preferences. The first two groups tended to be disproportionately male, white, 35-64 years old, middle to high income and college graduates. They wanted the nation to rely more on nuclear power, less on coal and oil, and were less concerned about nuclear power as a hazard than their counterparts. They were also more likely to live near a nuclear waste management facility and to have a friend or family member who worked at a nearby energy site. Finally, they were more likely to believe the U.S. energy policy should focus on protecting the environment. The no answer group was predominantly composed of women, many of them were Black or Latino American, they were relatively young (18e24) or older (65þ), and were not college graduates. The no answer groups were clearly the least supportive of greater reliance on nuclear power, indeed were the most concerned about health and environmental problems associated with nuclear power. They were the most supportive of relying of fossil fuels. The no answer group disproportionately wanted government energy policy to protect their health and tended not to live near a nuclear waste management facility.

Table 1 Public Knowledge of Disposition of Used/Spent Commercial Nuclear Fuel in the United States, % 2008. Source

Answers, Total sample

Six nuclear waste/or lab

Five nuclear power plant

National sample

Stored at nuclear power plants above ground Stored at Yucca Mountain underground Reprocessed and used as new fuel Dumped in the ocean Reprocessed into weapons grade materials No answer

9.6 22.8a 1.7 1.3 0.5 64.1

11.6 25.2a na na na 59.0

9.0 22.3a na na na 64.8

7.0 19.5a na na na 70.8

na ¼ no answer because of small numbers. Aggregate of three for six nuclear waste/labs was 4.2%, 3.9% for nuclear power plant sites; and 2.7% for national sample. a Stored at Yucca is significantly higher than other choices at P < .001 in all four comparisons.

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

135

Table 2 Predictors of Public Knowledge of Disposition of Used/Spent Commercial Nuclear Fuel in the United States, % 2008. Comparison of Averages Predictor

Right

Right-Wrong

No. answer

Demographic: Male (1 ¼ yes, 0 ¼ no) Demographic: White (1 ¼ yes, 0 ¼ no) Demographic: Black (1 ¼ yes, 0 ¼ no) Demographic: Latino (1 ¼ yes, 0 ¼ no) Demographic: Age 35-64 years (1 ¼ yes, 0 ¼ no) Demographic: Annual income $75,000 plus (1 ¼ yes, 0 ¼ no) Demographic: College graduate (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on nuclear power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on coal power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on solar power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on wind power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on oil power (1 ¼ yes, 0 ¼ no) Preferences-Values: Nuclear power is not a harmful source of energy (1 ¼ yes, 0 ¼ no) Preferences-Values: Nuclear site will have an accident exposing people and the environment (1 ¼ yes, 0 ¼ no) Preferences-Values: Nuclear site will discourage people and business from moving into an area (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S government energy policy should focus on protecting my health (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S government energy policy should focus on protecting the environment (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S government energy policy should focus on improving our state economy (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent trusts owner/operator of nuclear facility to protect health, safety and the environment (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent trusts federal agencies (DOE, EPA, NRC) responsible for nuclear facility to protect health, safety and the environment (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent trusts their state agencies responsible for nuclear facility to protect health, safety and the environment (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent is active supporter of environmental protection (1 ¼ yes, 0 ¼ no) Preferences-Values: Environment will be better in 25 years (1 ¼ yes, 0 ¼ no) Location: Respondent lives in host county (1 ¼ yes, 0 ¼ no) Location: Respondent, family member or friend works/ed at a nearby energy facility (1 ¼ yes, 0 ¼ no) Location: Respondent lives within 50 miles of nuclear waste management facility (1 ¼ yes, 0 ¼ no) Location: Respondent lives within 50 miles of nuclear power plant (1 ¼ yes, 0 ¼ no) Location: Respondent does not live within 50 miles of nuclear power plant or nuclear waste management facility (1 ¼ yes, 0 ¼ no)

.756 .866 .016 .042 .665 .504 .550 .738

.644 .895 .042 .050 .612 .418 .482 .629

398** .708** .160** .151** .535** .254** .292** .375**

.263

.255

.349**

.950

.954

.930

.907

.920

.896

.120

.150

.267**

.613

.521

.285**

.239

.337

.498*

.140

.233

.357**

.357

.360

.435**

.307

.329

.257**

.183

.192

.153**

.631

.580

.613

.740

.690*

.762

.682

.639*

.689

.302

.235

.203*

.204 .200 .303

.151 .174 .247

.161 .143* .136**

.507

.467

.398**

.331

.343

.359

.151

.189

.242**

Signifies that sample is significantly different from other two groups. Chi-square test: ** P<. 001; *P < .01.

4.3. Survey 2 4.3.1. Question 1: Public knowledge of energy use and locations of nuclear sites The second survey provided many more opportunities to examine public knowledge and its correlates. Table 3 lists the actual

proportion of electricity provided by coal, nuclear, solar and wind sources in the U.S. as well as participants’ beliefs about the proportions provided by each source. The median response for coal was about 10% lower than the actual (40% vs. 49%), and it was almost 10% higher for solar and wind (10% vs. <1%). The nuclear power median was remarkably close to the actual number (20% vs.

Table 3 Public Knowledge of Proportion of Electrical Energy Sources in the United States, 2009, %. Source

Actuala

Survey Median

Survey Meane

“Right”b

Right-Wrongc

Wrong- wrongd

Other

Coal Nuclear Solar Wind

49 19 <1 <1

40 20 10 10

38 27 18 17

13.5 18.3 39.5 40.6

9.2 14.1 16.9 15.6

0.3 13.3 NA NA

77.0 54.3 43.6 43.8

a b c d e

Source: EIA, 2009. http://tonto.eia.doe.gover/energyexplained/index.cfm?page¼electricity_in_the_united_states. Accessed 11/20/09. Right means within 5% of the right answer. Right-wrong is within 5% of the right answer and in the direction predicted by media coverage. Wrong-wrong means within 5% of the right answer and in opposite direction predicted by media coverage. Survey means are significantly different from actual values at P < .01.

136

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

19%). However, the latter does not imply that many people came close to the actual proportion. About one-quarter of respondents believed that 40% or more of electricity was generated by nuclear power and a quarter assumed that the proportion was 10% or less. Wind as a source had the closest fit of actual and estimates reliance. Over half the respondents were within 5%. However, that is somewhat misleading because the modal answer was 5% and 10% was the typical second estimate. Furthermore 20% was the third most frequent answer. In short, our respondents have an ordinal understanding (coal first, nuclear second, wind and solar third and fourth) of use, but not much more than that about reliance on energy sources. It is unreasonable to assume that members of the public would be aware of the exact proportions. For purposes of this analysis, the “right” answer was defined as within 5% in either direction of the actual proportion. The “right-wrong” answer was a further 5% away from the right answer but in the direction implied by media coverage. Extending 5% in the direction opposite the media coverage are the “wrong-wrong” answers. The remaining answers were labeled “other.” These definitions are admittedly arbitrary, and they were tested by extending them another 5% in both directions. The associations between the categories of answers and the correlates did not notably change. About 40% of the wind and solar answers were right, and another 16% were right-wrong. There were no wrong-wrong answers because the actual proportion is less than 1%. Summarizing, about 55% of the respondents gave the right or right-wrong answers to the solar and wind questions. The coal answers were much less accurate. Only about 14% had the right answer, another 9% had the right-wrong answer (lower), and less than 1% provided the wrong-wrong answer. This means that 77% of respondents had no answer or an answer not close to the accurate one. With regard to nuclear power for electrical generation about 46% gave the right, right-wrong, or wrong-wrong answers. There were slightly more right-wrong than wrong-wrong answers. The remaining 54% provided answers that were less accurate. Table 4 shows that respondents who provided the right answer for one of the four sources also tended to know the right answer for the others. Interestingly, the right answers for nuclear, wind and solar were more strongly associated with each other than they were with coal. Overall, the average respondent had 1.1 right answers and 0.6 right-wrong ones. Twelve percent had 3 or 4 right answers, and 2% had 3 or 4 right-wrong answers. In contrast, 43% had 3 or 4 “other” answers. Did respondents know if there were nuclear facilities in their state? Table 5 shows that the majority did know. About four out of five respondents knew if they had a uranium mine in their state. Proportions for nuclear power plant dropped to about three of four, and the proportions declined another 10% for nuclear waste management and nuclear laboratory facilities. The drop-off in knowledge regarding nuclear waste management and laboratory facilities, we believe, was explained by the fact that there are many

Table 4 Association of Accurate Answers About Proportion of Electrical Energy Sources, 2009. Source

Measures

Nuclear

Solar

Wind

Coal

Contingency coefficient Ratiob Contingency coefficient Ratiob Contingency coefficient Ratiob

.139a 1.33 xx

.091a 1.10 .337a 1.46 xx

.083a 1.13 .340a 1.45 .581a 1.89

Nuclear Solar a b

xx

Contingency coefficient significant at P < .001. Ratio of actual versus predicted by change for right answers.

Table 5 Association of Accurate Answers About Each Fuel Source and Knowledge of the Presence of an In-State Facility, 2009. Type exists & is thought to exist

Accurate answers, %

Contingency coefficient

Nuclear power plant Nuclear waste management Nuclear laboratory Uranium mine

74 64 64 79

0.387* 0.199* 0.173* 0.390*

*P < .001.

laboratory and waste management facilities that are smaller and not as well-known as the larger facilities at Oak Ridge and Los Alamos. These larger facilities were well known by respondents in those areas, but university laboratory facilities were much less well known. Were those who were aware of the existence of these nuclear facilities in their state the most knowledgeable about sources of electrical energy? The answer is yes, which is testified to the contingency coefficient of 0.185 (p < .001) between right answers to proportion of energy source questions and knowledge of in-state nuclear facilities. As noted earlier, the U.S. Environmental Protection Agency released a list of problematical coal waste impoundments while the survey was being conducted [44]. The survey was suspended for six days, and a question was added to determine if respondents during the course of last year heard or seen any stories about “problems with coal energy production or coal waste management facilities.” A total of 2512 respondents were asked this question. If they said that they had heard something, then they were asked to describe what they had seen or heard. A total of 37% had. Nearly all of these were aware of the coal waste management failure in Kingston Tennessee, which had flooded a valley, leading to considerable environmental and property damage [45]. We found an association between knowledge of this incident with coal impoundments and knowledge of information about the uses of nuclear, coal, solar, and wind as sources of electrical energy. The contingency coefficient between the right answers to the question about the proportion of coal as a source of electricity and hearing about the coal impoundment incident was .191 (p < .001). More specifically, of the 305 (out of 2512) respondents who had 3 or 4 right answers to the energy source questions, 54% percent of them knew about the Tennessee coal incident. This compares to only 30% of the groups who gave no answer or only one right answer to the energy source questions. 4.3.2. Question 2. Association between information sources and knowledge of sources of electrical energy Over 70% of the respondents indicated that the mass media (newspapers, radio and television) was their primary source of information about nuclear, coal, wind, and solar facilities in their states. Our expectation was that less dependence on these mass media sources would be associated with more personal study and more knowledge about energy sources. Indeed, the proportion of right answers was associated with less reliance on the mass media (Table 6). Beginning with solar and wind as energy sources, 33% and 32% of those who knew the right answers, respectively, cited internet/web, book and magazines, or personal connections as the most important information source. This compared to 23% and 25%, respectively, for “other answer” respondents. This expected finding extended to the nuclear and coal answers. The active research modes were 35% and 33%, respectively, for right answers respondents compared to 23% and 25% for their other answer counterparts. This is not a striking finding, but it is consistent with the expectation of the media as the main source of

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

137

Table 6 Sources of Information About Energy Sources, % 2009. Source

Answer

Newspaper

Television & radio

Internet searches

Books, magazines reports

Personal contacts

Coal

Right Right-wrong Opposite Other

29 36 27 34

39 36 36 40

15 8 19 11

8 5 6 6

10 15 12 8

Nuclear

Right* Right-wrong** Opposite Other

33 39 29 31

33 35 44 46

12 11 14 10

6 11 8 5

17 4 5 8

Solar

Right* Right-wrong** Other

34 35 31

33 40 46

16 5 10

7 9 5

10 11 8

Wind

Right* Right-wrong Other

33 31 34

35 46 41

14 8 10

7 8 5

11 7 10

*Distribution of sources associated with right answer is significantly different from chance at P < .05. **Distribution of sources associated with right-wrong answer is significantly different from chance distribution at P < .05. See text for explanation.

information about personally distant issues for the majority of people, but less so for those who were motivated to have more accurate information. Focusing on the right and right-wrong responses, chi-square tests found that 5 of 8 produced statistically significant results. Two of these were for nuclear energy. Disproportionately those who gave the right answer relied on personal contacts, whereas those who gave the right-wrong answer relied on newspapers. With regard to solar energy, the right answer was associated with disproportionate reliance on the web, whereas the right-wrong answer with reliance on television and radio. Finally, those who had the right answer for wind relied disproportionately on internet searches. None of the coal relationships were significantly different. Yet, these insights may be confounded by other predictor variables. The next section examines the association between knowledge and information sources compared to other predictors of more knowledge, such as demographic attributes, trust, and others.

indications, with two exceptions. The right answer group strongly supported the idea of more reliance on nuclear power and trusted the responsible parties, whereas the no answer group was less supportive of nuclear power and less trusting of authorities that manage energy facilities. With regard to location-related variables, right answers were associated with living in a host county, being familiar with a nuclear site, and assessing the economic impact of the local nuclear facility as positive. The no answer group had the opposite characteristics. The last variable in the table measures the dependence on mass media versus active information search methods. While every group depended on the mass media, the notable finding is that the right answer group had the least dependence. Specifically, 33% of those who gave right answers said that their first information sources was from web searches, books/magazines, and personal contacts compared to 24% for all other respondents.

4.3.3. Survey 2: Correlates of right and right-wrong answers We focused on the distinction between right, right-wrong and other answers. The associations were evaluated by one-way-analysis-of-variance and ordinal regression. Due to space limitations, only the ordinal regression results are presented (Table 7), although the analysis-of-variance produced similar results. The demographic differences among the three groups were striking. The right answer group members are disproportionately male, white, middle or high income, and college graduates. The other answer group members were disproportionately female; many were Latino or African-American, 65þ years old, less affluent, and less formally educated than their counterparts. The right-wrong group was less differentiated from the right answer and other answer groups by demographic characteristics. Compared to these, it was more likely to have many respondents 45 to 64 years old and few had only a high school education. Yet, the right-wrong group was more distinguishable by its political and environmental preferences. Those who fell into the rightewrong group were more committed to solar and wind energy sources than their counterparts and they wanted the federal government to allocate more research dollars to investments in conservation rather than to new energy production projects. These right-wrong answer respondents also tended to self-identify as Republican. In contrast, the right and other answer respondents did not strongly identify as strongly with these preference and value

5. Discussion This research sought to investigate three major questions. First, what is the state of energy-related knowledge among the public? Results revealed that the public has a much better understanding of what sources we use for electrical energy and whether their state houses a nuclear facility than it does about where high level commercial nuclear waste is managed. The large majority of participants were accurate in their understanding of whether their state had a nuclear power plant, nuclear waste management facility, nuclear laboratory or uranium mine. In terms of their knowledge of the proportion of energy sources for electricity in the U.S., many respondents overestimated the proportion of solar and wind energy, and underestimated our reliance on coal, as we had anticipated. Estimates of nuclear power reliance were much more centered around actual reliance on nuclear fuels. With regard to the destination of commercial spent nuclear fuel, less than 10% knew where it was managed, and even a smaller proportion of the U.S. sample as whole knew that it is managed on site; far more thought that it was brought to Yucca Mountain. Second, we aimed to determine which participant attributes are most important in predicting energy-related knowledge. The survey showed a clear demographic distinction between those who knew the right answers about energy-related issues and those who

138

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

Table 7 Predictors of Knowledge About Proportion of Electrical Energy Sources from Ordinal Regression Numbers in the table are B-values, which are logistic regression coefficient estimates. Predictor

Right answer

Right-wrong Answer

Other answer

Demographic: Male (1 ¼ yes, 0 ¼ no) Demographic: White (1 ¼ yes, 0 ¼ no) Demographic: Black (1 ¼ yes, 0 ¼ no) Demographic: Latino (1 ¼ yes, 0 ¼ no) Demographic: Age 18e24 years (1 ¼ yes, 0 ¼ no) Demographic: Age 45e64 years (1 ¼ yes, 0 ¼ no) Demographic: Age 65þ years (1 ¼ yes, 0 ¼ no) Demographic: Annual income $50,000 to $99,999 (1 ¼ yes, 0 ¼ no) Demographic: Annual income $100,000 plus (1 ¼ yes, 0 ¼ no) Demographic: Less than high school education (1 ¼ yes, 0 ¼ no) Demographic: College graduate (1 ¼ yes, 0 ¼ no) Identity: Identifies as Republican (1 ¼ yes, 0 ¼ no) Identity: Identifies as Democrat (1 ¼ yes, 0 ¼ no) Identity: Identifies as Independent (1 ¼ yes, 0 ¼ no) Identity: Concerned that discrimination is a major problem in the U.S. (1 ¼ disagree, 0 ¼ agree) Identity: Wealthy people should be able to spend their resources as they choose (1 ¼ agree, 0 ¼ disagree) Identity: Problems in society result from decline in traditional family where man works and woman stays home (1 ¼ disagree, 0 ¼ agree) Preferences-Values: U.S. should rely more on nuclear power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on coal power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should rely more on solar and wind power (1 ¼ yes, 0 ¼ no) Preferences-Values: U.S. should allocate more energy-related science investments to conservation projects than to new energy projects (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent is active supporter of environmental protection (1 ¼ yes, 0 ¼ no) Preferences-Values: Environment will be better in 25 years (1 ¼ yes, 0 ¼ no) Preferences-Values: Respondent trusts public and private authorities to manage major energy faculties (1 ¼ least trust, . 3 ¼ most trust) Location: Respondent lives in host county (1 ¼ yes, 0 ¼ no) Location: Respondent, family member or friend works at a nearby energy facility (1 ¼ yes, 0 ¼ no) Location: Respondent assesses economic impact of nearby energy facilities as positive (1 ¼ yes, 0 ¼ no,)

.885*** .528*** .109 .787*** .236 .136* .368*** .426*** .389*** .407** .718*** .126 .014 .115 .113

.097 .114 .250 .065 .194 .177** .055 .072 .066 .907*** .030 .286* .056 .190 .072

.826*** .397*** .227 .699*** .324** .237** .349*** .460*** .311** .733*** .604*** .070 .043 .218* l.015

.051

.006

.049

.080

.006

0.084

.586*** .105 .030 .152

.073 .111 .543*** .341*

.470*** .152* .012 .092

.145

.062

.060

.091* 2 ¼ .188** 3 ¼ .252** .247** .225**

.057 2 ¼ .126 3 ¼ .019 .170 .104

.125* 2 ¼ .519*** 3 ¼ .478*** .042 .230**

1 ¼ .539***

1 ¼ .004

1 ¼ .534**

2 ¼ .511** .374*

2 ¼ .065 .125

2 ¼ .494** .215

0.262

0.047

0.243

Information: Mass media (TV, newspaper, radio) mentioned as source for energy-related information Nagelkerke pseudo r2 ***P < .001,**P < .01, and *P < .05.

did not. Right answers disproportionately came from collegeeducated and relatively affluent white men. They not only were more familiar than their counterparts with these selected facts but also wanted the nation to rely more on nuclear energy. The much larger no answer group was different from the right answer one with regard to demographic indicators. They were less likely to be college educated and were less affluent, were less interested in the nation relying more on nuclear energy, and they were disproportionately women. The right-wrong group was the most distinct. It sought greater reliance on solar and wind energy, less on coal, and desired investments in conservation all of which is consistent with the direction of mass media messages. Cultural identity did not predict knowledge about energy-related issues, which was surprising, and we do not have any insights why these were not significant predictors. Third, we sought to examine the relationship between energyrelated knowledge and information sources. The first glimpse at this relationship comes from our finding that respondents underestimated the proportion of coal for electricity generation and overestimated the proportion of solar and wind electricity generation. The media’s portrayal of fossil fuels such as coal in a negative light and renewables such as solar and wind in a positive light undoubtedly plays a role in these misperceptions. Give the mixed messages members of the public get from the mass media about

nuclear energy, it was not surprising that the media estimate was only 1% away from the actual reliance and that there were only a few more who overestimated use than underestimated it. We speculate that if asked 20 years ago, these electrical energy source questions would have produced more underestimates of dependence on nuclear power than overestimates. The most interesting overall insight is that those who rely on mass media for their energy-related information are less likely to correctly identify the proportion of electricity from different energy sources. The results of these two surveys are suggestive but inconclusive. Caution is warranted for four reasons. First, the surveys contained relatively few knowledge questions. A full study would require more questions about human and environmental health effects of radiation and other contaminants, greenhouse gases, and related issues. Second, the sample was geographically weighted toward locations with major nuclear facilities. We expected and found more knowledge among respondents from the nuclear areas than from respondents as a whole. The nuclear area focus weighted the results toward more knowledge. Third, we deliberately asked openended knowledge questions to limit guessing. For balance, a full knowledge-based survey should include a mixture of open-ended and fully labeled questions. Fourth, the battery of questions about sources of information was general and needs to be more specific. For example, we need to know which television show, which radio show, which newspapers and magazines, which web sites, and

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

other information sources are providing the information people remember and rely on. We are satisfied with the set of demographic indicators, preference, value and location indicators. More specifically, we think a great deal more work needs to be done on the role of specific sources. Readers will recognize that the coverage of energy-related issues by the CBS, NBC, ABC, Fox, MSNBC, CNN, local networks and many others can be quite different. Radio news has the same wide variation. Newspapers range from the national audiences for the New York Times, Wall Street Journal, Washington Post and Los Angeles Times to smaller daily regional papers and weekly local papers. It seems to us to be infeasible to measure all of these variations in one large study. Logic suggests that articles collected about newspapers from sources previously mentioned are a good place to start to compare major media coverage. The authors would like to see a study of Spanish-language newspaper coverage of energy compared to English-speaking coverage in the same regions. Data bases also exist for television coverage. Frankly, however, the authors would love to see a comparison of web-based coverage of energy choices and issues because while it may not the major source today, the trends point to a rapid increase in web-use for information gathering. With these caveats and suggestions for expanding the research noted, are these findings important for public policy? In some respects yes, in others, no. Knowledge of isolated facts does not mean that an individual understands the breadth of information needed to make an informed decision. Furthermore, gaining knowledge of isolated facts may be the result of existing preferences rather than a decision-making step. For example, those who already favor greater reliance on nuclear power may learn as many facts as possible to support their preference. Perhaps the most important conclusion to be drawn from this study is that it provides a dose of reality to those who believe that providing information is the key to changing public opinion about nuclear power. Studies show that the proportion of U.S. residents who favor nuclear power is increasing [46,47], despite limited knowledge about where nuclear waste is managed and even how much the nation relies on nuclear power. In pursuit of an informed citizenry it is probable that the public is using the information they are exposed to in order form mental models that favor solar, wind and other renewables and to oppose coal and other fossil fuels. The coverage of nuclear energy appears to be increasingly more balanced [33,34,48]. While only a preliminary finding, this study supports the idea that the mass media’s role in providing information to the general public about distant facts like energy and waste management is paramount but it is mediated by personal and geographical characteristics. Acknowledgments This research was prepared with the support of the U.S. Department of Energy, under Cooperative Agreement Number DE-FC01-06EW07053 entitled The Consortium for Risk Evaluation with Stakeholder Participation III awarded to Vanderbilt University. The opinions, findings, conclusions, or recommendations expressed herein are those of the authors and do not necessarily represent the views of the Department of Energy or Vanderbilt University. This report was prepared as an account of work sponsored by an Agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. We would like to thank Dr. Marc Weiner for his

139

assistance with the survey instrument and administration, and Charles Powers and David Kosson for their ongoing encouragement of this research. References [1] Kasperson R, Renn O, Slovic P, Brown H, Emel J, Goble R, Kasperson J, Ratick S. The social amplification of risk: a conceptual framework. Risk Analysis 1988;8 (2):177e87. [2] Mazur A. True warnings and false alarms: evaluating fears about the health risks of technology, 1948e1971. Washington, DC: Resources for the Future; 2004. [3] McCombs M. Setting the agenda: the mass media and public opinion. Malden, MA: Blackwell; 2004. [4] Fortunato J. Making media: the influence of constituency groups on mass media content. Mahwah, NJ: Lawrence Erlbaum Associates; 2005. [5] Graber D. Mass media and american politics. 7th ed. Washington, DC: CQ Press; 2006. [6] Krimsky S, Plough A. The emergence of risk communication studies: Social and political context. Science, Technology, and Human Values 1987;12(3/4):4e10. [7] Committee on Technical Bases for Yucca Mountain Standards. Technical Bases for Yucca Mountain Standards. Washington, DC: National Academy Press; 1995. [8] Macfarlane M, Ewing R, editors. Uncertainty underground: Yucca Mountain and the nations high-level nuclear waste. Cambridge, MA.: MIT Press; 2006. [9] Gable E. Yucca mountain: Reid declares storage proposal dead. E&E, August 13, 2009. [10] Cooper H, Broder J. Obama presses case for renewable energy. New York Times, www.nytimes.com/2009/10/24/us/politics/24obama.htm; October 23, 2009 [accessed 26.10.09]. [11] Rutenberg J. Solution to greenhouse gases is new nuclear plants, Bush says, www.nytimes.com/2006/05/25/washibngton/25/bush/html; 2006 [accessed 22.12.09]. [12] Kanter J. Sweden takes another look at nuclear power. New York Times, www. nytimes.com/2009/02/06business/./06nuke.html; 2009 [accessed 22.12.09]. [13] Siegrist M, Cvetkovich G. Perceptions of hazards: the role of Social trust and knowledge. Risk Analysis 2002;20(5):713e20. [14] Knight A. Intervening effects of knowledge, morality, trust and benefits on support for animal and plant biotechnology applications. Risk Analysis 2007;27(6):1553e63. [15] Sjoberg L. Limits of knowledge and the limited importance of trust. Risk Analysis 2002;21(1):189e98. [16] Sjoberg L, Drottz-Sjoberg B-M. Knowledge and risk perception among nuclear power plant employees. Risk Analysis 2006;11(4):607e18. [17] Vandermoere F. Hazard perception, risk perception, and the need for decontamination by residents exposed to soil pollution: the role of sustainability and the limits of expert knowledge. Risk Analysis 2008;28(2):387e98. [18] Grasmuck D, Scholz R. Risk perception of heavy metal soil contamination by high-exposed and low-exposed inhabitants: the role of knowledge and emotional concerns. Risk Analysis 2005;25(3):611e22. [19] Finucane M, Slovic P, Mertz CK, Satterfield T. Gender, race, and perceived risk: the ‘White Male’ effect. Health, Risk & Society 2000;2:159e72. [20] Flynn J, Slovic P, Mertz CK. Gender, race, and perception of environmental health risks. Risk Analysis 1994;14(6):1101e8. [21] Greenberg M. Concern about environmental pollution: how much difference do race and ethnicity make? a New Jersey case study. Environmental Health Perspectives 2005;113(4):369e74. [22] Nye J, Zelikow P, King D. Why people don’t trust government. Cambridge, MA: Harvard University Press; 1997. [23] Pew Research Center. Deconstructing distrust: Americans view government. Washington, D.C.: Pew Research Center; 1998. [24] Poortinga W, Pidgeon N. Exploring the dimensionality of trust in risk regulation. Risk Analysis 2003;23(5):961e72. [25] Kahan D, Braman D, Gastil J, Slovic P, Mertz CK. Culture and identityprotective cognition: explaining the white male effect in risk perception. Journal of Empirical Legal Studies 2007;4(3):465e505. [26] Costa-Font J, Mossialos E, Rudisill C. Optimism and the perceptions of new risks. Journal of Risk Research 2009;12(1):27e41. [27] Whitfield SC, Rosa Ea, Dan A, Dietz T. The future of nuclear power: value orientations and risk perception. Risk Analysis 2009;29(3):425e37. [28] Stern P. Toward a coherent theory of environmentally significant behavior. Journal of Social Issues 2000;56(3):407e24. [29] Earle T, Cvetkovich G. Social trust, towards a cosmopolitan society. London: Praeger; 1995. [30] Baxter R. Some public attitudes about health and the environment. Environmental Health Perspectives 1990;86:261e9. [31] Gillespie MUS. Public worries about toxic waste, air and water pollution as key environmental threats, http://gallup.come/poll/releases/pr990325.asp; 1999 [accessed 19.5.03]. [32] Baldassare M, Katz C. The personal threat of environmental problems as predictor of environmental practices. Environment & Behavior 1992;24 (5):602e16.

140

M. Greenberg, H. Truelove / Socio-Economic Planning Sciences 44 (2010) 130e140

[33] Bakst B. Effort to scrap anti-nuclear law in Minnesota ramps up, news/ yahoocom/ap/s/20091124/ap_on_bi_ge/mn_minnestoa_nuclear_power.html; 2009 [accessed 25.11.09]. [34] Brett P. The dilemma of aging nuclear plants, www.nytimes.com/2009/10/20/ business/global/20renuke.html; 2009 [accessed 20.10.09]. [35] Horrie C. From the birth of the daily mirror to the death of the tabloid nation. Andre Deutsch; 2003. [36] Greensdale R. How newspapers make profits from propaganda. Macmillan; 2003. [37] Greenberg M. NIMBY, CLAMP and the Location of new nuclear-related facilities: U.S. national and eleven site-specific surveys. Risk Analysis, An International Journal 2009;29(9):1242e54. [38] Greenberg M. Energy sources, public policy, and public preferences: analysis of US national and site-specific data. Energy Policy 2009;37:3242e9. [39] Zukin C. The future is here! Where are we now? And how do we get there? Public Opinion Quarterly 2006;70(3):426e42. [40] Cantor J, Brownlee S, Zukin C, Boyle J. Implications of the growing use of wireless telephones for health care opinion polls. Health Services Research 2009;44(5):1762e72. [41] Wald M. Leaks trouble nominees for nuclear panel. New York Times, http:// nytim.ms/ammxN0; February 10, 2010 [accessed 11.2.10]. [42] Ling K. What does $36B in nuclear loan guarantees buy Democrats? Greenwire. http://nytim.ms/diJoYl. [accessed 11.2.10]. [43] Learn S. Despite billions spent, Hanford won’t be clean for thousands of years. Portland Oregonian. http://bit.ly/9Xijnl. [accessed 11.2.10]. [44] US Environmental Protection Agency. Fact sheet: Coal combustion residues (CCR) e surface impoundments with high hazard potential ratings. EPA 530-F-09-006, www.epa.gov/epawaste/nonhaz/industrial/special/fossil/ccrsfa/index.htm; 2009 [accessed 29.6.09]. [45] Ward K. A year after Tennessee disaster, fight over coal-ash rules just beginning, http://wvgazette.com/News/200912190281; 2009 [accessed 21.12.09]. [46] Bisconti Research, Inc. National survey of nuclear power plant communities. for Nuclear Energy Institute, JulyeAugust, www.nei.org/newsandevents/ newsreleases/nuclearpowerplantneighborsaccept.html; 2007 [accessed 26.11.07].

[47] Bisconti Research, Inc. U.S. public opinion about nuclear energy. Report for nuclear energy institute, Washington, DC: May 5e9, 2005. [48] Gertner J. Atomic balm? New York Times; July 16, 2006. section six, 36e47, 56, 62, 64.

Michael Greenberg is Professor and Director of the National Center for Neighborhood and Brownfields Redevelopment of Rutgers University; Director of the U.S. DHS-funded National Transportation Security Center of Excellence at Rutgers University; and Associate Dean of the faculty of the Edward J. Bloustein School of Planning and Public Policy. Professor Greenberg received his B.A. from Hunter College of CUNY, and his M.A. and Ph.D. in geography from Columbia University. His recent books include Environmental Policy Analysis and Practice (2008) and The Reporter’s Handbook on Nuclear Materials, Energy, and Waste Management (2009) He has received awards for research from the United States Environmental Protection Agency, the Society for Professional Journalists, the American Public Health Association, the Association of American Geographers, and Society for Risk Analysis. He serves as Associate Editor for environmental health for the American Journal of Public Health, is a member of the Editorial Board of Socio-Economic Planning Sciences, and is Editor in-Chief of Risk Analysis: An International Journal. Professor Greenberg’s extensive refereed research has appeared in a wide variety scholarly journals.

Heather Barnes Truelove is a Postdoctoral Fellow at Vanderbilt University, Nashville, TN, in the Vanderbilt Institute for Energy and Environment, and the Consortium for Risk Evaluation with Stakeholder Participation. Dr. Truelove received her B.S. in psychology from University of Florida, Gainesville, her M.A. in general psychology from University of North Florida, Jacksonsville, and her Ph.D. in experimental psychology from Washington State University, Pullman. Her research interests center on the social psychology of pro-environmental behavior with specific focuses on attitudes toward nuclear energy and behaviors that mitigate climate change. She has presented her work at numerous conferences, with refereed papers in the Journal of Environmental Psychology and Environment and Behavior.