A Longitudinal Study of Newspaper and Wire Service Coverage of Nanotechnology Risks

1.1. Public Trust, Risk, and the Media

Consumer trust is important for the acceptance of nanotechnology products, particularly those related to health and food applications, but such trust can rise or fall depending on how well information about health and environmental risks is gathered and disseminated. According to Siegrist et al., the “importance of trust for the positive perception of new technologies suggests that a preventable event with significant negative consequences must be avoided” because it would indicate lack of concern for public welfare and decrease acceptance of nanotechnology.⁽¹⁸⁾

Many risk perception studies have shown that lack of trust is a critical factor that drives numerous technological controversies, including nuclear issues.^{(19, 20)} Such studies also have shown that social and psychological issues and values, which may not be part of a risk assessment, affect public responses to risk. These values include uncertainty, perceived inequality of the distribution of risks and benefits, involuntary exposures, lack of control over the risk, affect or emotional responses, gender, race, and worldview,⁽²¹⁾ and religiosity.⁽²²⁾ Scheufele and Lewenstein⁽²³⁾ noted that various heuristics and biases might create concerns as the public learns more about nanotechnology. Kahan et al. found through an experimental study that cultural worldviews appeared to influence perceptions of the risks and benefits of nanotechnology. For example, people who did not favor technology were less likely to be exposed to nanotechnology information, but when they were exposed to it, they were more likely to react negatively.⁽²⁴⁾

When most people have little experience with a new technology, newspaper coverage can provide heuristics for understanding and assessment, according to Scheufele and Lewenstein.⁽²³⁾ Many studies have shown that the mass media are important in conveying science and technology information to laypeople.^(25,26) Brossard et al. found that science media use plays an important role in shaping positive attitudes toward a technology in its early stages of development.⁽²²⁾ Conversely, media information can play an important role in signaling and amplifying risks, having an impact on how risks are constructed, communicated, and transformed, according to the social amplification of risk framework (SARF).⁽²⁷⁾

1.2. Nanotechnology Media Content, Coverage Patterns, and Salience

Numerous content analyses of U.S. and U.K. newspaper articles have found a focus on the positive contributions that nanotechnology can make in the economy, medicine, new materials, and information technologies, among other fields.^(28-33) Content analyses of newspaper articles in Germany⁽³⁴⁾ and the Netherlands⁽³⁵⁾ also found news about nanotechnology benefits to be the dominant coverage pattern.

Although most researchers described this positive framing of nanotechnology in terms of social progress and economic development or competitiveness, Weaver et al. portrayed nanotechnology frames by using a different set of issue frames in their study of 10 U.S. newspapers and the Associated Press (AP) wire service from 1999 to 2008. Their frames included progress, regulation, conflict, and generic risks, and they found that nanotechnology framing became more diverse over time. During 2005–2006, they saw a brief surge in articles in the regulation frame. However, journalists fell back on the progress frame in the absence of conflict or debate.⁽³³⁾

Pidgeon et al. also found in a series of deliberative workshops in 2009 that “benefit framing currently dominates understandings of the future of nanotechnology in both the United States and United Kingdom and persists even when participants are provided with the opportunity for balanced engagement with a range of information and perspectives regarding potential risks.” The researchers noted that it was not clear whether this positive perception would endure or change if a significant nanotechnology health, environmental, or safety issue were to occur. They noted that: “Any such event, if significantly amplified through media coverage, would likely provide the (currently missing) ‘mental model’ or narrative allowing people to connect nanotechnology risks in more concrete terms to their everyday lives.”⁽³⁶⁾

Also finding a strong benefits emphasis in a study of U.K. nanotechnology news coverage from 2003 to 2004, Anderson et al. found relatively little coverage of the risks, implications, and uncertainties of nanotechnologies.⁽³⁷⁾ Nanotechnology safety issues received little sustained media attention, appearing in only 9% of the coverage and mostly in the elite or “quality” newspapers in the sample.⁽³⁸⁾ Our earlier work on nanotechnology risk coverage in U.S. and U.K. newspapers and wire services from 2000 to 2004 also showed little coverage of risks issues.⁽³⁰⁾

This research adds to our earlier work by looking for changes in coverage patterns and narratives over a 10-year period from 2000 to 2009. As with the earlier study, we were particularly interested in the possibility that the news media would treat potential nanotechnology health and environmental risks as a salient topic and draw readers’ attention to it.

Often, people learn how much importance to attach to a topic by the emphasis placed on it in the news, according to agenda setting theory. Basically, agenda setting refers to the idea that there is a strong correlation between the emphasis that the mass media place on certain issues and the importance attributed to these issues by the public.⁽³⁹⁾ Much mass communication research has been done on various interpretations of the media's role in agenda setting, including agenda setting not only for public issues or topics but also for attributes of these issues or topics—that is, the characteristics and traits that describe the issue or topic. According to McCombs: “The power of the news media to set a nation's agenda, to focus public attention on a few key public issues, is an immense and well-documented influence.” Having cues repeated “day after day effectively communicate[s] the importance of each topic.”⁽⁴⁰⁾ Some cues that indicate the importance or salience of a topic include the volume of the coverage, position of the story, and its length, as well as the prominence of the news outlets that carry the story.⁽³⁹⁾

From a salience and SARF perspective, extensive and prominent media coverage of threats of long-term illness or environmental contamination from nanotechnology products or manufacturing could draw public attention to a variety of concerns about nanotechnology. In fact, they could be prime examples of the types of risks that media coverage would amplify, which could then have ripple effects on companies, industries, and other nanotechnologies.⁽²⁷⁾ Concerns about such possibilities have been expressed for some time by U.S. officials associated with the National Nanotechnology Initiative and other government agencies.⁽⁴¹⁾

To evaluate the impact of U.S. and U.K. newspaper and wire service coverage of nanotechnology risks, this descriptive study tracked a number of issues related to three basic research questions.

Among the issues tracked to answer these questions were:

2.1. Study Sample

For this 10-year study, 20 U.S. and 9 U.K. newspapers and the Associated Press (AP) and United Press International (UPI) wire services were searched from 2000 through 2009. It was especially important to conduct this study over a long period of time because longitudinal studies of media coverage provide repeated observations that can distinguish short-term from long-term effects. Having a variety of newspapers from both countries and the two wire services allowed for a wide geographic and demographic distribution in two countries with industrial, scientific, and governmental interest in nanotechnology. The newspapers used encompassed all of those from the United States and the United Kingdom that were included in the “Major Newspapers” section of the LexisNexis Academic database as it existed in 2004. Adjustments to retain continuity were made to match changes made to this database by LexisNexis during the later years of the study. The terms used for the LexisNexis searches were “nanotechnology,”“risk or problem or issue or concern or toxicity or safety,” and “environment or health.” The Lexis “wildcard” designation was used for each term to allow for a variety of word endings in the search.

The main U.S. news sources for the study in descending order of the number of articles found in each were the UPI wire service, The New York Times, The Washington Post, AP wire service, The San Francisco Chronicle, The Houston Chronicle, The Boston Globe, and The Christian Science Monitor. Other U.S. newspapers included the San Diego Union-Tribune, The Sacramento Bee, USA Today, Chicago Sun-Times, The Oregonian, St. Louis Post-Dispatch, The Milwaukee Journal Sentinel, Tampa Tribune, Pittsburgh Post-Gazette, St. Petersburg Times, The Atlanta Journal-Constitution, The Columbus Dispatch, San Antonio Express-News, and The Denver Post.

The U.K. newspapers listed in descending order of the number of articles found in each were The Guardian, The Financial Times, The Times and The Sunday Times, The Independent and The Independent on Sunday, The Irish Times,³ The Daily Telegraph and The Sunday Telegraph, The Scotsman and Scotland on Sunday, The Herald, and The Observer. Articles from the U.K. Sunday newspapers, excepting The Observer, were counted as part of the total contributions with their daily counterparts to maintain comparability with the Lexis listings of U.S. newspapers, which do not have separate names or listings for daily and Sunday publications.

Using the U.K. newspapers included in Lexis’ major newspapers section during 2004 emphasized elite newspapers and did not include mid-market or tabloid newspapers. This selection was supported by the 2003 to 2004 study by Anderson et al., which sampled nanotechnology articles from 10 daily and 8 Sunday U.K. newspapers and found that most articles that dealt with nanotechnology risks were in the elite newspapers such as The Guardian, The Times, The Financial Times, and The Telegraph. The Observer was the only Sunday newspaper to include an article that discussed nanoparticle risks.^(37,38)

About 2,200 articles were retrieved from Lexis using the search terms between January 1, 2000 and December 31, 2009. The senior author screened these articles to determine whether they specifically included information on a nanotechnology health or environmental risk or whether the search terms appeared randomly in articles on irrelevant topics. This filtering process determined that 367 of the articles included information about health or environmental risks. The selected articles included news and feature articles and columns, but omitted editorials or letters to the editor.

The low number of articles that result from a filtering review is often common with Lexis as these searches provide any article within the search domains that includes the search terms. A similar filtering activity in the study by Weaver et al. reviewed 1,763 articles and found only 137 relevant for that study.⁽³³⁾

2.2. Coding and Intercoder Reliabilities

At the outset of the study, a detailed 11-page coding document containing 49 questions was developed and tested to provide answers to the study's research questions for articles from 2000 to 2004. The document provided quantitative coding categories for information in four major segments: (1) general newspaper and reporter data; (2) health, environmental, and societal risk data; (3) public attitude coverage data; and (4) source data. A brief fifth segment included several questions on clarity and depth of the articles, reader relevance, and valence judgments about the headline, individual paragraphs, and the overall article.

In 2005, a new 19-page section containing 79 questions was added to the coding document to reflect increasing media coverage of nanotechnology regulation and funding. There were three segments in the regulation section. The first focused on the people or groups calling for new or tightened regulations about nanotechnology risks; the second section concerned regulatory strategies, plans, or proposals from nongovernmental persons, groups, or organizations; and the third concerned actions, regulations, or legislation from elected or government officials or government bodies. This third section was divided into three subsections, preliminary, proposed, or passed/put into effect. A section on funding concentrated on discussions about the need for more funds for environmental, health, or safety studies.

The unit of analysis was the individual article. The majority of the codes were dichotomous, indicating presence or absence of an item. There were a few open-ended questions to provide more detailed information for the dichotomous questions, and very few scaled questions. Most of these related to value judgments, which will be discussed below. Codes for specific health, environmental, and societal risks allowed for multiple positive responses in different categories. Most of the research questions required simple descriptive answers, and frequencies, cross-tabulations, and Pearson's chi-square tests were calculated using the Statistical Package for the Social Sciences. Chi-square tests were used to calculate statistical differences between U.S. and U.K. data and the results appear in the text when they are statistically significant beyond the 0.05 level.

A coding instruction booklet was developed in 2004 that gave directions on how to annotate articles for coding, how to code the information in each of the original five sections, and how to define terms and concepts. The coding booklet was updated yearly and used for multiple training sessions with three coders.

Intercoder reliability scores averaged 86% for articles from 2000 to 2004. This was calculated by having a proportion of articles double coded and then checking for exact agreement on each question. At that time, some problems occurred in a few of the scaled questions that required value judgments about the valence of the articles and paragraphs, which lowered the overall reliability score. Because of the complexity of the subject matter and the value judgments required, these scaled questions were dropped from the study and the coding document in 2005 and were not reported thereafter.

In 2005, additions were made to the coding instruction booklet on how to code the complex regulatory information. Definitions of what constituted regulation were developed, including word cognates such as ban, oversight, guideline, and so on. Multiple training sessions and tests were done. To ensure accuracy, the senior author identified articles that had regulation sections in them in advance of coding. All the articles from 2005 to 2009 were coded using the 30-page document. In addition, articles from 2000 to 2004 were recoded to include the new regulation and funding sections. Consequently, the contents of all 367 articles were analyzed using the 30-page coding document.

To ensure high reliability because of the increasing complexity of nanotechnology discussions, all articles from 2005 to 2009 plus the regulation and funding sections of articles from 2000 to 2004 were each coded by two individuals. After comparing the responses from the separate coders for each article, the senior author arbitrated any discrepancies. Intercoder reliabilities were 95.8% from 2005 to 2009 and they were calculated in the same manner as the earlier articles.

3.1. Salience Indicators and Event Orientation

As will be seen from several indicators of salience, including volume of coverage, prominence of news outlets, article placement in the publications, and their length,^(39,40) the nanotechnology risk story was not an important media topic from 2000 to 2009, although there were a few instances when a particular risk topic or event drew modest media coverage. One major factor that exerted its influence throughout the study was the volume of coverage. Only 367 articles about nanotechnology health or environmental risks met the research criteria—248 from the United States and 119 from the United Kingdom. This is 5.9% of the more than 6,200 articles found during a very general search in Lexis using only the term “nanotechnology” in the same newspapers and timeframe.⁴

Although the total number of articles was small, an increasing number appeared in the United States from 2001 to 2006, followed by a decline in 2007, a similar level in 2008, and a drop in articles by 50% in 2009 (Fig. 1). At first it appeared as if 2006 indicated a new level of U.S. interest in nanotechnology risk coverage, but it was an outlier year. In the United Kingdom, after a large increase in 2003, only coverage in 2004 and 2008 showed moderate increases.

Prominent newspapers published these articles but not often enough to indicate the importance of the subject to readers. Over a 10-year period, 33 nanotechnology risk articles appeared in The New York Times, 31 in The Washington Post, 29 in The Guardian, and 22 in The Financial Times. The largest number of articles from one source was 78 from the UPI wire service, an organization associated with a smaller number of media outlets than the larger and more powerful AP, which published 28 articles. Even though some prestigious newspapers carried these articles, they were not prominently displayed. Only 4.1% appeared on the first page of any newspaper, and 26.7% appeared elsewhere in the first section of the newspaper.

Many of the early U.S. articles were of good length, with a mean number of words per U.S. article of 1,188.4 words in 2004. However, this mean dropped to 553.7 words in 2009.

U.K. articles had a mean of 899.3 words per article in 2004, which rose slightly in the next two years and then dropped in 2007 and 2008. The mean rose again in 2009 to 988.3 words, which was almost double the length of the U.S. articles that year.

Another factor that indicated little salience was a lack of consistency in the various nanotechnology risk topics covered by the media. Although this study identified a general but relatively small focus on health and environmental risks as well as a few clearly defined issues, the majority of the articles in both countries focused on discrete news events. Even with defined issues such as nanosilver or sunscreen concerns, there was little continuity over time so readers could follow changes occurring in a particular topic. Much of this was because the coverage was predominately oriented toward covering news events. About 68.5% of the U.S. and 58% of the U.K articles were based on the release of a study or report, an event, or statements by well-known people. So when an important report was released such as one from the Royal Society and Royal Academy of Engineering in July 2004 or when an important person such as Prince Charles made a comment, there was an increase in coverage. Without such events, the number of articles dropped. Reinforcing this point, 69.5% of the articles in both countries were news articles, which usually provide less information to help readers understand technical issues than feature, investigative, or explanatory articles.

3.2. Narratives and Patterns of Risk Coverage

During the time period of this study, risk coverage involved three main narratives. The earliest was one that focused on runaway technology, which included coverage of concerns about science-fiction-type issues such as nanobots and grey goo. It ran from about 2000 to 2004. A science-based narrative began around 2004, increased considerably for several years and, although declining, still was important at the end of this study. It focused on coverage of scientific studies and reports about general and specific health, environmental, and societal issues. Finally, a regulation narrative started with coverage in the United Kingdom in 2004 and in the United States from 2005, with a drop off in 2007 to 2009, although it too was present at the end of the study. Other studies have found similar patterns in both U.S. and U.K. coverage.^{(33,37,38,42)}

3.2.1. The Runaway Technology Narrative

In the runaway technology narrative, risk coverage reflected issues raised by Eric Drexler in his book Engines of Creation, Bill Joy in his article in Wired, Michael Crichton in his novel Prey, and by Prince Charles in statements and letters to newspapers that self-replicating nanobots could run amok and consume all natural matter on the planet, leaving behind only grey goo.⁽⁴³⁾ Coverage of these early risk issues appeared in 28.6% or 105 of the total articles. U.K. coverage was significantly higher, the issues appears in 49.6% of these articles versus 18.5% in the United States (p= 0.001). The number of U.K. articles on runaway technology was highest in 2003 and 2004 spurred by comments against nanotechnology by Prince Charles and the Royal Society report and then dropped off (Fig. 2). The major topics covered were grey goo, discussed in 41.2% of the U.K. and 9.3% of the U.S. articles, and self-replicating nanobots, discussed in 39.5% of the U.K. and 16.5% of the U.S. articles.

The runaway technology narrative was most prevalent in the United Kingdom, particularly because of Prince Charles’s prominent role, which received both support and derision in the media. Anderson et al. found similar results, reporting that Prince Charles’s comments on grey goo appeared in 39 of 344 articles they reviewed. These investigators felt that coverage of the Prince's comments took on a more “celebrity” tone with both political and humorous tones.⁽³⁷⁾

3.2.3. Health Risk Coverage

Health risk coverage (73.6%) was more prominent in the total number of articles for both countries than environmental (67%) and societal (66.5%) risk coverage. Of the U.S. articles, 185 or 74.6% included discussion of health risks, and of the U.K. articles, 85 or 71.4% did so. More than 57% of the health risk articles in both countries appeared between 2006 and 2009 (Fig. 3). Even in 2009, with its paucity of articles, health risks appeared in 82.4% of the U.S. and 62.5% of the U.K. articles.

Most of the health risk discussions concerned nonspecified risks instead of or in addition to a more specific risk: 55.6% in United States and 50.4% in the United Kingdom (Fig. 4). Over time, however, more specific health risk references occurred, probably as a result of an increased number of scientific studies published after 2003. For example, coverage of nanoparticle risks for lungs and brains occurred after scientific studies showed damage caused by accumulation of nanoparticles in rodents and fish. Concerns about using nanoparticles in cosmetics and sunscreen skin products began appearing more frequently in 2004. There were no statistically significant differences between the U.S. and U.K. coverage categories.

3.2.4. Environmental Risk Coverage

Of the 367 risk articles, 67% or 246 were environmental risk articles. Environmental risk articles appeared in 70.2% of the U.S. and 60.5% of the U.K. articles (Fig. 5). Interest in environmental risks in the United States remained fairly consistent between 2004 and 2008. Even in 2009, when only 10 U.S. environmental articles appeared, this accounted for 58.8% of the coverage. The U.K. media did not exhibit as much interest in the topic except in 2003, 2004, and 2008, and the number of articles never surpassed 16 in any year. Similar to the health risk articles, most of the environmental risk articles focused on general or nonspecified risks: 59.7% in the United States and 52.1% in the United Kingdom (Fig. 6).

Differences in U.S. and U.K. coverage of specific environmental risks were statistically significant for five categories involving the threat of nanoparticles to: contaminate water sources (p= 0.016), damage viruses, bacteria, or other microorganisms (p= 0.009), damage plants, species, or wildlife (p= 0.006), contaminate soil or landfills (p= 0.003), and harm sewage treatment plants with nanosilver (p= 0.001). Most of the differences in environmental coverage in the two countries were probably because the majority of the environmental studies reported occurred in the United States and received more coverage there.

3.2.5. Societal Risk Coverage

Although this study sought articles involving health and environmental risks, about 66.5% or 244 of the total articles also included information on various societal risks, even though societal risks were not included in the search terms. Given the presence of societal risks in so many articles that included health and environmental risks, it was decided to track these societal risks over time in the belief that the three risk topics tended to aggregate in newspaper coverage. Consequently, all of the articles discussing nanotechnology societal risks that may have appeared in the newspapers and wire services studied may not be included here.

U.K. articles included more information about societal risks, 77.3% compared to 61.3% of the U.S. articles (p= 0.002). In the United States, the two highest years were 2005 and 2006, with almost steady numbers for 2003, 2004, 2007, and 2008. A large drop occurred in 2009. In the United Kingdom, the same three high years appeared as in the environmental risk category: 2003, 2004, and 2008 (Fig. 7).

Concern about safety issues was the largest category in both countries—appearing in 41.5% of the U.S. and 49.6% of the U.K. articles (Fig. 8). Safety was a general term used by the journalists that they did not specifically link to a health or environmental risk. The next major area of concern for both countries was the side effects of runaway technology, and as noted previously, there was significantly more discussion of this risk in 34.5% of the U.K. articles compared to 15.3% of the U.S. articles (p= 0.001). U.K. newspapers also paid significantly more attention to social issues than did U.S. newspapers (p= 0.006).

It may be that the larger concerns for safety and social issues found in U.K. newspapers are a result of past experience and extensive media coverage of similar issues involved in mad cow disease, genetically modified foods, and stem cell research. In particular, with genetically modified foods, food safety and labeling have been of far less concern in the United States than in the United Kingdom or Europe.⁽⁴⁴⁾

3.2.6. The Regulation Narrative

Articles that fell into the regulation narrative started appearing about the same time as those in the science-based narrative, but grew at a different rate. Although fewer articles discussed nanotechnology regulation than health and environmental risks—43.6% of the 367 articles—more articles appeared later in the study period. In the United States, 109 articles mentioned regulation issues with the major year being 2006 when they composed 65% of the U.S. total. From 2007 through 2009, regulation was discussed in 50%, 41%, and 47% of the U.S. articles, respectively, even though the total number of U.S. articles dropped each year. In the United Kingdom, 51 articles included discussion of regulation, with it occurring primarily in 2004 when 56% of that year's U.K. coverage was involved. In 2008, 50% of the U.K. coverage concerned regulation issues (Fig. 9). Differences between the coverage in the two countries were not statistically significant.

Three-quarters of the regulation articles focused on calls for new or tightened regulations, in 76 U.S. and 40 U.K. articles. Again, the difference was not statistically significant. The two leading reasons for these calls in both countries were to protect public health and safety and to protect the environment. The third leading reason, particularly in the United States, was that existing rules and regulations did not apply to nanotechnology.

A small group of environmental organizations called most frequently for new and tightened regulations in the U.S. coverage, including the ETC Group and the International Center for Technology Assessment, followed by the Project on Emerging Nanotechnologies (PEN) of the Woodrow Wilson International Center for Scholars. Scientific organizations, including both the Royal Society and the Royal Commission on Environmental Pollution, as well as individual scientists, were most often identified as those seeking more or better regulation in the U.K. coverage.

3.3. Top Risk Events Covered

Coverage of specific events often alerts readers about risk concerns and contributes to risk amplification, as postulated by the SARF. Ten events were identified as generating small clusters of newspaper articles between 2000 and 2009. These were, in chronological order:

This list was developed by looking for topic groupings in either or both countries that occurred at a particular time. A study by Weaver and Bimber using Goggle News data in major U.S. newspapers and the AP from January 2006 through August 2007 identified spikes in coverage for many of the same events during that time period.⁽⁴⁵⁾ This type of concentrated coverage, however, was the exception rather than the rule. In general, as described earlier, most topics or events, particularly in the United States, were reported by only one newspaper or wire service at a time, so there were many topics, such as a scientific study, that appeared in only one article. Because of the small number of articles involved in each, probably none of the 10 risk events became salient or acted as a risk amplifier for newspaper readers. However, these topics were discussed on many websites and may have amplified concerns and alerted some Web readers.

One incident that did have the potential for amplification was a major health risk incident in August 2009 in which two Chinese factory workers died and five developed severe health problems from working in a paint factory using nanoparticles.⁽⁴⁶⁾ Yet this incident received no coverage in the U.S. newspapers studied or even in almost all of the major U.S. newspapers indexed in Lexis in 2009. It was covered by The Telegraph in the United Kingdom, the Reuters wire service, and a large number of websites. One can speculate that the lack of U.S. newspaper coverage was probably because the incident was work-related, in China, and involved a technology that was not known to many people. Therefore, U.S. editors and reporters did not see it as newsworthy. Clearly, some U.K. news outlets found this incident important to cover, but even there, the incident quickly faded from sight.

3.4. Scientific Uncertainty and Cautions

The nanotechnology scientific literature is full of statements about the uncertainty of the science, and this was reflected in about half of the coverage in both countries. Uncertainty statements were included in 52.4% of the U.S. and 55.5% of the U.K. risk articles. The most commonly used uncertainty explanations fell into a category described as “little is known about the risks or hazards of nanotechnology,” which was found in 25% of the U.S. and 30.3% of the U.K. articles. The next most frequent category was “there are uncertainties about the risks of nanoparticles, nanotubes or nanotechnology,” found in 14.1% of the U.S. and 5.9% of the U.K. articles (p= 0.02). A statement similar to “more research or studies are needed before health or environmental issues or their potential toxicity can be addressed” was found in 45.6% of the U.S. and 47.1% of the U.K. articles.

In contrast to the fairly frequent uncertainty coverage, little attention was given to cautions about proceeding too quickly with nanotechnology research and manufacturing. Several organizations, including the Royal Society and the ETC Group, had called for a moratorium on nanomanufacturing or distribution until more was known. However, moratorium information was reported in only about 11% of the articles in each country. Even fewer articles mentioned invoking the precautionary principle for nanotechnology, 7.3% in the United States and 6.7% in the United Kingdom.

3.5. Information Sources

The credibility of information sources and the levels of trust accorded them play a role in the impact their statements have on a risk issue. People and groups who were information sources for journalists in this study were categorized by whether they raised concerns about nanotechnology risks or whether they responded to risk issues by discussing information that would reassure readers that nanotechnology risks were not serious or could be managed. Sources could be counted in both the raised and responded categories if they did that in the same article, which happened somewhat frequently.

There were 623 risk-raising sources in 367 articles, an average of about 1.7 per article. University scientists and engineers were the leading risk raisers, not environmental or consumer organizations as some members of the nanotechnology community had expected. They were followed by the Wilson Center's PEN and other nanotechnology organizations. As would be expected, a lower number of sources who responded to risk issues appeared in the study, 436 in 367 articles, averaging about 1.2 per article. University scientists and engineers also led this source group, followed by industry and government organizations. Almost all of the main risk raisers probably would be considered trustworthy by the public because of their roles as university scientists and engineers.⁽²⁶⁾