The political climate for science communication has changed even in the 4 years since the second Arthur M. Sackler Colloquium on the Science of Science Communication. People are more likely to get scientific information from social media or politically partisan news sources. Political polarization has made some people less likely to trust information from
particular news outlets. Scientists are increasingly likely to communicate directly with lay audiences through social media, sometimes before peer review of their research.
Several speakers at the colloquium explored the challenges facing science communication in a politically charged environment, including the challenges that arise in communicating information about immigration, climate change, and vaccination.
Americans’ sense of political identity has resulted in an increasingly powerful form of ingroup-outgroup polarization, said Shanto Iyengar, Chandler Chair in Communication at Stanford University. Surveys by the American National Election Studies, The Pew Charitable Trusts, and other organizations document increasing hostility toward the outparty, along with increased social distance depending on party affiliation. Other results suggest that implicit bias against the outparty exceeds the levels of implicit bias based on race (Iyengar and Westwood, 2015). Behavioral measures of interpersonal trust demonstrate that the party cleavage is deeper than social cleavages even in divided societies.
Social identity theory holds that humans tend to gravitate to groups and that once they form a group they develop hostile feelings toward outgroups, Iyengar pointed out. In the political domain, researchers have coined the term sorting to indicate that the political divide between Democrats and Republicans is now being amplified through reinforcing cleavages. This sorting—where a partisan divide converges with racial, gender, age-based, religious, and urban–rural divides—contributes to a situation in which partisan opponents are considered what Iyengar termed “the repugnant cultural other.” As an example of this effect, disapproval of interparty marriage has grown dramatically in the United States, though not in the United Kingdom, since the 1960s (Iyengar et al., 2012). Partisanship has become a litmus test for interpersonal relationships, including those based on family, kinship, friendship, and online networks. As a result, “groups are increasingly politically homogeneous,” Iyengar pointed out.
This polarization has had a profound effect on media in the United States. Many news organizations cater to partisan preferences. Partisans, especially on the right, impute bias to mainstream news—for example, a recent poll found that 46 percent of respondents think that the media make up stories about President Trump. Partisan news sources have seen increased circulation, creating the potential for an “echo chamber,” where people mostly receive information that reinforces their preferences.
When partisans encounter information at odds with their identity, they counterargue and often move further away from the evidence-based
positions, said Iyengar. The credibility of universities, think tanks, and government agencies is considered suspect. When a political party has developed a reputation or a stance on a particular policy issue, the adherents of that party tend to follow the cues supplied by party leaders.
Provision of expert information can change the preferences of policy makers, Iyengar observed, though sometimes in counterintuitive ways. For example, when researchers presented information to policy makers showing that needle exchange programs do not increase drug use, liberal policy makers’ support for these programs remained close to 100 percent before and after the treatment. But conservative policy makers tended to disagree with the programs even more after they received the information. “If anything, the suggestion from this study is that expertise can have a boomerang effect,” he said.
Iyengar concluded by proposing several ways of dampening the polarization. Gentler rhetoric from high places would help, but “I would not be losing a lot of sleep over that possibility.” Increased interpersonal contact across the party divide could reduce animus; for example, in marriages that mix Democrats and Republicans, evaluations of the 2016 presidential candidates tended to be less extreme. Increasing the mobilization of nonpartisans would be helpful but not easy. Finally, with topics such as immigration, Iyengar pointed out that framing the issue in terms of individuals rather than policy can increase public support for a position.
Doug Massey, Henry G. Bryant Professor of Sociology and Public Affairs at the Woodrow Wilson School of Public and International Affairs at Princeton University, has been engaged with public communication on immigration issues since he got his Ph.D. in 1978 from Princeton. He has penned op-ed articles, written and edited books for general readers, appeared on radio and television programs, and testified in front of the U.S. Congress. “I’ve gotten better at communicating,” he said. “But in my experience, it’s not the communication that has become the problem. It’s the context into which you’re trying to communicate, which has changed dramatically.”
Massey served on the committee that produced the report The Integration of Immigrants into American Society (NASEM, 2015) and he was a reviewer of the report The Economic and Fiscal Consequences of Immigration (NASEM, 2017d). The messages of both reports were well received by the mainstream press, he said. The headline for the latter report in The New York Times was “Immigrants Aren’t Taking Americans’ Jobs, New Study Finds.” The headline in The Wall Street Journal was “Immigration Does More Good than Harm to Economy, Study Finds.” However, other
organizations cast the report’s findings in a very different light. The restrictionist Center for Immigration Studies described the report under the headline “National Academy of Sciences Study of Immigration: Workers and Taxpayers Lose, Businesses Benefit.” The Heritage Foundation’s headline was “National Academy of Sciences Report Indicates Amnesty for Unlawful Immigrants Would Cost Trillions of Dollars.” The Washington Examiner wrote “Immigration costs at least 278 billion, ‘illegals’ mentioned only three times.” Breitbart’s headline was “National Academies’ Study Shows $500 Billion Immigration Tax on Working Americans.” Infowars said, “Report Explains Financial Cost of Illegal Immigrants for American Taxpayers,” adding that the “wall could save taxpayers $64 million over the next decade.”
These diametrically opposed interpretations point to a major structural shift in the media and in America, Massey said. Before the U.S. Congress deregulated the cable industry in 1984, there were five major sources of broadcast news: ABC, NBC, CBS, PBS, and CNN. After deregulation, these five were joined by Fox in 1986, CNBC in 1989, Bloomberg in 1990, FSTV in 1995, MSNBC in 1996, Blaze in 2011, One America in 2013, and CBSN and Newsmax in 2014. Before 1987, the Fairness Doctrine required broadcasters to devote some of their airtime to discussing controversial matters of public interest and to present contrasting views regarding those matters, and controversial issues had to be presented in a manner that was, in the view of the Federal Communications Commission, honest, equitable, and balanced. In 1987 the commission eliminated the doctrine.
Following the spread of the Internet in the 1990s, social media became a major new source of information and Internet news sites became alternatives to the mainstream press. Internet trolls began to sow discord, disbelief, and discontent on the Internet by starting arguments or upsetting people, by posting inflammatory, extraneous, or off-topic messages in online communities, and by otherwise trying to undermine facts and knowledge. In 2010 the U.S. Supreme Court ruled that freedom of speech prohibited the government from restricting independent political expenditures by profit or nonprofit corporations, which unleashed a wave of dark money into politics.
“Unfortunately, science and support for science has become a partisan issue,” concluded Massey. While trust in science has remained largely unchanged among liberals and moderates since about 1980, it has declined markedly for conservatives. The result has been an increase in obfuscation and misdirection. Communicators therefore need to use all the tools available to them to support their messages, he insisted. “It’s not enough simply to communicate clearly and effectively. You have to have a counterinsurgency tactic ready to fight the inevitable information wars.”
A long-term collaboration between M. Granger Morgan, Hamerschlag University Professor of Engineering at Carnegie Mellon University, and Wändi Bruine de Bruin, university leadership chair in behavioral decision making at the Leeds University Business School, has produced several key insights and lessons learned in communicating about climate change research.
In one project, Morgan and Bruine de Bruin studied messages about removing carbon dioxide from power plant emissions and putting it deep underground, thereby helping to curb climate change by preventing carbon dioxide from entering the atmosphere. A series of interviews in the early 2000s found that most interviewees had not heard of carbon capture and sequestration (CCS) and were, at best, lukewarm about the technology after they heard a brief description of it. Instead, they tended to want to discuss and compare alternative strategies for low-carbon electricity generation, such as solar power and wind power (Palmgren et al., 2004). Participants expressed a lower degree of acceptance for geological and oceanic CCS than for alternative low-carbon strategies, including nuclear power. “What might have happened in that survey is that people focused on the negative information about CCS more than the positive information,” said Bruine de Bruin. “We know from decision research that negative information gets more attention than positive information and losses loom larger than gains. What we should have been doing is give people information about the risks, costs, and benefits about all technologies so that people could recognize that all technologies have downsides and upsides.”
In follow-up studies with graduate student Lauren Fleishman Mayer, the investigators wrote multiple-attribute descriptions of 10 low-carbon technologies, with the descriptions pilot tested with the intended audience to ensure that they were understandable (Fleishman et al., 2010). The descriptions covered how the technology works, carbon dioxide releases, cost, and safety. After studying the materials at home, participants preferred coal plants with CCS versus coal plants without CCS, particularly with the newer types of plants. “What we concluded from the project is that, if you give people positive and negative information about a range of technologies, they may be willing to accept the technology that you’re trying to communicate about,” Bruine de Bruin observed.
Based on these and other studies, Morgan, Bruine de Bruin, and their co-workers have been developing various communication materials on climate change for lay audiences. However, Morgan has increasingly come to the conclusion that, to be informed participants in public discourse about climate change issues, people need to know just three things. First, burning fossil fuel produces carbon dioxide. Second, higher levels of
carbon dioxide in the atmosphere warm the planet and cause the climate to change. Third, once carbon dioxide gets into the atmosphere, much of it stays there for hundreds of years.
Studies that Morgan did with Ann Bostrom at the University of Washington had shown that progress has been made in increasing public understanding of the first two points. However, understanding of the third point was making much less headway. Therefore, in a collaboration with Ph.D. student Rachel Dryden, Morgan, Bruine de Bruin, and Bostrom set out to learn what people thought about how long carbon dioxide stays in the atmosphere. In a survey, they asked people how long they thought it would take for common air pollutants such as smog and for carbon dioxide to return to preindustrial levels if humans quit emitting them. The results showed that people did not understand the difference between the two—even though common air pollutants would leave the atmosphere quickly while carbon dioxide would stay there for hundreds of years (Dryden et al., 2017). “We find that disturbing,” said Morgan, “because if you believe that, then people would be able to think: ‘I don’t know if this climate change stuff is real or not. But if it ever gets serious enough, we’ll just fix it by reducing emissions in the same way that we fixed air pollution in places like Pittsburgh and Los Angeles.’ And of course that doesn’t work. Once the stuff is there, we’re stuck with it.” Current work is looking at communications that can help people understand why different pollutants have very different residence times in the atmosphere.
Morgan and Bruine de Bruin have drawn several conclusions from their study of climate change communication. First, communication is more effective if it addresses what people need and want to know, not just if it covers what experts think is important. Second, communications need to be worded in an understandable way so that people can use the information in their decisions. Third, it is important to test the effectiveness of communications before they are disseminated. “If you don’t test your communications, then you run the risk of spending time and effort in disseminating something that doesn’t help people make more informed decisions,” said Bruine de Bruin.
More broadly, Morgan and Bruine de Bruin identified a need for research that brings together domain experts who can provide technical expertise and social scientists who can provide expertise on ensuring understanding with the intended audience. Interdisciplinary collaborations can pose challenges, they acknowledged, but these challenges can also be seen as strengths. Interdisciplinary researchers may aim to solve different problems, but their work together can be more useful as a result. Theories and methods may differ, but the resulting work ends up being grounded in multiple fields. Researchers may differ in academic language and culture, but they can learn to understand one another. Interdisciplin-
ary papers may not fit into disciplinary journals, but enough top journals want interdisciplinary projects to make publication possible. Similarly, traditional academic departments may be confused by interdisciplinary research, but the departments that welcome such researchers can break new academic ground. Morgan and Bruine de Bruin have had adequate funding, supportive environments, recognition in their fields, and excellent students who have helped them balance their approaches. They have also have been fortunate to work at institutions that promote interdisciplinary work.
“Vaccines work,” stated Christine Finley, immunization program manager at the Vermont Department of Health. The incidence of rubella, mumps, hepatitis A, hepatitis B, measles, chicken pox, diphtheria, whooping cough, and polio—which together used to plague millions of Americans, mostly children—has dramatically decreased as vaccinations for these diseases have become widely available. Today, both polio and rubella no longer exist in the United States, and only 37 cases of polio worldwide were reported in 2016. Diphtheria, which once caused 150,000 deaths per year, is now extremely rare.
Despite these dramatic results, one in three parents in the United States today has some concern about vaccinating their children. This unease has a variety of sources, said Finley. One is that perceived benefit matters. Most parents have never seen a child with measles or who has been paralyzed by polio. However, many parents have encountered someone with autism and have heard that autism may be related to vaccines. Many other forces also come into play. Imposed risks feel scarier than risks that are chosen. Misinformation is widely disseminated through a variety of venues, especially social media. Distrust in pharmaceutical companies and in government causes people to rely on other sources of information. Many parents today believe that natural is better and fixate on the manmade nature of vaccines, even though the reaction to a vaccine is a natural response of a functioning immune system.
Vaccines are mandated for enrollment in child care and kindergarten in all 50 states, but 47 states still provide loopholes to escape these requirements. Only three states do not allow exemptions for philosophical or religious reasons—California, Mississippi, and West Virginia—with the exemptions being eliminated by California’s legislature in 2015 after a measles outbreak at a Disney theme park.
“Parents’ need for education and guidance in order to vaccinate with confidence is higher than ever, but we don’t know exactly what that should look like,” observed Finley. Nevertheless, a few things are known
that work, she added. It is important to understand the audience and not to confront them in a judgmental way. It also is important to focus on the fence sitters. “People who are completely opposed to vaccines will never be interested, but they’re usually less than 1 percent.” The credibility of health care providers can be leveraged. “They’re our partners in this, and they’re very well trusted among patients.” Finally, personal stories and testimonials can be very powerful.
Though many communication strategies are under way in the United States, they are often underfunded and not evaluated, Finley observed. She and Brendan Nyhan, professor in the Department of Government at Dartmouth College, have therefore formed a partnership to evaluate new approaches to promoting vaccination by determining the most effective messages and strategies in countering vaccine hesitancy. Their prospective study was one of two described at the colloquium that were chosen through a competitive process for support by the Rita Allen Foundation and the Arthur M. Sackler Colloquia.
“We should distrust our intuitions about what works and instead test them rigorously,” said Nyhan. He has led two studies designed to evaluate the effectiveness of giving corrective information to people about vaccines, and both have had counterintuitive results (Nyhan and Reifler, 2015; Nyhan et al., 2014). In randomized controlled trials with nationally representative samples, the studies found that exposure to corrective information caused people to express less intention to vaccinate rather than more. Negative responses were concentrated among the people with the least favorable attitudes toward vaccines—precisely the group of greatest concern.
Finley and Nyhan’s project aims to fill gaps in knowledge by bringing together scholars and practitioners who can exchange and combine insights and expertise. Practitioners have the unique perspective of engaging directly with human behavior. Academics, by partnering with practitioners, can analyze actual behavioral measures on vaccination rates rather than relying on surveys. Practitioners can also provide valuable local insights into communication problems that academics working at a broader level might overlook.
Despite the expertise that practitioners bring to the partnership, many are not trained in experimental and survey research methods. Academics can design, execute, and analyze research conducted in health care settings and apply the rigorous techniques needed to produce concrete evidence. They may be able to spot limitations in the measurement capabilities of practitioners and program officers, and they can contribute funding support to the partnership through institutional backing and grants. Nyhan pointed out that new programs can be randomized across a variety of
factors. “Instead of haphazardly rolling it out, let’s randomize how we roll it out, and let’s partner with someone to evaluate it and see if it works.”
These kinds of partnerships could become much more common, Nyhan observed. Scholars and practitioners are both interested in collaborating, “but helping them find each other, connecting people who have common interests and expertise, is harder than it might seem.” Research and practice can inform each other, such partnerships are not expensive, and they can be incorporated into existing programs and activities. “What’s most exciting about this partnership is the ability to have real-world impact,” said Nyhan. “We’re going to provide the most rigorous evidence to date of the effectiveness of vaccine messaging, which is a critically important social issue.”
Doron Weber, vice president of programs and program director at the Alfred P. Sloan Foundation and one of three discussants for Finley and Nyhan’s presentation, pointed out that the mindsets of hesitant parents can be changed through direct or indirect means. Monetary incentives can directly affect actions, beliefs, or behaviors. Indirectly, films, plays, books, television shows, and social media can influence the culture and the decisions people make within that culture. For example, the Sloan Foundation’s support of the book Hidden Figures, which led to the movie of the same name about African-American female mathematicians working for NASA, could very well do as much to encourage women and underrepresented minorities to pursue science, technology, engineering, and mathematics (STEM) careers as more directly focused efforts, Weber argued. Similarly, the foundation has supported a play on vaccination and projects on the public health heroes Jonas Salk and D. A. Henderson. These cultural influences “are as powerful a weapon as we have in our arsenal, and we should not forget them.”
Suzanne Ffolkes, vice president of communications at Research!America, pointed to the critical importance of the messenger and trustworthiness. Surveys have shown that Americans trust scientists and medical professionals, and the nature of this trust could be probed in research collaborations. This is one of many ways, she added, in which social and behavioral scientists could make major contributions through collaborations between practitioners and scholars. Finally, she called attention to the effects of cultural differences on vaccine hesitancy—for example, some population groups may be more likely to worry about possible side effects of vaccines than others.
Paul Hanle, president and chief executive officer of Climate Central, pointed to the parallels between the approach Finley and Nyhan are taking and the approach needed to understand public perceptions of climate change. Studies of vaccine hesitancy could yield “lessons that we might be able to take forward to apply, for example, to the extraordinarily
entrenched views that are in the climate conversation.” Responses may differ among groups and locations, and such differences can provide “an interesting opportunity to look at something that is more localized.”
Finally, Greg Boustead of the Simons Foundation emphasized the importance of gaining greater understanding of the power of storytelling, particularly among people who might be skeptical on an issue. Science Sandbox is a strategic portfolio for grantmaking that has targeted people who do not identify as science enthusiasts, “and there is a lot we don’t know and understand about what resonates with these communities.” Partnerships between social and behavioral scientists and content creators can help produce the kinds of outcome measures and evaluations that funders are seeking.
Robert Hornik, Wilbur Schramm Professor of Communication and Health Policy at the University of Pennsylvania’s Annenberg School of Communication, offered another perspective on the prevalence of vaccination in the United States. People have made the claim that vaccine rates have fallen because of false claims about a link between vaccine and autism, he noted. However, the percentage of 19- to 35-month-olds in the United States who have completed the seven-vaccine series rose from 2009 to 2011 and stayed constant, at about 70 percent, from 2011 to 2015. Similarly, the vaccination rate for measles, mumps, and rubella (MMR) stayed constant, at about 90 percent, from 2000 to 2016.
Few studies have specifically tested the effects of beliefs about autism on vaccination levels. Smith et al. (2008) found that selective nonreceipt of the MMR vaccine rose from 0.75 percent in 1995 to 2.00 percent in 2000 after the first autism-MMR study appeared in the medical literature, but it returned to baseline before any substantial media coverage of the issue, suggesting that the decline in vaccination rate was linked to professional wariness and not to public concerns. Other studies have assessed the presence of antivaccine content on traditional and social media, have examined beliefs about vaccine risks versus vaccine benefits, and have done cross-sectional associations of beliefs with exemption requests or, rarely, with vaccination rates. The association may be relevant in some regions or for some vaccine-resistant subgroups. But, together, the studies provide “quite limited support for the specific hypothesis that the false beliefs about vaccination and autism drive low vaccination rates.”
Hornik called attention to a different association—between vaccine rates and economic status. Completion rates for the seven-vaccine series are substantially lower for children living below the poverty line than children living above the line. “Is there a difference in beliefs, or a differ-
ence in misinformation between these groups? Or is it more likely that there’s a difference in structural circumstance—the ability to deal with the logistic and financial complexity of completing the entire series in a timely way?” Hornik asked. A study that Hornik and his associates did in poor areas of Philadelphia showed that individual beliefs about benefits and risks of vaccination were not important correlates of vaccination levels. What mattered was the structural circumstances people faced: long wait times at overburdened clinics, missed opportunities for vaccination, older siblings who needed the car, the cost of taking time off work, and so on. “Our advice to the city of Philadelphia was to forget about a communication campaign and fix the clinical system,” said Hornik.
He recommended starting with the behavior of concern and considering misinformation as one among other competing explanations of that behavior. If it does turn out to be an important explanation, remediation can be considered. “Scientific misinformation may matter, but it does not always matter. We need to show that it does.”
Adding to the complexity of science communication is the problem of science, health, and cultural literacy, said Susan Scrimshaw, co-chair of the board of directors for the Nevin Scrimshaw International Nutrition Foundation and former president of the Sage Colleges. According to the National Center for Education Statistics, about one in seven people in the United States cannot read, and another 21 percent read below a fifth-grade level, she observed. The challenge is to empathize, be supportive, and form communities with people whose literacy is limited.
Low literacy makes communication about health and infectious diseases particularly difficult, Scrimshaw observed. Even most people with college educations are not going to understand such terms as diurnal, herd immunity, genetic load, endemic, or transmission. Further confusion can result when evidence leads to changes in findings and recommendations, as has happened recently with prostate cancer screening and with blood pressure treatment recommendations.
As an example of translation difficulties, Scrimshaw cited work she and her colleagues have done in Spanish-speaking populations where they used the term “riesgo,” which is the literal translation of “risk.” But communications research with the intended audiences showed that the word “peligro,” or danger, resonated much more with women in that population. “We never would have thought of that from the psychology or public health perspectives. The women taught us how to communicate with them.”
Language, culture, and education create variation in what people hear. Ethnicity and culture, language, gender, age, low literacy or low Internet literacy, economic status, blindness, and deafness all heighten the risk for miscommunication, said Scrimshaw. The risks of Zika infection are an example. People need specific information about the risks of infection and of a baby being affected. But the complexities of the behavior of mosquitoes make the task of conveying information difficult. Furthermore, when sexual transmissions proved to be occurring, discussion of condoms was a nonstarter in some political climates. Public health officials might assume that what they recommend is possible. But, for example, it may not be possible to avoid getting bitten by mosquitoes following a hurricane, when the power is out and buildings are damaged, which is what occurred in 2017 in the U.S. Gulf Coast and the Caribbean.
In general, people get information about health and infectious diseases in different ways than in the past. They have shifted away from newspapers and television to Internet sites and social media contacts. Judging credibility, developing trust, and understanding sources are all different for these media than for traditional media. Furthermore, culture, education, language, and other factors influence these processes, even as the ways people get information continue to change.
Scrimshaw drew several conclusions from her observations. First, communicators need to understand the cultural and linguistic parameters for the groups they are trying to reach. “One message does not fit all,” she said. Iterative consultative and action relationships with communities help build trust. Telling stories that illustrate data and desired behaviors can be powerful influences on beliefs and behaviors.
Communicators also need to understand the feasibility of the behaviors they are seeking in terms of access, economic constraints, priorities, and other factors, said Scrimshaw. They then need to develop strategies to enable the desired behaviors. Social mobilization and community engagement can foster communication, as can integration of the social, communication, and health sciences. Long-term funding for capacity building and preparedness require political will and understanding. Finally, the role of the National Academies of Sciences, Engineering, and Medicine in building capacity and serving as science “advisers to the nation” is critical, she said.