The last decade has seen repeated calls for scientists to communicate their work to legislators in order to promote the inclusion of “usable” scientific information in policy decisions [1, 2]. But to what extent are political leaders interested in listening to scientists, particularly when contending with politically polarized issues such as global climate change (GCC) [3]? Accessible scientific information has the potential to cut through political polarization and gridlock by presenting objective facts to counter subjective opinions, thus facilitating the ability of legislators to make science-based decisions regarding policies and funding priorities [4, 5]. This need for science-based policy has become particularly dire in the United States where members of Congress are debating the science and societal implications of GCC, while at the same time some are attempting to wrest decisions regarding scientific funding priorities away from heads of national science agencies such as NSF and NASA [6, 7]. While these observations are independently alarming, several questions important for diagnosing and overcoming the gridlock remain unanswered. For instance, has science overall become polarized to the extent that outward interest in science (and not just climate science) might harm partisan identity? Similarly, does interest in scientific information actually influence political decision making, or are such decisions made mostly on the basis of partisan group identity? Answers to these questions have the potential to help maximize the ability of the scientific community to effectively communicate both environmental and climate science [8]. Both Republicans and Democrats have been accused by their political opponents as being “anti-science,” to a large extent on the basis of policy decisions regarding issues as diverse as climate change, evolution, vaccination and genetically modified organisms [9, 10]. Several studies have suggested that conservatives have become increasingly distrustful of science [11] and political differences among members of the American public in their level of trust in climate science and “belief” in anthropogenic climate change have been well-established [12, 13]. Other studies, however, have suggested that such generalizations are too simplistic, arguing for example that self-identified economic (as opposed to social) conservatives may be as or more scientifically literate than economic leftists [14]. Explorations such as these are important if we as scientists hope to engage constructively with the multiple publics on pressing societal issues such as GCC.
While a growing body of literature has examined perceptions of and interest in science among the lay publics, we know considerably less of how such differences may play out among elected officials, in part because they may be less willing to discuss their personal views openly. We analyzed the public Twitter profiles of U.S. senators as a window into these questions. The Twitter handles that a user follows serve as an indicator of the people, organizations, and issues in which the user is interested [15, 16], and are arguably an objective reflection of the public image a user wishes to portray. Twitter is thus now commonly used by the offices of U.S. political leaders not only for information sharing and issuing “calls to arms” to their constituents, but perhaps even more importantly serves as a platform for self-promotion and a public expression of their core values [17]. Social media platforms like Twitter also offer a means of rapidly disseminating information, and instantly sharing links to news stories of interest between a user and their followers. In this way, networks or communities of users with common interests often form, within which information is then shared [17]. For example, Twitter is a place where interested followers can be alerted to scientific breakthroughs in real-time and where “buzz” about newly published research is measurable [18]. A recent study by the Pew Research Center found that 63 % of Twitter users report that they use the platform as a source of news, up from 52 % in 2013 and almost twice that of Facebook users [19]. The accounts that a user opts to follow thus determine the information that she or he sees on Twitter, as well as a statement of the other users with which the person wishes to be associated publically.
A 2015 survey by Jasny and colleagues [20] emphasized that in political spheres information on climate change often reverberates within a series of “echo chambers” with relatively few opportunities for cross-fertilization. Analyses of the general public based on Twitter, in contrast, suggest that while information on political issues is exchanged primarily among users with similar ideologies, information regarding other current events can be more dynamic and at least initially starts as a “national conversation” that may eventually devolve into partisan bickering [16]. And, while some studies have demonstrated substantial differences among Republicans and Democrats in the intensity to which they respond negatively to scientific information that conflicts with their world view [21], others have suggested that “…scientific recommendations on public policy are taken seriously by partisans of all stripes” [22]. While numerous surveys have documented strong partisan differences in understanding and acceptance of GCC among members of the general public [12, 13] and have pointed to the role of partisan media in widening such gaps [23], less is known about how political leaders receive and share such information, especially using increasingly important social media platforms such as Twitter. Specifically, are political leaders exposed to similar scientific information but responding to it based on partisan identity [21], or are they interested in and exposed to information from fundamentally different sources, as shaped by the communication networks to which they belong [20, 24]? Has interest and trust in science- and notably not just climate science- truly become a partisan wedge issue in the 21st century?
We analyzed the 78,753 unique Twitter accounts followed by U.S. senators to gauge their interests as well as the public images that their offices project to the public. We were particularly interested in determining if higher levels of overt interest in and exposure to science and climate-related Twitter feeds had any correlation with senators’ positions on GCC. We took advantage of votes (made in January 2015) on a series of amendments associated with the Keystone oil pipeline legislation clarifying the U.S. Senate’s position on GCC. We then used ordination and network analysis to identify influential people and organizations that might best provide an avenue for injecting scientific information into the network(s).