英文摘要: | Diverse methods have been applied to understand why science continues to be debated within the climate policy domain. A number of studies have presented the notion of the ‘echo chamber’ to model and explain information flows across an array of social settings, finding disproportionate connections among ideologically similar political communicators. This paper builds on these findings to provide a more formal operationalization of the components of echo chambers. We then empirically test their utility using survey data collected from the community of political elites engaged in the contentious issue of climate politics in the United States. Our survey period coincides with the most active and contentious period in the history of US climate policy, when legislation regulating carbon dioxide emissions had passed through the House of Representatives and was being considered in the Senate. We use exponential random graph (ERG) modelling to demonstrate that both the homogeneity of information (the echo) and multi-path information transmission (the chamber) play significant roles in policy communication. We demonstrate that the intersection of these components creates echo chambers in the climate policy network. These results lead to some important conclusions about climate politics, as well as the relationship between science communication and policymaking at the elite level more generally.
Environmental politics continue to be highly contentious, and nowhere has this debate become more deeply entrenched than in the issue of climate change. Despite a well-documented scientific consensus on the causes and drivers of global climate change, legislation has yet to be passed in the United States at the federal level to address these issues. As scientists continue to warn decision makers about the need to act1, 2, 3, 4, 5, the political debate remains polarized. Furthermore, this political polarization often manifests among political elites as debates over the veracity and legitimacy of established scientific consensus6. In January 2015, while debating the Keystone XL oil pipeline in the United States Senate, for example, an amendment was offered to get the ‘sense of the Senate’ about whether humans contribute significantly to climate change7. The vote was split, 50-49, with 49 Senators refusing to affirm that climate change is anthropogenic. Numerous studies have aimed to understand why the science of climate change continues to be challenged within policy circles, focusing on the media coverage of the issue8, 9, 10, the role that conservative think tanks have played in creating a countermovement11, 12, 13, 14, and the ways the issue has been discussed by the US Congress6, 15, 16. Within this literature, scholars have invoked the notion of echo chambers to describe how information has become a partisan choice, and how those choices bias towards sources that reinforce beliefs rather than challenge them, regardless of the source’s legitimacy17. Within the broader literature, echo chambers are described as social network formations that transform the ways in which information is transmitted and interpreted by actors18, 19, 20, 21, 22. Scholars have applied this concept to illustrate the dynamics of issue positions of candidates and political parties23; the public consumption of media24, 25; the homophily of online communication networks26, 27, 28 (defined as the presence of ties among actors who share the same attribute29); and multiple aspects of blogs and blog-based discussion17, 29, 30, 31, 32, 33. This paper builds on this extant literature by providing a more formal operationalization of the components of echo chambers and then testing empirically for their presence against competing network mechanisms within the US climate policy network. We conceive of the echo chamber as being comprised of two distinct processes. First, information is an ‘echo’ when it repeats what one already believes. Called ‘confirmation bias’ in the psychology literature, information is perceived to be more credible when it matches the recipient’s world view34, 35, or when individuals hear the same information from different sources, even if that information ultimately came from one original source35, 36. Furthermore, hearing repeated messages has been found to intensify viewpoints further and push some to extreme opinions37, 38, 39, 40. Although this process of influence homophily in information transmission involves some element of time, in this first examination of the echo chamber, we test this operationalization with a static model. The second mechanism is the formation of ‘chambers,’ or structures that provide the space needed for information to echo. Our ‘chamber’ is the smallest network structure that provides the conditions for the same information to be transmitted from one source to one recipient via different paths. In other words, as we describe in detail below, the ‘chamber’ involves at least three actors: a speaker, a receiver, and a mediating actor through which the information can travel. This directed multi-path transmission distinguishes echo chambers from other polarization mechanisms16, 38, 39, 40, 41, 42. The combination of homophily with the bonding social capital found in cohesive triads19 makes these echo chambers a fundamentally different network structure. In the pages that follow, we depart from the previous literature on echo chambers by examining the interaction of the echo and chamber mechanisms within a statistical framework. To accomplish this goal, we examine the information networks that supplied members of the climate policy community in the United States in 2010 with research, advice and perspectives on climate change. Below, we briefly describe the data collected, operationalize our understanding of echo chambers using social network methods, and apply exponential random graph (hereafter, ERG) model simulation methods to test for the presence and significance (relative to tie formation) of such echo chambers among members of the US climate policy network. For a full discussion of the policy network approach, which uses policy actors as the unit of analysis, see the work of Knoke41 and Laumann and Knoke43. Our network is comprised of the set of the policy actors in our sample who responded to our survey (64 in total) and all reported directed communication within this population. For example, when actor B states that they received information from actor A, there is a directed tie from A to B (Fig. 1a). Given this network formalism, we can now specify the two characteristics of echo chambers.
| http://www.nature.com/nclimate/journal/v5/n8/full/nclimate2666.html
|