The influence of corporate funding on two topics identified by computational analysis within the climate change contrarian movement. The horizontal axis spans the years 1993–2013. The vertical axis indicates how much the topic was written about, expressed as a decimal fraction. The red line represents the prevalence of the topic in the texts of contrarian organizations who received money, and the black line represents the prevalence of the topic for contrarian organizations who did not receive money. Source: Farrell (2015) Proceedings of the [U.S.] National Academy of Sciences, published online, www.pnas.org/cgi/doi/10.1073/pnas.1509433112.
A rigorous computer-driven analysis of organizations producing documents and speeches deemed to express contrarian views on climate change, the texts of these documents, and energy companies providing funding to those organizations has been carried out. Entities receiving funding are more closely connected to one another than are those not supported. The entities most active in generating communication are highly influential in the contrarian movement. Several contrarian topics, from energy industry-funded entities, were the subjects of communications that increased in number strongly in the years since about 2007 compared to those from nonsupported entities. Entities receiving funding were more likely to have produced contrarian writings. These findings provide some understanding of why public opinion in the U. S. is more dismissive of the findings of climate science than it is in other industrialized countries.
The United States Congress in recent times has opted against enacting national policies that would lower the annual rate of emission of greenhouse gases (GHGs), especially carbon dioxide (CO2). We in the U. S.have already suffered significant damages and harms brought on by the effects of global warming. These include fair weather ocean flooding and heat-driven droughts in the Midwest and Southwest. Damage from extreme events has high costs for recovery which ultimately end up being paid by individuals directly, or indirectly due to higher taxes needed for the extra, unforeseen government services. On the other hand, federal legislation to combat warming and its harms would recognize, and counteract, the increasingly significant role of human activities in the energy economy that result in higher GHG emissions.
In a democratic republic such as the U. S. we idealize that our representatives are responsive to the positions taken by their constituents. In reality, however, the American political system has long been accused of being in the thrall of powerful economic interests who contribute to the election campaigns of our Congresspersons. Campaign support from the fossil fuel industry is a prime focus in such discussions. This undercurrent, while widespread among the electorate, is difficult to prove.
Justin Farrell has recently published two reports on potential connections between funding from energy producers and public discourse. Farrell’s interest is the pronounced polarization in policy discussions of climate change. He points out that earlier work in this field has been conducted at an individualized level of inquiry. In contrast, he has taken advantage of the power of contemporary computing power to analyze effects that funding contributions from the energy sector may have had on climate change discussions from 1993 to 2013 (see Details at the end of this post).
In one article (Farrell (2015) Nature Climate Change, published online) the author applies computational social science to analyze networks among individuals and organizations producing discourse promoting climate-contrarian points of view, as well as the effect that funding from two fossil fuel corporations have on the networks (see Details). Organizations receiving funding from the corporations are more closely tied to one another with high significance, in a test of connectivity, than those that are not so funded. The funded entities thus “have greater influence over flows of resources, communication, and the production of contrarian information.”
The most important factor in this result is the very fact of being funded. The funded entities achieve a higher level of similarity in phrasing and expression in news reports, to a very high level of statistical significance. Power within the organizational network is not evenly distributed, but rather is highly centralized among that smaller group of organizations having ties to private sector entities. In Farrell’s view, these “findings … help to explain why climate science rejection is so pronounced in the United States compared to other developed nations.”
In his second article (Farrell (2015) Proceedings of the [U.S.] National Academy of Sciences, published online) the author examined relationships among the contrarian organizations, their ties to the funders, and the effects of funding on the thematic material produced. First, to a high degree of statistical significance, the “results suggest that organizations within the movement who made an effort to produce textual discourse about climate change are the most central to the movement itself, providing them more influence over the transfer of information.” Other analyses yielded the topics prevalent in the document texts, established clustered relationships among the topics, and compared the time-dependent production of topics by funded entities vs. nonfunded entities. Examples of timelines for the topics “CO2Is Good” and “Climate Change is a Long-Term Cycle” are shown below.
Importantly, energy corporate funding is seen to have influenced the extent of polarization in climate change writing over the 20 years ending 2013. In particular, thematic content on topics such as “CO2 Is Good” and “Climate Change is a Long-Term Cycle” (left and right panels, respectively, in the graphic above), deemed polarizing and subject to debate, produced by funded organizations was far more prevalent in the last few years of this period, compared to content from nonsupported entities.
Farrell concludes that “corporate funding influences the actual language and thematic content of polarizing discourse.” Entities that received funding from the energy companies were more likely to have produced texts characterized as polarizing discourse on climate change than those not so funded. He also concludes “organizations that received corporate funding were more likely to have written and disseminated contrarian texts”.
The author emphasizes that, because of the computationally large size and objective analytical approach using robust procedures (see Details), these findings are highly significant, confirming earlier but more poorly substantiated concepts on climate change knowledge and politics.
Farrell has analyzed the interrelationships among sources of contrarian discourse that were identified by others, their contrarian writings, and the existence of funding from energy company sources (exemplified here by ExxonMobil and the Koch family foundations). The period considered is from 1993 to 2013. He finds statistically strong correlations between funding and the connectedness between sources of discourse, such that funded organizations wield considerable influence in establishing the frames of discussion. Corporate funding affected the wording and overall themes of discussion on topics considered to be polarizing in the sense that the positions taken are either contrary to, or dubious in the face of, reality. The author concludes that his characterizations help us understand the basis for the relatively high prevalence of rejection of climate science in the U. S., when compared with other industrialized countries of the world.
The energy companies studied by Farrell are representative of those in the energy industry more generally. In a time of great ferment worldwide seeking to reduce GHG emissions to near zero, energy companies clearly have an interest in preserving the status quo, so that the demand for their products continues unconstrained. Farrell’s work shows that in the U. S. the financial support provided by the two energy companies studied here correlates strongly with the written and verbal communications produced by the organizations receiving the funding. These writings run contrary to the objective findings of the worldwide climate science community and the policies that these findings strongly suggest.
The nations of the world are coming to realize the need for a universal effort to limit bring annual emissions to very low levels by mid-century. They, and their populations, have to resist the flow of contrarian writings and speeches. The science underlying contemporary climate change is beyond questioning, and has reached the stage requiring concerted, meaningful actions. Contrarian communication from energy companies seeking to preserve “business as usual” must be resisted.
In both articles (Farrell (2015) Nature Climate Change, published online; and Farrell (2015) Proceedings of the [U.S.] National Academy of Sciences, published online) the author used computational analysis first, of institutional and social network structure among individuals and organizations promoting climate contrarian points of view, and second, of analysis of texts containing climate contrarian viewpoints. Farrell labeled “contrarian organizations [as] those identified by prior peer-reviewed research as overtly producing and promoting skepticism and doubt about scientific consensus on climate change”.
The social network comprises 4,556 individuals connected to 164 organizations promoting contrarian viewpoints. The organizations include think tanks, foundations, public relations firms, trade associations, and ad hoc groups.
The texts were produced between 1993 and 2013, and include written and verbal content from the organizations identified, as well as from three major news outlets, the U. S. presidents and floor speeches in the Congress (40,785 documents containing over 39 million words). They were analyzed computationally to produce contextually-driven topics using a process called Structural Topic Modeling, with sets of word stems identified for each topic. This process was chosen because it permits consideration of document attributes such as the year written, or organizational attributes considered in the article such as corporate funding. Importantly, the topics are not assigned manually in advance. Rather, they are identified as a result of the self-learning features incorporated into the analysis. In addition, Farrell considered whether the individuals or organizations had ties to energy industry entities.
Energy companies were restricted in the end to only ExxonMobil and the three Koch family foundations, the philanthropic organizations established by Koch Industries, because of the greater reliability of the information on contributions available for them. (Koch Industries is a conglomerate active in the exploration, production, refining and distribution of petroleum and its products, among other endeavors.)