Addendum to ‘Assessing ExxonMobil’s climate change communications (1977–2014)’ Supran and Oreskes (2017 Environ. Res. Lett. 12 084019)

In our 2017 study ‘Assessing ExxonMobil’s climate change communications (1977–2014)’, we concluded that ExxonMobil has in the past misled the public about climate change. We demonstrated that ExxonMobil ‘advertorials’—paid, editorial-style advertisements—in The New York Times spanning 1989–2004 overwhelmingly expressed doubt about climate change as real and human-caused, serious, and solvable, whereas peer-reviewed papers and internal reports authored by company employees by and large did not. Here, we present an expanded investigation of ExxonMobil’s strategies of denial and delay. Firstly, analyzing additional documents of which we were unaware when our original study was published, we show that our original conclusion is reinforced and statistically significant: between 1989–2004, ExxonMobil advertorials overwhelmingly communicated doubt. We further demonstrate that (i) Mobil, like Exxon, was engaged in mainstream climate science research prior to their 1999 merger, even as Mobil ran advertorials challenging that science; (ii) Exxon, as well as Mobil, communicated direct and indirect doubt about climate change and (iii) doubt-mongering did not end after the merger. We now conclude with even greater confidence that ExxonMobil misled the public, delineating three distinct ways in which they have done so.


Introduction
In our recent article (Supran and Oreskes, 2017 Environ. Res. Lett. 12 084019 [1]), we assessed whether ExxonMobil has in the past misled the general public about anthropogenic global warming (AGW) (we refer to Exxon Corporation as 'Exxon' , Mobil Oil Corporation as 'Mobil' , ExxonMobil Corporation as 'ExxonMobil Corp' , and generically refer to all three as 'ExxonMobil'). Presenting an empirical document-by-document textual content analysis of the company's private and public climate change communications-including peer-reviewed and non-peer-reviewed publications, internal company documents, and paid, editorial-style advertisements ('advertorials') in The New York Times (NYT)-we concluded that it has.
After our study was published, we became aware of additional relevant ExxonMobil advertorials not included in our original analysis. Here, we present a document-by-document content analysis of 1448 advertisements, which include these additional materials. Our original finding is reinforced: between 1989-2004, Mobil and ExxonMobil Corp advertorials overwhelmingly expressed doubt about AGW as real and human-caused, serious, and solvable. By including additional advertorials in this expanded analysis, we now conclude with even greater confidence that Exxon, Mobil, and ExxonMobil Corp misled the public.
We also address a critique that ExxonMobil Corp has raised about our original study: that it 'obscur [ed] the separateness of the two corporations' , Exxon and Mobil, thereby rendering our conclusions invalid [2,3]. This was never the case: our article's citations explicitly attributed each individual advertorial to one of Exxon, Mobil, or ExxonMobil Corp; we did not obscure anything. It is the case that to avoid overcomplicating or belaboring the point, our original article focused on how the three companies-Exxon, Mobil, and ExxonMobil Corp-have collectively misled the public. We considered this approach appropriate, because when Exxon and Mobil merged, ExxonMobil Corp inherited legal and moral responsibility for the parent companies. We reject the implied argument that ExxonMobil Corp is somehow not responsible for the actions of Exxon or Mobil, whatever they may have been. Here, we show ExxonMobil Corp's critique to be incorrect both statistically and at the level of individual documents. We delineate three distinct ways in which the data demonstrate that Exxon, Mobil, and ExxonMobil Corp have all, variously, misled the public about AGW.

Method
Previously we demonstrated that between 1989-2004, available advertorials-paid, editorial-style advertisements on the Op-Ed page of the NYT-published by Mobil and ExxonMobil Corp overwhelmingly expressed doubt about AGW as real and humancaused, serious, and solvable [1]. In this study, we analyze additional advertorials that came to light after our study was published.
We adopt the same methodology as in our prior study, characterizing each document's manifest content in terms of its (i) topic, (ii) position with respect to AGW, and (iii) position with respect to risks of stranded fossil fuel assets [1]. Results from our original analysis of the 32 Internal memos, 72 Peer-Reviewed articles, and 47 Non-Peer-Reviewed articles made available by ExxonMobil Corp are carried forward (see table 1). As before, our analysis compares these documents with Mobil and ExxonMobil Corp's public outreach in the form of advertorials in the NYT.
We previously analyzed 36 AGW-relevant advertorials from a collection of 97 compiled by Pol-luterWatch based on a search of the ProQuest archive [1,6,7]. Here, we add to this dataset of 36 by running two additional Boolean ProQuest searches (see section S1, supplementary information for details). In the first, we query for all advertisements in the NYT between 1923 and 2018 that refer to 'Mobil' or 'Exxon' or 'ExxonMobil' and to one or more of 13 keywords pertaining to AGW (based on a word frequency analysis of all advertorials included in [1]): 'climate' or 'climate change' or 'greenhouse' or 'global' or 'warming' or 'Kyoto' or 'carbon' or 'CO2' or 'dioxide' or 'temperature' or 'GHG' or 'Fahrenheit' or 'Celsius' . This relevance sample search yielded 1412 documents [8]. In our second search, we query for all advertisements published in the NYT on Thursdays between 1970 and 2018, and that refer to 'climate change' or 'global warming' or 'greenhouse gas' or 'greenhouse gases' or 'greenhouse effect' or 'carbon dioxide' or 'CO2' . (This search specifically targets Mobil and ExxonMobil Corp's 'every Thursday'  and 'every other Thursday' (2001+) advertorials [9,10].) This search yielded 138 documents. Combining the above three datasets and removing redundancies yielded a total of 1448 documents spanning 1924-2013 (see table S4, supplementary information). Despite our comprehensive search, additional unidentified advertorials may, of course, exist. We would welcome ExxonMobil Corp making publicly available a complete online database of its-and Mobil's-advertorials in all newspapers (archived versions of the company's website show that in the past, some-but not all-advertorials were listed, albeit misrepresented as 'Op-Eds' [11]).
Eight research assistants conducted an initial, high-level content analysis to filter for relevance the 1412 documents generated by the first ProQuest search. The assistants downloaded and inspected each individual document within assigned publication windows spanning one to ten years. Applying a standardized procedure, they binned each document as either 'irrelevant' or 'not irrelevant' (subcategories of 'relevant' , 'generic' , and 'ambiguous') to AGW, erring heavily on the side of caution (even most 'not irrelevant' documents do not, in fact, express any positions on AGW). The remainder of the 1448 documents were likewise binned by one of the authors. To verify intercoder reliability, each analyst independently coded a random subset of 100 documents (approximately 7% of the total number of documents; equivalent, on average, to 61% of the number of documents analyzed by each assistant). In sum, this yielded 267 'not irrelevant' advertorials (intercoder reliability: percentage agreement = 92%; Krippendorff 's α = 0.77; these are conservative lower-bounds owing to Type I errors, the true value is close to unity-for details see section S1, supplementary information). The authors then coded these 267 advertorials according to the content analysis scheme detailed in [1]. (This included occasional reevaluations of codes assigned in our original analysis. ) We have also obtained additional non-peerreviewed documents not included in our original study, such as company reports, webpages, and speeches. These inform our interpretation of the results of our content analysis. The sources for these additional documents include the Climate Files archive maintained by Climate Investigations Center, ExxonMobil webpages, and digital archives (Wayback Machine) of earlier ExxonMobil webpages [12,13]. Unlike other document categories, which are bound sets, non-peer-reviewed documents are virtually limitless in potential number and scope (see footnote on p. 2, [1]). Accordingly, while we introduce specific new non-peer-reviewed documents in this paper in order to inform our Discussion, we do not systematically assess their positions using content analysis. Table 1 and figures 1 and 2 reflect only those nonpeer-reviewed documents included in our original study. Table 1. Inventory of documents analyzed. Shown for each document category are the total number of documents, their date range, source(s), and assigned types. The internal, peer-reviewed, and non-peer-reviewed documents are those studied in [1]. Among peer-reviewed and non-peer reviewed documents, eight publications were found to be redundant, with similar or identical wording to seven other (strictly unique) publications. All [14,15]. The advertorial was condemned by the former director of the National Assessment Coordination Office: 'To call ExxonMobil's position out of the mainstream is…a gross understatement' [16]. Another 2000 ExxonMobil Corp advertorial says that 'climate change may appear as confusing as a maze' [17]. Expanding beyond our original analysis to include 4 and 18 new advertorials published pre-1989 and post-2004, respectively, figures 1(a) and 2(a) (left bars) show that 'Doubt' continues to account for half of all positions (31/61 = 51%), though it loses some ground to the 'Acknowledge' stance (23/61 = 38%). The remaining positions are shared between 'Reasonable Doubt' and 'Acknowledge and Doubt' (5/61 = 8% and 2/61 = 3%, respectively). Examples of 'Doubt' include three ExxonMobil Corp advertorials in 2007, which, despite acknowledging 'the risks of climate change' , variously say that 'climate science remains extraordinarily complex' , that it is 'evolving' , and that 'areas of uncertainty do exist' [18][19][20]. Of those advertorials expressing 'Acknowledge' from 2005 onwards, 93% (14/15) do so only implicitly (EP3a), almost exclusively by discussing mitigation (such as energy efficiency and technology innovation) rather than climate science. None explicitly say that climate change is real and humancaused.

Endorsement Levels (ELs)-AGW as real and human-caused
Accompanying the emergence of implicit acknowledgments is a rhetorical framework focused on 'risk' . 'Risk(s)' of AGW (or of greenhouse gases) becomes ExxonMobil Corp's watchword, appearing at least once in 87% (13/15) of these advertorials (table S4, supplementary information). A characteristic example is a 2007 advertorial entitled 'Saving Energy and Reducing Greenhouse Gas Emissions' , which refers to 'steps ExxonMobil is taking to address the risk of climate change' and says that 'industry, consumers and policymakers all have a role to play in addressing the risks of climate change' [21]. A 2008 advertorial discusses lower-carbon fuels and other approaches to 'addressing the risks posed by rising greenhouse gas emissions' , but without mentioning AGW [22].
These observations-of implicit acknowledgments and 'risk' rhetoric-are part of a wider trend. Regarding the former: across all advertorials in all years, only two express any form of explicit acknowledgment (EP2). One, a borderline case in 2005, does so only indirectly, by quoting a statement from the Group of Eight (G8) that does not address causation [23]. The other, in 1989, is not in fact an advertorial, but an advertisement in The New York Times Magazine that may or may not have actually included Exxon among its industry sponsors [24]. All other acknowledgments are implicit: they avoid directly addressing climate science and the issue of human causation, instead discussing emissions reductions strategies. Figure S1, supplementary information, shows that from the late 1990s onwards, advertorials focused on mitigation rapidly outnumbered those focused on methods and climate science-cumulatively, by more than three-to-one.
We shall address the wider trend concerning 'risk' rhetoric in a forthcoming study. See   [16]] [14].

2008
'To meet this [higher future global energy] demand, while addressing the risks posed by rising greenhouse gas emissions, we will need to call upon a broad mix of energy sources' [22].

2007
'Climate remains an extraordinarily complex area of scientific study. But the risks to society and ecosystems from climate change could prove to be significant-so despite the areas of uncertainty that do exist, it is prudent to develop and implement strategies that address the risks' [20]. (continued)

1996
UN-sponsored climate action 'is likely to cause severe economic dislocations…If developed nations act alone to reduce emissions, the staggering cost imposed on energy-intensive industries will drive nations to export much of their industrial base to countries with less stringent controls. World economic health will suffer as nations are forced to switch from fossil fuels, saddled with large carbon taxes and driven to prematurely scrap many factories and machinery. The dislocations will be even more severe if the solutions are not implemented globally…Jobs and livelihoods are at stake [in deciding on climate policy]' [26].

2007
'Businesses, governments and NGOs are faced with a daunting task: selecting policies that balance economic growth and human development with the risks of climate change' [18,19].   [26]. The 2000 ExxonMobil Corp advertorial discussed earlier claims that the US National Assessment 'report's language and logic appear designed to emphasize selective results to convince people that climate change will adversely impact their lives'implying that it will not [14,15]. A third example is a 1993 Mobil advertorial entitled 'Apocalypse No' [25], which claims that 'dire predictions of global warming catastrophes' in 1992 were 'media hype' that 'did not properly portray the consensus of the scientific community' . It goes on to argue that 'what's wrong with so much of the global warming rhetoric' is 'the lack of solid scientific data' , and alleges 'a noticeable lack of evidence of the sky actually falling' and 'colder than normal temperatures' in the US The advertorial quotes prominent climate contrarian Robert C. Balling, who argues 'that the apocalyptic vision is in error and that the highly touted greenhouse disaster is most improbable' . The advertorial also quotes physicist S Fred Singer, well known at the time for challenging the scientific evidence of stratospheric ozone depletion, claiming that: 'the net impact [of a modest warming] may well be beneficial' [27]. Expanding beyond our original analysis to include all years has little effect on the overall result: 'Doubt' continues to dominate (19/29 = 66%), while 'Acknowledge' and 'Acknowledge and Doubt' make up the difference (5/29 = 17% apiece). Post-2004, advertorials are virtually silent about the seriousness of AGW (beyond generic 'risk' statementssee [1]). In other public communications, however, this doubt has continued (a few examples are given in table 3-see ExxonMobil Corp statements from ∼ 2008 onwards).
A characteristic example of doubt that AGW can be effectively addressed (table 2, bottom row) is a 2000 ExxonMobil Corp advertorial (not included in our original dataset) that says the Kyoto Protocol to the United Nations Framework Convention on Table 3. Examples of public doubt about AGW either directly communicated or indirectly funded by ExxonMobil Corp following the merger of Exxon and Mobil. Quotations are sourced from documents not included in our content analysis, such as company reports, speeches, newspaper accounts, and archived websites. Although we do not formally code the positions of these statements on AGW, and the relative 'strengths' of doubt vary from statement to statement, ExxonMobil Corp's direct representations through 2007/8 appear to express doubt about AGW as real and human-caused. Through to the present day, the company continues to itself question the 'competency' of climate models and the role of humans as the 'principal drivers of climate change' , yet emphasis also shifts to promoting doubt about AGW as serious and solvable (as indicated, most statements also include 'risk' rhetoric). Examples are also given of third-party individuals and organizations funded by ExxonMobil Corp that have communicated doubt about AGW as real and human-caused, serious, or solvable in the recent past and/or present.

Year
Publication Quotation

2000
Company report (preface by CEO Lee Raymond) [ Academic (non-peerreviewed) article funded by ExxonMobil (also Charles G Koch Charitable Foundation and American Petroleum Institute) [111] '[I]t is highly premature to argue for the extinction of polar bear [sic] across the circumpolar Arctic within this century…It is certainly premature, if not impossible, to tie recent regional climatic variability in this part of central Canada to anthropogenic greenhouse gases and, further, to extrapolate species-level conditions on this basis…[T]here is no ground for raising public alarm about any imminent extinction of Arctic polar bears' .
(continued) ExxonMobil affiliate, Syncrude [117] Syncrude submits that the production and consumption of petroleum fuels is not dangerous and does not pose a risk to human health or safety' .

2015
Senator Rex Tillerson, Congressional testimony [120,121] 'I understand these [greenhouse] gases [due to 'combustion of fossil fuels'] to be a factor in rising temperature, but I do not believe the scientific consensus supports their characterization as the 'key' factor' . Risk rhetoric: 'risk' .

1992-2018
American Legislative Exchange Council, funded by ExxonMobil [122][123][124] 'Global Climate Change is Inevitable. Climate change is a historical phenomenon and the debate will continue on the significance of natural and anthropogenic contributions' . (2020) 2002-present National Black Chamber of Commerce, funded by ExxonMobil [125][126][127] 'There is no sound science to support the claims of Global Warming' .
Climate Change involved 'highly unrealistic carbon reduction goals' that were 'not possible' for the US to meet [28]. 'Ambitious public policies and international treaties that assume very rapid change in total energy use are simply unrealistic' and 'attempts to mandate such change are fraught with risk' . Another ExxonMobil Corp advertorial, which appeared twice in 2007, says that 'businesses, governments and NGOs are faced with a daunting task: selecting policies that balance economic growth and human development with the risks of climate change' [18,19]. These advertorials echo two of the prominent themes of 'Doubt' identified in our original analysis: (i) an alleged dichotomy between climate mitigation and poverty reduction, and (ii) the allegedly severe adverse economic impacts of mitigation [1]. A third example is a 1996 Mobil advertorial that states: '[UNsponsored climate action] is likely to cause severe economic dislocations at a time when many nations are striving for growth and jobs...World economic health will suffer as nations are forced to switch from fossil fuels, saddled with large carbon taxes and driven to prematurely scrap many factories and machinery…Jobs and livelihoods are at stake' [26].
As might be expected, the content and tone of advertorials change with time. As the scientific evidence of AGW strengthened in the early 2000s, advertorials began to include discussion of options for greenhouse gas emissions reductions, such as investment in energy efficiency and technology research and development. This is the context in which the third 'Doubt' argument we identified in our original study appears: insisting on the limitations of renewable energy [1]. A 2001 Exxon-Mobil Corp advertorial expresses a characteristic sentiment: 'Though promising, renewable energy's potential should be tempered with realism' [29]. The advertorial points out that wind power 'generally enjoys tax subsidies' , yet says nothing about the much larger subsidies that fossil fuels receive [30][31][32]. In various forms, the advertorials reinforce the presumed inevitability of continued fossil fuel dominance [33][34][35][36].

Stranded fossil fuel assets
As discussed in [1], 24 of the analyzed documents allude to the concept of stranded fossil fuel assets. Our updated analysis finds that, as before, no advertorials address the issue. Therefore, the contrast across document categories remains clear and statistically significant: the threat of stranded assets is recognized in internal and academic documents, but never mentioned in advertorials (FET: (all years) p = 3.3 × 10 −7 ; (1989-2004) p = 3.2 × 10 −6 ).

Summary of results
Our ProQuest searches described herein add 18 advertorials expressing positions on AGW (real and human-caused, serious, or solvable) to those included in our original analysis spanning 1989-2004, and 26 outside of these years (these new documents are  indicated by yellow highlights in table S4, supplementary information).
An updated analysis of the period 1989-2004 continues to yield statistically significant results, and our conclusions therefore remain unchanged: between 1989-2004, Mobil and ExxonMobil Corp advertorials overwhelmingly expressed doubt about AGW as real and human-caused, serious, and solvable. Indeed, having augmented our archive of advertorials, and with our prior document codings undisputed by ExxonMobil Corp's critiques, our original conclusions are now strengthened [2,3].
Most of these recent 'Acknowledgments' are ambiguous. As described in section 3.1.2, the vast majority (93%) are implicit: in no case does Exxon-Mobil Corp state that climate change is real and human-caused. Nor do they acknowledge a change in their position. In this sense, the acknowledgments are asymmetric compared to the doubt promoted in earlier advertorials. Earlier advertorials explicitly challenged climate science; later ones merely sidestepped it, citing undefined 'risk(s)' of climate change (87% of post-2004 advertorials) and discussing options for emissions reductions without stating why they are necessary.

Discussion
Our results imply at least three ways in which Exxon, Mobil, and ExxonMobil Corp have, variously, misled the public about AGW. Sections 4.1-4.3 address each of these in turn.

Exxon and ExxonMobil Corp misled with discrepant communications
The first way the public was misled derives from the results of our content analysis and relies on a line of reasoning presented in our original paper: comparison across company document categories. Figure 2(d) shows that from 2000 through 2004 (after the Exxon-Mobil merger), the overwhelming position of ExxonMobil Corp advertorials on AGW as real and human-caused continued to be 'Doubt' (12/16 = 75%). The discrepancy between this doubt and the predominant acknowledgment in Exxon and ExxonMobil Corp peer-reviewed, non-peerreviewed, and internal documents shown in figure  1(a) is statistically significant (FET: p = 8.5 × 10 −8 , p = 0.0079, and p = 1.6 × 10 −5 , respectively, for all peer-reviewed, non-peer-reviewed, and internal documents through 2004). From a statistical standpoint it is essentially certain that whereas Exxon and Exxon-Mobil Corp's private and academic documents predominantly acknowledge that climate change is real and human-caused, ExxonMobil Corp's advertorials disproportionally-and overwhelmingly-promote doubt on the same matter. This unambiguously reaffirms our original conclusion.
The contrast across document categories-that is, evidence of misleading communications-is also clear when analyzed at a year-to-year scale ( figure  1(a)). During the early 2000s, ExxonMobil Corp's peer-reviewed publications and advertorials in the same years contradict one another. For instance, in 2004, one peer-reviewed ExxonMobil Corp publication refers to 'the fraction of anthropogenic CO 2 emissions that remains in the atmosphere, and contributes to the radiative forcing of climate'; another presents 'cumulative CO 2 emissions' for a '550 ppm stabilization trajectory'; and a third discusses 'CO 2 disposal as an option to mitigate climate change from an enhanced greenhouse effect' [37][38][39]. Yet, that same year, one ExxonMobil Corp advertorial stressed the alleged 'debate over climate change' and fostered uncertainty that AGW had been observed, saying 'last year's record summer heat in Europe does not confirm a warming world' (climate attribution assessments have since disproved this claim [40]). They insisted that 'in the face of natural variability and complexity, the consequences of change in any single factor, for example greenhouse gases, cannot readily be isolated and prediction becomes dif-ficult… scientific uncertainties continue to limit our ability to make objective, quantitative determinations regarding the human role in recent climate change or the degree and consequences of future change' [41]. Another advertorial the same year emphasized the 'gaps and uncertainties that limit our current ability to know the extent to which humans are affecting climate and to predict future changes caused by both human and natural forces' [42].
Given these discrepancies it is clear that Exxon-Mobil Corp misled the public over this period. The historical record categorically refutes ExxonMobil Corp's recent claims that only Mobil was responsible for misleading advertorials (and for other misleading communications, as we discuss below). Misleading advertorials did not cease when Exxon and Mobil merged.
Additionally, peer-reviewed, non-peer-reviewed, and internal documents from Exxon and ExxonMobil Corp acknowledge the risks of stranded assets (24 times), whereas ExxonMobil Corp's advertorials do not (p = 3.3 × 10 −7 , FET). This imbalance has not been disputed by ExxonMobil Corp in its critiques of our original study [2,3].
The significance of these discrepancies is compounded by the imbalance in the physical and intellectual accessibility of advertorials versus other document categories. As evidenced in our original study, ExxonMobil contributed to scientific articles with likely average readerships of tens to hundreds, yet raised doubts about that science in newspapers potentially read by millions of people [1].
Non-peer-reviewed Exxon and ExxonMobil Corp documents also communicate greater doubt about AGW as real and human-caused and solvable than peer-reviewed Exxon and ExxonMobil Corp publications (and, with respect to real and human-caused positions, than Exxon and ExxonMobil Corp internal documents) (figures 1(a) and (c)). Although this discrepancy is smaller, it is statistically significant at or below p < 0.1 (FET: (real and human-caused) p = 0.044 for peer-reviewed publications and p = 0.077 for internal memos; (solvable) p = 0.0076), suggesting that Exxon and ExxonMobil Corp's non-peerreviewed communications, which tended to be more orientated towards non-scientific audiences (such as industry groups and journalists) than peer-reviewed papers, were sometimes misleading.
The non-peer-reviewed documents demonstrate that the doubt ExxonMobil Corp expressed in advertorials post-merger was not an unintentional or isolated incident: it was part of the company's broader public communications effort. As noted in our original paper, there are countless non-peer-reviewed materials beyond those included in our corpus [1]. Table 3 lists just a few examples, among them 'climate talking points' that ExxonMobil Corp distributed to reporters in 2001 as part of a press release specifically promoting their publication of two advertorials ('major ads') in the Los Angeles Times, NYT, The Wall Street Journal, and The Washington Post [43]. In step with the advertorials, the talking points question the scientific authority of the Intergovernmental Panel on Climate Change (IPCC) and the validity of the 'Hockey Stick' graph showing global warming, which was a centerpiece of the 2001 IPCC report [44].

Exxon, Mobil, and ExxonMobil Corp misled with misinforming advertorials and non-peer-reviewed publications
The second way the public was misled also derives from the results of our content analysis and relies on a line of reasoning presented in our original paper: comparison of public company communications against available scientific information.
ExxonMobil Corp has not disputed any of our original document codings, including those identifying numerous expressions of doubt-some, factual misrepresentations-about AGW (notably in Mobil and ExxonMobil Corp advertorials and Exxon and ExxonMobil Corp non-peer-reviewed publications) [2,3]. Using as proxies for mainstream climate science both the conclusions of the IPCC (our analysis filters for 'reasonable' doubt-see [1]) and the science of Exxon and ExxonMobil Corp itself (Exxon-Mobil Corp says its 'researchers recognized the developing nature of climate science at the time…[and] mirrored global understanding'), it is evident that Exxon, Mobil, and ExxonMobil Corp's public communications were inconsistent with available scientific information and therefore misled the public [45,46].

What did Mobil know?
ExxonMobil Corp's critiques of our original study imply that Mobil was oblivious to the insights and warnings of mainstream climate science, even as it ran advertorials attacking that science [2]. Yet a 1997 Mobil advertorial suggests otherwise: 'We continue to sponsor research at universities…At Columbia's Lamont-Doherty Geophysical Observatory, we supported work on the role that oceans play in the climate system' [47].
Additional documents not included in our original analysis confirm that Mobil, like Exxon, had direct access to the insights of mainstream climate science [48][49][50][51]. For example, as a 1997 report by Mobil's Anthony R. Corso  According to a newly discovered internal budget proposal, '1994 Mobil Foundation Grant Recommendations' , Mobil's funding at Columbia University included $25 000 per year in 1991 and 1992 and would continue at the same rate in 1993 and 1994 [49]. Mobil described the university's Lamont-Doherty laboratory as 'a world-wide leader in earth and atmospheric studies' and said the purpose of the grant was to 'develop an improved computer model [that] will become part of the larger models predicting the impact of increased greenhouse 2 Correct spelling is Lamont-Doherty. gas emissions on global climate' . 'Ultimately' , they noted, 'these models will be the basis for regulatory action' . 'Benefits to Mobil Foundation' included '[t]echnical information and understanding…key to Mobil's ability to participate in the debate on [potentially imminent greenhouse gas] regulations...Mobil scientists involved in the global warming issue can gain first hand understanding of the role of the oceans in global warming and develop personal relationships with some of the key experts…[P]articipating at this level is far more valuable to Mobil than merely reading papers…' .
In other words, Mobil had scientists studying AGW and learning from some of the same groups of independent climate experts as Exxon scientists. (For example, from the late 1970s through the mid-1980s, Exxon spent tens of thousands of dollars funding a 'cooperative program with Lamont-Doherty' in which scientists at Exxon and Columbia University collaboratively co-authored AGW project proposals and conducted AGW research [52][53][54][55][56][57][58][59]. ExxonMobil Corp has continued to fund the Lamont-Doherty Earth Observatory throughout most of the 2000s to present [60][61][62][63][64][65][66][67][68][69][70][71].) In turn, those Exxon scientists overwhelmingly acknowledged AGW as real and humancaused. Mobil's access to these same mainstream scientific resources preceded and paralleled its publication of advertorials attacking climate science and its implications, further demonstrating that Mobil knowingly misled the public.
Mobil was also an active member of the American Petroleum Institute (API), and numerous documents record API's early awareness of the potential AGW dangers of its products. These include APIcommissioned research on carbon dioxide at the California Institute of Technology in 1955; an in-person warning to API by physicist Edward Teller in 1959; API monitoring of warnings about AGW by President Johnson's Science Advisory Committee in 1965; and API-commissioned research on AGW at Stanford Research Institute in 1968 and 1969 [72][73][74][75].

Exxon and ExxonMobil Corp misled with additional direct and indirect climate denial
The third way the public was misled relies on an additional line of reasoning that was not explicitly discussed in our original paper: comparison of the results of our content analysis against an extensive literature of scholarly research and investigative journalism that has chronicled the company's history of directly and indirectly perpetuating climate science misinformation.
ExxonMobil Corp has not disputed our document codings, which reveal overwhelming acknowledgement by both Exxon and ExxonMobil Corp scientists that AGW is real and human-caused [2,3]. At the same time, it is well-documented (based on documents beyond those included in our analysis, as well as on some non-peer-reviewed documents included herein) that (i) from at least the 1990s until at least 2015 (and arguably to this day), Exxon and ExxonMobil Corp have sometimes publicly promoted doubt about climate science through direct company communications; and that (ii) from at least the late 1980s through to the present, Exxon and ExxonMobil Corp have funded groups and individuals and participated in organizations that cast doubt in public on climate science [27, (table 3 provides a  few examples). To our knowledge, ExxonMobil has never disputed its history of direct and indirect climate denial. Likewise, Exxon and ExxonMobil Corp have a track record of directly and indirectly promoting public doubts about AGW as serious and solvable that are inconsistent with the views of company scientists chronicled by our analysis (again, see table 3 for examples).
This comparison-between what ExxonMobil knew and its broader history of climate denial and delay-is an inherent, central line of reasoning in many journalistic and legal investigations of the company. It highlights an important point: Our work does not stand in isolation. At the onset of our study, substantial evidence already existed to suggest that ExxonMobil had misled the public on a variety of aspects of AGW and in a variety of ways [27,[77][78][79][80][81][82]. The purpose of our study was to bring to bear an additional, complementary empirical methodology to test the hypothesis that ExxonMobil misled the public. Our results show this to be the case.

Conclusion
We have updated our original analysis to include additional Mobil and ExxonMobil Corp advertorials in the NYT, and have also introduced new documents never previously analyzed in the peer-reviewed literature. Among other things, we have shown that misleading communications, direct and indirect, emanated from both Exxon and Mobil before their 1999 merger, and continued thereafter. We have also introduced new evidence that Mobil was aware of developments in mainstream climate science, even as they took out advertorials that challenged it. We now conclude with even greater confidence that Exxon, Mobil, and ExxonMobil Corp misled the public about climate change.
The history of ExxonMobil's communications about AGW is consistent with what scholars have labeled merchandising doubt, manufacturing doubt, or doubt-mongering [27,[128][129][130][131][132][133][134][135]. A party whose interests are threatened by scientific findings may seek to protect those interests by casting doubt on the science: 'emphasiz[ing] the uncertainty' , as a 1988 Exxon strategy memo put it, focusing on 'debate' , and suggesting that remedies are unavailable, unrealistic, too expensive, or otherwise undesirable [136]. Often these claims are not made outright, but are insinuations, which are harder to refute. They may also attack scientists, suggesting they are unreliable or biased. Many of these strategies are evident in Exxon-Mobil's communications, as well as in their public and private critiques of our work that we have here addressed.