How “An Inconvenient Truth” Contributed to Partisan Polarization on Climate

Joe Romm of Center for American Progress Misrepresents Polling Data


Al Gore’s “An Inconvenient Truth” did contribute to partisan polarization of public opinion on global warming, despite what Joe Romm recently claimed at Climate Progress in his critique of our New York Times oped. A wealth of Gallup and Pew polls find that partisan polarization increased significantly between 2006 and 2008. But beyond polls, Romm elides the nuances of the studies we cite, all of which arrive at the conclusion that mass communication efforts are unlikely to raise the salience of global warming and, to the degree that those same efforts increase elite polarization, they are likely to be counterproductive.

April 11, 2014 | Michael Shellenberger & Ted Nordhaus

Scroll down to read an update from the authors, written on April 14, 2014

Joe Romm of Climate Progress misrepresented polling data in his critique of our recent New York Times op-ed when he claimed Al Gore and An Inconvenient Truth did not contribute to partisan polarization of public attitudes toward global warming.

Gallup and Pew both find that partisan polarization increased significantly over the two years following the release of An Inconvenient Truth across a range of measures. Between 2006 and 2008, the split between Democrats and Republicans reporting that they personally worry a great deal or a fair amount about global warming rose from 29 percent to 41 percent. On the question of whether the effects of global warming had already begun, the split rose from 24 percent to 35 percent and on the question of whether the rise in temperatures was primarily due to human activities or natural causes, the split went from 20 percent to 31 percent. Similarly, the split over the question of whether global warming is being exaggerated in the news rose from 36 percent to 41 percent.

Pew also finds rising polarization during this period, with the split between Democrats and Republicans on the question of whether global warming is human caused rising from 26 percent to 31 percent and the split on the question of whether global warming constituted a very serious problem rising from 30 percent to 35 percent.

Beyond Pew and Gallup, many scholars have concluded that Gore’s movie did in fact drive increasing polarization. Anthony Leiserowitz, a Yale University researcher, told an interviewer this week:

"We actually studied this: What was the impact of An Inconvenient Truth? What we found is that a remarkable number of people saw the documentary and were moved by it …  But viewers also tended to be liberal Democrats with higher education, to be women, and so on… But for every person the movie engaged, there were at least as many who were turned off because they disliked Al Gore long before An Inconvenient Truth … And we still see that in our research today: Many ‘Dismissive’ — when they think of global warming, they think of Al Gore, feel intense dislike, and thus disregard his message."

By contrast, Romm references three studies, none of which conclude what he claims they do. The first combines responses to several questions about global warming asked by Gallup between 2001 and 2010. It finds, as noted above, that the split between Democrats and Republicans grew significantly between 2006 and 2008, just as we claimed.


But that’s not actually the chart that Romm posted. Instead, he posted a chart tracking the differences between self-identified liberals and self-identified conservatives. That chart shows polarization beginning earlier.


The difference illuminates an important difference, one that Romm elides rather than illuminates. The split among party elites happened earlier and drove the split among rank and file Democrats and Republicans, a dynamic that the authors conclude has been the central dynamic driving the rising polarization of the issue.

Little surprise, then, that the release of a major documentary, heavily promoted in the media, featuring a famous partisan and former Democratic presidential nominee using highly loaded language about the issue – Hurricane Katrina was like a “nature walk through the Book of Revelations” – and making strong claims about the implications for Americans – “we are going to have to change the way we live our lives” – would prefigure a rise in polarized opinion among self-identified Democrats and Republicans in the years that followed.

Romm then claims we misrepresent a 2010 study which found that “dire warming” messages decrease support for climate action because the same study finds that when such messages are paired with solutions, they increase belief in climate science. In fact, our op-ed explicitly addresses this dynamic. “What works, say environmental pollsters and researchers, is focusing on popular solutions.” And the popular solution that the 2010 study tested was of “technological ingenuity” — the very thing the two of us and Breakthrough have advocated since 2002.

Moreover, clean energy solutions are popular with the public irrespective of whether fear-based messages about climate change are offered alongside them. In this regard, climate concerns are just one among many reasons that the public supports policies designed to develop cleaner energy technologies, alongside a range of non-climate related economic, social, and environmental considerations.

Our oped also cited the work of Daniel Kahan and colleagues, which finds that the nature of the solutions offered matters. When Kahan’s team primed respondents by first having them read an article about carbon regulations as the solution to climate change, polarized views of climate science were significantly higher than the control condition, in which respondents read an unrelated article. By contrast, when respondents read articles about geoengineering or nuclear energy as key climate solutions, views were less polarized. In short, when the climate agenda is narrowly focused on carbon caps and renewable energy subsidies, and excludes solutions conservatives tend to support (eg, nuclear energy), it increases rather than decreases polarization.

Offering solutions clearly helps overcome fatalistic and skeptical responses to fear-based climate messages. But when those solutions conflict with the recipient’s underlying ideological orientation, they polarize. In our oped, we cite a Nature Climate Change study which concludes, “Broadening the debate to encompass outcomes that are related to deniers' willingness to act … may be more likely to foster the widespread consensus and support that governments need to enact effective mitigation policies.”

The third study Romm cites makes no particular claim about the relationship between An Inconvenient Truth and polarization. But it does, ironically, quite explicitly debunk Romm’s longstanding claim that more media communication linking extreme weather and natural disasters to climate change will result in policy action. Brulle and colleagues conclude:

The implication would seem to be that a mass communications effort to alter the salience of the climate change issue is unlikely to have much impact … this analysis shows that these efforts have a minor influence, and are dwarfed by the effect of the divide on environmental issues in the political elite [emphasis added].

In other words, mass communication efforts are unlikely to raise the salience of global warming and, to the degree that those same efforts increase elite polarization, they are likely to be counterproductive. This is precisely what An Inconvenient Truth, by most scholarly accounts, did.

And this is what the producer’s of Showtime’s Years of Living Dangerously, with its heavy dependence on liberal celebrities, catastrophic messaging, and dearth of solutions, appear to have entirely failed to learn.

Update April 14, 2014:

Paul Thacker and Jim Naureckas have posted further critiques of our New York Times op-ed. Both largely repeat Romm's talking points from Thursday, with Naureckas providing a fine example of how closely tied green apocalyptic messaging is to the long-standing dream of running the world entirely on renewable energy sources. 

Thacker quotes from a series of articles we have written over the last seven years about the legacy of An Inconvenient Truth, claiming we have contradicted ourselves. However, even a close reading of the quotes Thacker posted reveals that’s not the case. We have claimed, in 2007 and again in 2009, that Gore's movie did little to raise the salience of the issue among the public. This is demonstrably true. The number of Americans who believed that global warming is happening or will happen during their lifetime was 65 percent in 1998. It was 73 percent in March 2006, before An Inconvenient Truth was released, 75 percent in March 2008, and is today again 65 percent. The number of people believing that global warming will pose a serious threat to their way of life was 25 percent in 1998. By March 2006 it had risen to 35 percent. In March 2008 it was 40 percent, and today it is once again 36 percent.

Gallup has asked a number of other questions about global warming starting more recently. In 2003, the number of people telling Gallup that humans primarily caused global warming was 61 percent. In 2006, before the movie, it was 58 percent, and in 2008, after the movie, it was still 58 percent. Similarly, 65 percent of Americans told Gallup that most scientists believed global warming was occurring in 2006 and again in 2008.

Meanwhile, the issue has consistently remained a low priority for most Americans compared to other issues, even other environmental issues. That was the case before An Inconvenient Truth and has remained so since. So no, An Inconvenient Truth did not raise the salience of the issue among Americans, but it did contribute to polarization on the issue.

Thacker quotes Robert Brulle and Aaron McCright, who claim, "No reputable researcher would claim that Al Gore contributed to partisanship on climate." But in the post above, we quote a highly reputable researcher, Yale University's Anthony Leiserowitz, who claims that Gore contributed to partisanship on climate change. A range of other scholars have reached the same conclusion (see NisbetHoffman).

Even McCright, in the conclusion to his polarization study, reaches a conclusion that is difficult to reconcile with the claim that Gore's movie did not contribute to polarization: "Current flows of political messages and news concerning global warming are likely contributing to the growing divide. Americans’ political orientations moderate educational attainment and self-reported understanding in ways consistent with the predictions of the elite cues hypothesis and information-processing theory. Given the bifurcated flow of conflicting information on global warming from elites on both sides of the political spectrum, ideological, and partisan camps in the general public likely receive very different information on global warming—in ways that reinforce their existing political differences."


  • You both continue to elide the billions spent by fossil fuel interests on misinformation since 2006 — and the effective takeover of the Republican party on this issue over the same period.

    This is a grave sin of omission—and it seems actively dishonest in this context. You both know better and you’re misleading people.

    By tim dickinson on 2014 04 11

    Reply to this comment / Quote and reply

  • You’ve got to be joking. You’re basing this whole idea on one data point on the republican side in 2008 that you’d not be able to distinguish from the margin of error? What about all the other potential influences going on during that year? What about the overall trend?

    You know, you guys seem to think there’s a huge problem with “enviros” using scare tactics. The problem is, the facts ARE scary. It’s just that telling the truth about what the science is telling us is very scary.

    What I hold is that, ultimately, nearly everyone wants to do the right thing. People just have different visions of what the “right thing” is. The problem today has little to do with the left attempting to scare people with facts. It has everything to do with the fiduciary duties of fossil fuel corporations. It’s highly effective to sow seeds of doubt in the mind of the public so they cannot clearly evaluate what the right thing is.

    You have a board of directors. A CEO, CFO, VP’s, etc. Their job is to maximize returns for shareholders. Their jobs do not require them to take into account any issues that are outside the reasonable lifespan of those shareholders. In fact, the primary pressure on them is to make sure that quarterly figures on on target.

    For the fossil fuel industry it’s a no-brainer. Funneling millions into a network of groups trying to misinform the public about the crisis has an extremely high ROI (re: Merchant’s of Doubt). This has been shown to help the tobacco industry and others as well. It’s not a matter of truth. It’s a matter of making sure to maximize returns over the near term for large corporations. That’s their job.

    So, with any of these charts, how can you possibly assume that it’s Gore’s AIT that is the influence rather than the millions going into misinformation every year? The answer is, you can’t.

    It is completely irrational to pin on one man (and his movie) the effect you’re ascribing. Informing people does not cause polarization (per our twitter exchange with Dr Hayhoe). The dual tracks of having some people informing and other people misinforming DOES cause polarization. And all of that is taking place within a much broader landscape of other issues, like abortion and gay marriage and taxation and healthcare and wealth distribution and…

    By Rob Honeycutt on 2014 04 11

    Reply to this comment / Quote and reply

  • I really struggle to understand how the Breakthrough Institute have so mis-read this programme.

    I read the New York Times op-ed earlier in the week. But was surprised to see the first episode freely on-line, so I watched it.

    Instead of this deeply polarising show that the op-ed led me to expect, I got the opposite.

    Notably the show features one climate scientist talking-head, and it is an open evangelical Christian. I mean, how climate scientists on the planet are Evangelical Christians? I can probably count them on hand, or perhaps one finger.

    So, instead of polarisation, the show goes out of its way to accommodate religious people, with Katherine Hayhoe telling us that there is no conflict between science and religion. This view is one that I, and many scientists, view as ridiculous. Yet it is put forward unchallenged. Is this polarisation?

    A sop to the religious right as far as I can see, but the absolute opposite to what Shellenberger and Nordhaus led us to believe in their op-ed.

    This whole thing leaves me perplexed. The approach taken by the first episode seems to be one that the Breakthrough Institute would normally approve of. Yet you have attacked the show for being culturally polarising. These feelings are not mine alone. I have spoken to a few other secularist friends who were deeply annoyed by how the show went out of its way to play nice with religious people. Politically this may be OK, and an approach the BTI normally approves of, but it is rather unsatisfying intellectually.

    The criticism then of the show for being culturally polarising does not seem to have any legs to stand on.

    The Breakthrough Institute is often criticised for being in “punch a hippy” mode. Normally I find this to be deeply unfair, and lazy criticism, but in this case I find it difficult to disagree. Sadly.

    By Robert Wilson on 2014 04 11

    Reply to this comment / Quote and reply

  • Robert…  I can tell you exactly why BTI took this stance on the Years Living Dangerously. It’s solely because Joe Romm is a key advisor for the series.

    That’s it. They’re letting their petty grievances get in the way of their stated goals.

    By Rob Honeycutt on 2014 04 11

    Reply to this comment / Quote and reply

  • I agree with some stuff here but I think this guy got some good critics:

    By Evets Lefurac on 2014 04 11

    Reply to this comment / Quote and reply

  • Tim Dickinson and Rob Honeycutt: We don’t elide the issue. We were quite clear in our op-ed that fossil fuel interests and conservatives had played a role. Substantial funds have been spent on both sides to communicate about the issue. Substantially more by climate advocates than opponents if you are talking about direct and explicit communications about climate change.

    It is the two of you, not us who want to hang polarization on a single data point - skeptical communication by fossil fuel interests. Of course Gore and his movie are not entirely responsible for polarization on the issue. We have never suggested that. But he clearly played a role, in a context in which polarization has been rising across the board. So in that context, and knowing that we have multiple data points - polling, social psychology, cognitive science, media content analysis - all pointing to the movie, the messenger, and the message being polarizing, why would you double down on that strategy and expect a different result? Only if you thought you were simply going to vanquish the political other, which you are not going to do on this issue.

    By Ted Nordhaus on 2014 04 11

    Reply to this comment / Quote and reply

    • Substantially more by climate advocates than opponents if you are talking about direct and explicit communications about climate change.

      Really? How do you come about that conclusion when one side uses open reporting of funding, while the other uses a shell game of 501(c)(3) organizations for funding?

      ...skeptical communication by fossil fuel interests.

      It’s not skepticism that is being utilized by fossil fuel interests. It’s misinformation. Try this, Ted. Go to the Idso’s website called Go to their subject index and start reading their synopses of various research papers. They have methodically gone through and created an alternate universe on climate. They essentially go through and re-write conclusions to 100’s of research papers so that they fit what they want people to believe.

      Of course Gore and his movie are not entirely responsible for polarization on the issue.

      From the data you’ve provided, there is no way that you can possibly infer with any realistic confidence that Gore had anything at all to do with polarization. Based on the margin of error you can’t even know if there was any polarization from the conservative side.

      I think you guys should take a trip over to Dr Reich’s office at Berkeley. The polarization you’re talking about, I believe, and I think Dr Reich would agree, is taking place not because of how any particular issue is playing out. It’s happening due to income inequality.

      If we saw polarization only on the climate change issue, then I might actually agree with you guys. But it’s not. It’s occurring across the board with nearly every issue.

      By Rob Honeycutt on 2014 04 11

      Reply to this comment / Quote and reply

      • Actually Rob, I’d remind you a) that skeptic groups are not alone in not disclosing their donors. Most of the big green groups don’t either and b) you can pull the 990 on any 501-C3 and see what the budget is and what it is spent on. So the source of the money may be secret, but the amount raised and spent is not. Take your list of skeptic organizations, pull their 990’s, and you can figure out how much in total they are spending. If it’s CO2 Science, you can pretty well assume its all being spent on climate related communications. If its the Chamber of Commerce, very little of it is. In any event, if you take your list of organizations that are directly and explicitly disinforming about climate science and you add up every dollar you can conceivably spend on those activities, it adds up to not much more than a mole hill. Compare that with the $300 million that Gore raised and spent on his various post-AIT climate communication campaigns alone and there is no comparison. Or the year in year out, well funded communication strategies of all the big green groups on climate. That doesn’t mean that climate denial efforts haven’t been effective. But it does suggest to quote Joe Romm’s house sociologist, that “mass communications effort to alter the salience of the climate change issue is unlikely to have much impact.”

        As far as Gore’s impact on polarization on the issue, it is not a single date point, nor is it within the margin of error. Both Pew and Gallup are large sample national surveys. We are seeing increases in polarized response to multiple items over the two years that spanned the release of the movie and Gore’s Nobel and Oscar of over 10%. Nor do we rely simply on polling. We have well established social psych, cognitive science, and sociological theory. We have multiple controlled psychological experiments. We have extensive cognitive analysis and qualitative research. All of it points to the same conclusion. You have to be really committed to your position not to connect all those dots.

        By Ted Nordhaus on 2014 04 11

        Reply to this comment / Quote and reply

    • about as clear as footnote 9

      By Eli Rabett on 2014 04 11

      Reply to this comment / Quote and reply

  • Robert Wilson: I’m glad you liked the first episode. I imagine many liberals and environmentalists will. But I’d remind you that in that first episode, the show basically attributed the civil war in Syria and the closing of a meatpacking plant in Texas to global warming and took a laid off worker from the plant and made her drive six hours to be lectured by a climate scientist all without so much as offering a single solution.

    Nelly’s conversion told the audience exactly what they wanted to here - keep moralizing, with a little evangelical gospel thrown in for good measure, and those ignorant bible thumping texans will see the light. One of the show’s producer has acknowledged that they aren’t going to talk alot about solutions because it doesn’t make for good television. I’d be willing to bet that any discussion of solutions will focus entirely on solar, wind, and energy efficiency.

    I don’t know that any of this will actually alienate key audiences, because I’m not sure any are watching. This is the network after all, that brought us Weeds and Californication. But I’m pretty certain that it will only reinforce the worst instincts of its upscale liberal audience, and that does not bode well for a less polarized climate discourse anytime soon.

    By Ted Nordhaus on 2014 04 11

    Reply to this comment / Quote and reply

  • This post implicitly claims that Leiserowitz found that AIT did more harm than good on climate awareness. Not only does this remains unproven (there has been a great deal of public interest in it, followed by some gain in awareness) but in fact the polarization did occur mostly because it was Al Gore (former Vice-President, Democrat, Presidential Candidate, etc.) who was associated to it rather than because of the content or the approach of the documentary.

    By Evets Lefurac on 2014 04 11

    Reply to this comment / Quote and reply

  • Though I agree that the fear-based approach used in the Years docu here has to evolve.

    By Evets Lefurac on 2014 04 11

    Reply to this comment / Quote and reply

  • Public service marketing has shown many times that fear-based marketing is not very effective except in preaching to the choir. We’ve seen this with smoking, drugs, seat belts and numerous other issues. Climate change advocacy has some special problems, because you have powerful groups fighting it.

    That said, AIT was an excellent documentary, but one which impressed mostly people who were already receptive to the idea of anthropogenic global warming. Influencing public opinion among those who are neutral or skeptical about the reality of AGW is going to be a difficult task.

    By Robert Kelley on 2014 04 11

    Reply to this comment / Quote and reply

  • The op-ed’s claim that the “scholarly consensus” is against the use of threatening messages is wrong. The most recent and most methodologically rigorous meta-analysis on fear appeals would suggest that a focus on energy solutions, raising the audience’s sense of efficacy, will only work if threatening climate information is also included.

    To quote the authors,“efficacy only has an effect is threat is high”.

    See here:

    By Neil Stenhouse on 2014 04 11

    Reply to this comment / Quote and reply

  • Oops - quote should be “efficacy only has an effect IF threat is high”.

    By Neil Stenhouse on 2014 04 11

    Reply to this comment / Quote and reply

  • Thanks for posting this Neil but I don’t think you got the conclusion quite right. The main conclusion of the analysis appears to be that threatening communications backfire unless efficacy is high. The authors have little to say regarding the prospects for raising efficacy without threatening communications, perhaps because their main focus was on examining under what conditions threatening communications were effective. It needs to be noted also that this analysis looks at public health campaigns targeting behavior change. Climate change is a very different animal. I post the key section below:

    First, intervention developers should employ utmost caution when using severity- based threatening communication (note that emotionally neutral, susceptibility- based information appears less dangerous and equally effective; de Hoog et al., 2007). Threatening communication only works when either the baseline efficacy is high in the relevant behaviour-population combination, or when the intervention also includes a potent efficacy-enhancing element. An overview of effective methods to enhance efficacy is provided on pages 342

    By Ted Nordhaus on 2014 04 11

    Reply to this comment / Quote and reply

  • Rest of section here: Note that to enhance efficacy, a simple recommendation (e.g., the recommendation to quit smoking, to call a phone number, to moderate drinking) is insufficient. The efficacy manipulations reported in the studies included in this meta-analysis entailed at least a paragraph of text, and in fact, using active rather than passive methods is recommended (Albarrac ́ın et al., 2005). Mere recommendations at best enhance response efficacy, but without high self-efficacy, threatening communication will have no effect, or worse, backfire. Given this danger of backfiring, developing a behaviour change intervention that uses threatening elements is ineffective and unethical unless pilot tests indicate that the intervention also reliably enhances response and self- efficacy. In general, the evidence indicates that intervention developers should look to the theory for guidance, as theory-based interventions have been found to be more effective than non-theory-based interventions (Peters, Kok, Ten Dam, Buijs, & Paulussen, 2009; Webb et al., 2010). When restricted to mass media, it will probably be wisest to resort to a behaviour change method that does not involve emphasising negative consequences of a behaviour, and if that cannot be avoided, at least make sure the communication is not threatening, emotional or confronting (see also de Hoog et al., 2007).

    By Ted Nordhaus on 2014 04 11

    Reply to this comment / Quote and reply

  • “...threatening, emotional or confronting.”

    Are there three other words that better describe Showtime “scientist” Joe Romm?

    By Jean Singer on 2014 04 11

    Reply to this comment / Quote and reply

  • Right. Threats backfire unless efficacy is high. I completely agree that raising efficacy and a positive vision of energy solutions is massively important, probably the single most important thing the advocacy community needs to do more of.

    My concern was just that people might get the impression they should ignore threat information altogether, when a lot of literature suggests it is better to include threat *and* efficacy.

    You’re right that the meta-analysis was health messages which is very different to climate. I’ve found one experimental study that tested threat-only, efficacy-only and threat-and-efficacy messages for climate action. The threat-and-efficacy message worked best. Will post that in a sec when I find it.

    By Neil Stenhouse on 2014 04 11

    Reply to this comment / Quote and reply

  • Here it is:

    van Zomeren, M., Spears, R., & Leach, C. W. (2010). Experimental evidence for a dual pathway model analysis of coping with the climate crisis. Journal of Environmental Psychology, 30(4), 339-346.

    By Neil Stenhouse on 2014 04 11

    Reply to this comment / Quote and reply

Submit a comment

About Michael Shellenberger & Ted Nordhaus

Ted Nordhaus and Michael Shellenberger are leading global thinkers on energy, climate, security, human development, and politics. They are founders of the Breakthrough Institute and executive editors of Breakthrough Journal.

Click here to view their recent articles.