Solving the climate change attitude mystery

Brandon Keim writes,

Over the last year and a half, the number of Americans who believe the Earth is warming has dropped. The decline is especially precipitous among Republicans: in January 2007, 62 percent accepted global warming, compared to just 49 percent now. . . . The confounding part: among college-educated poll respondents, 19 percent of Republicans believe that human activities are causing global warming, compared to 75 percent of Democrats. But take that college education away and Republican believers rise to 31 percent while Democrats drop to 52 percent.

That strikes me [Keim] as deeply weird. I don’t even have a snarky quip, much less an explanation.

This does seem a bit weird: you might think that college grads are more likely to go with the scientific consensus on global warming, or you might think that college grads would be more skeptical, but it seems funny that it would go one way for Democrats and the other for Republicans.

Things become clearer when I looked at the graph (which was thoughtfully presented next to Keim’s article):

pewclimateducation_2.jpg

Among college grads, there is a big partisan divide between Democrats and Republicans. Among non-graduates, the differences are smaller. This is completely consistent with research that shows that people with more education are on average more politically polarized (see, for example, figure 9a of my paper with Delia). Basically, higher educated Democrats are more partisan Democrats, and higher educated Republicans are more partisan Republicans. On average, educated people are more tuned in to politics and more likely to align their views with their political attitudes.

From this perspective, it’s really not about the scientific community at all, it’s just a special case of the general phenomenon of elites being more politically polarized (a phenomenon that we discuss in chapter 8 of our forthcoming book, and which is related to divisions between red and blue states).

P.S. I followed the link from Andrew Sullivan. And here’s the detailed Pew report (and, remember, Pew gives out raw data!).

34 thoughts on “Solving the climate change attitude mystery

  1. I don't think the graph above does show "more polarization"; I think that's an interpretation that's too kind to Republicans who went to college.

    The bar graph format makes it difficult to see what the line graph alternative makes clear: that everyone is more likely to believe that global warming is anthropocentric if they got an education, except Republicans. I don't call that polarization, I call that "did Republicans really get into college on their academic merits, and did they really study while there?"

    See the President for more evidence that rich Republicans go to college more because of their daddy's oil money than because of their smarts. In Britain of course, we've been familiar with this phenomenon for centuries :-)

  2. Alternative hypothesis: perhaps the science that humans drive climate change is not settled, and Republicans are more aware of that fact because of talk radio and their alternative media sources.

    Before you call anthropogenic global warming settled, please show me the data that shows that the same factors that created the numerous cycles of climate change before humans invented the internal combustion engine are no longer at work. Call me naive, but I think the sun's output has cycles. Occam's razor at work.

    Cars may increase the warming at the margin, but the climate system is so complex, noisy, and poorly understood that we cannot quantify the magnitude of the human effect with any degree of precision.

    Have any of the statisticians out there reviewed the guts of the these climate change models? Are they pasting together alternate sources of data responsibly? Do they even share their models or are they black boxes?

    As a libertarian I can see this is a fun topic because we get to call Republican stupid, but I get very skeptical when the strongest argument for a scientific theory is that it is 'already settled.' Science is not democracy where the majority wins. All it takes is one scientist to disprove the most popular theory.

    There are numerous examples in the history of science of the consensus being wrong. My favorite is the scientists that went against the grain and demonstrated that most ulcers are not caused by stress, but by a bacterium. It was very difficult for them to get their work published because the science was 'settled.' Even when the data supported their theory, their cure rate was markedly higher than the alternate hypothesis, and they had a great p-value (ha ha), they still had an uphill battle.

    I would more interested in an in-depth discussion of the statistical challenges of climate modeling than in the political angles of the question.

  3. Sean,

    I agree that media sources have alot to do with this result. But on the accuracy of the ones republicans rely on, how do I say this? You're just wrong. A Swedish chemist called Svante Arrhenius figured out the physics of CO2 climate forcing, and then calculated, all other things being equal the effect of rising CO2 on the earths temperature using a simple model. He came quite close, on the mean, to what has actually happened. The year? 1896.

    Solar forcings are not as large as CO2 forcings, we understand quite well the major influences on Climate in the past several million years. The key here is timescale, the current RATE of rise of CO2, CH4, NO2, is simply unprecedented in Earths history and we don't know how the rest of the complicated climate system will respond. We do know however the physics of the greenhouse forcing vs the solar forcings

  4. Alternative alternative hypothesis:

    Republicans are less likely to _care_ about global warming and/or are less willing to challenge the current party elites on the issue. It's too easy to say that Republicans are dumb. And, Sean, I would guess that readers of this blog know better than to think that your post presents anything but a "political angle." Have data and models been reviewed and presented? C'mon.

  5. Question is really why the difference. Why is does education make republicans more sure of the negative and democrats more sure of the positive. The proposed explanation is feasible.
    My take on why the explanation is true is:

    College educated people are more aware of their party line. This coupled with the fact that these are people that have choosen to reject the moniker "indepedent" leads to them saying essentially what they are supposed to say.

    On a side note which has nothing to do with the study, I also think republicans from uneducated to educate have a tendency to prefer the ancedotal and are extremely suspicious of statistical inference.

    True story I once had an argument that went like this:
    Me: African americans have almost twice the unemployment rate as whites in the United States. This is evidence of a continued issue with race.

    Republican: There is highly successful African American on the radio who says there is no longer a race issue in america, and he should know! (a story emphasising individual power over their destiny)

    I don't even think these two pieces of evidence belong in the same league. One is a fact measured against the entire US population. The other is an ancedote.

    Yet by and large without a study to back up my claim (there is some irony here), I swear to you that the republicans I know when confronted with the above statistics about the population versus an ancedote emphasising the individual will give more weight to the ancedote.

    The point of this is I think present day republicans are just conditioned to be suspicious of statistics because of some prolonged arguments in the US in which they have had to buck population measures in order to emphasize the individual. I am not saying this directly manifest here; however, because they have essentially had to reject studies in the social science setting, it is almost impossible to convince republicans I meet of anything via a p-value.

    The hypothesis from this perspective is that it may not be possible to get rank and file republicans on board global warming via superior statistical analysis. You may be able to get democrats to drop it, however, by driving home the point that the math isn't there.

  6. "did Republicans really get into college on their academic merits, and did they really study while there?"

    Who is more likely to get into college or professional school with lower qualifications through affirmative action: Republicans or Democrats?

  7. I'd also love to see an in-depth discussion of climate modeling that wasn't dominated by the "denialist" cries of the true believers. At the moment, my attitude is basically "New crisis, same old stuff." It's the "Population Bomb" or "Limits to Growth" furors revisited.

  8. Shall we have physicists vote on the existence of dark matter? Consensus has nothing to do with fundamental science.

    A major problem with the GCM models that predict AGW is their treatment of the cloud physics. The water vapor amplifier is the major influence on surface temperature, and not the direct IR absorption by CO2, which is actually pretty small compared to the water vapor. More water vapor means more clouds, and that could cause negative feedback because of the increased reflectivity of the incoming sunlight. The GCMs can't really can't model the cloud physics properly because it occurs on a scale that's much smaller than the resolution cells in the GCMs. As such the GCM parameters are adjusted to make predictions closer observed data. We don't have classic science here were a theory is tested against observation. There is plenty of room to reasoned, informed and intelligent skepticism.

  9. You're just wrong. A Swedish chemist called Svante Arrhenius figured out the physics of CO2 climate forcing, and then calculated, all other things being equal the effect of rising CO2 on the earths temperature using a simple model. He came quite close, on the mean, to what has actually happened. The year? 1896.

    Your implication that the physics of warming from CO2 is so simple that it was known in the 19th century, and hence skeptics are over 100 years behind the times, is incorrect. After a simple google search, I found the following, at http://www.aip.org/history/climate/simple.htm#L_A

    Arrhenius's model planet was mostly static. He deliberately left aside factors he could not calculate, such as the way cloudiness might change over the real Earth when the temperature rose. He left aside the huge quantities of heat carried from the tropics to the poles by atmospheric movements and ocean currents, which also might well change when the climate changed. Most important, he left aside the way updrafts would carry heat from a warmer surface into the upper atmosphere. In 1963, when a scientist made a calculation using roughly comparable assumptions, but with the aid of improved data on the absorption of radiation and an electronic computer, he found a far greater greenhouse warming — indeed impossibly greater. The assumptions left out too much that was necessary to get a valid answer.

    So any match between Arrhenius's result and the currently accepted result is pure coincidence.

  10. A. Zarkov at May 15:

    Ah, yes, the "water vapor" argument. Intelligent rhetoric. See the 6 April 2005 post on the issue (and exhaustive discussion) at realclimate.org.

  11. To all: The divide between Republicans and Democrats isn't a surprise, given the politicization of this issue and its connection to other environmental issues. (To put it another way, I could imagine a world in which Republicans were worried about climate change and Democrats weren't–it doesn't seem like a core issue in an ideological sense–but given the current arrangement of political attitudes, the difference between the parties is not a big surprise.)

    Also, the fact that the Dem-Rep gap is bigger among highly educated people is consistent with everything we know about more educated people being more ideological and more likely to connect parties with issues.

  12. "Ah, yes, the "water vapor" argument. Intelligent rhetoric. See the 6 April 2005 post on the issue (and exhaustive discussion) at realclimate.org."

    If you actually read the realclimate post and mine you will see that we say the same thing: water vapor is not a forcing, it's a feedback. Water vapor amplifies the warming caused by CO2. But what realclimate does not discuss there is the additional feedback of water vapor to cloud cover, or the problems with the cloud physics Moreover Freeman Dyson makes the same argument I do only in more detail.

    A critical comment is fine, but please take the time read carefully and understand what I wrote.

  13. A. Zarkov:

    Like this blog's host I am not a climatologist, but I believe what the realclimate.org people are saying (and the papers they cite) is that the magnitude of the effect you mention is not shown to have enough influence to cast doubt on the magnitude of anthropogenic global warming. But climate blogs are the place for that discussion.

    The point relevant to this thread is that many Republicans and Libertarians (esp. the college-educated ones, apparently) are more willing to toe the line of the party elites and make misstatements (or lie outright) about the science to do so. Sure there are democrats out there who do the same, but it seems to a lesser degree than those in parties to the right.

  14. I will limit my comments to the political angle here because the science discussion has devolved into two sides talking past each other. I regret that I even started it, but I do get worked up when someone writes that the science is settled. There are respected scientists on both sides of this debate.

    I don't think the political differences due to education level and party affiliation are all that mysterious.

    Two assumptions:
    1) The more educated the person, the more partisan they are likely to be
    2) The more educated a person, the more likely they are to think ahead and see the results of their positions. Of course, they may still be wrong, but at the margin they will try to think ahead more than the less educated.

    Combine these ideas and the Republicans see that buying into global warming will accrue more power to the government (higher taxes & more regulation), which they are against so they seek out scientific data to fit with this bias. Educated Democrats also see the endgame of a stronger government and weaker private sector, so they accept the scientific 'consensus' as is. For some people, this may function on a subliminal level. There is variation in humans so I am not declaring some universal law.

    If you know the result you want (i.e. bigger or smaller gov't), and you see where the discussion is going, then is it worth the time to dig into the details? Confirmation bias at work across the political spectrum.

  15. I might be being hopelessly naive, but I don't think that people base their decisions on the extent of anthropogenic global warming around implications towards the strength of government.

  16. Well, at least we're all agreed about education and partisanship. But in your post, Sean, you reveal a bias toward the party line but not reality. Republicans in recent years have shown themselves to be less concerned with "big government" than who reaps the benefits – which they cover by ranting about how they are the party of "small government." The climate issue is a good example of how party makeup (i.e. who holds power in the party) influences the discussion. On this issue Democrats don't want "more" government, they want the existing one change the way it operates – in turn shifting policy away from the favor of its long-established beneficiaries.

    And with regard to climate science – the fact is that the more that AGW skeptics question the science, the stronger the science is against them. So, in realization of this fact, skeptics resort to rhetorical techniques like the ones found in this thread.

  17. … skeptics resort to rhetorical techniques like the ones found in this thread.

    This is a striking illustration of the confirmation bias mentioned by sean. Just look at the first comment in this thread: "See the President for more evidence that rich Republicans go to college more because of their daddy's oil money than because of their smarts". With some exceptions, the pro-AGW comments don't get much better after that. But the poster above thinks that (presumably disreputable) "rhetorical techniques" were employed by the other side in this thread?!

    Any explanation for attitudes regarding AGW has to start with acknowledging that no more than 1% of the people with opinions have the ability to actually judge the scientific issues, and then only with considerable investment of time. Maybe a few percent more can sensibly judge the credability of the scientists involved, again with considerable time investment. The rest are just adopting the views of people who for some reason they consider authorities. Who you consider an authority often depends on your political ideology.

  18. anon:

    The April 6, 2006 RealClimate post you cite does not discuss problems with cloud physics, nor does it cite any literature that deals with this issue. One commenter brought up the issue of negative feedback from increased cloud cover. Another commenter said it was not significant, but he gave no cites. Go do a search on the word "cloud."

    I refer you to "A Climate Modelling Primer" by McGuffie (2005). This is a book written by practicing climate modelers, and it provides a nice introduction to the subject at a mature label. The authors discuss the problems with negative feedback from increased water vapor. They also discuss the implications of the various work arounds the modelers use: a large uncertainty in the climate sensitivity factor. The large uncertainty interval for this factor has remained essentially unchanged for the last 30 years indicting little or no progress in modeling all the feedbacks. To quote McGuffie, "The role of clouds in climate prediction remains on of the dominant sources of uncertainty. "

    A word on the culture of physicists. The best physicists are not team players. They are critical and ever skeptical. Any important theory gets a real going over and your feelings won't be spared. That's just the way the game is played. Until AGW. Now critics are guilty of a thought crime for being skeptical of a modeling enterprise full of flaws. "Oh you don't believe in global warming, you must be one of those dumb Republicans." I can tell you from up close experience that in some shops public criticism of AGW can be a career killer. Anything that's a threat to the budget is not tolerated.

  19. I think it less about ideology than about change. Liberals are more open to opinion, argument, and evidence. The weakness would be they are easily swayed. Conservatives are more closed and skeptical of anything other than what reinforces their own prior beliefs. The weakness would be they fall for anything that reinforces those beliefs uncritically. If this is true, belief in intelligent design should also be higher among educated conservatives.

  20. It would be interesting to see this data broken down by college major. It's not just that you went to college but what you studied there that matters.

  21. A. Zarkov,

    Yes, models have uncertainty in cloud feedback, but even within the bounds of different cloud parameterizations, you don't get very low climate sensitivities. You generally get sensitivities between 2 K and 5 K, which is pretty much the IPCC standard 3 +/- 1.5 K figure.

    Furthermore and perhaps more robustly, you can get at climate sensitivity from an observational perspective, instead of trying to calculate all of the individual feedbacks from first principles in a big AOGCM simulation. You can measure the radiative forcings (GHGs, solar irradiance, aerosols, dust, etc.) and the observed climate response, and calculate how large a total feedback you need for the observed forcing to produce the observed response. (Of course you must also take into account the uncertainty in forcing and response.) That approach can't tell you how much of that feedback is due to water vapor or clouds, or even the signs of individual feedbacks, but it lets you estimate their overall sum. And it too gives results which pretty much overlap the canonical IPCC range.

    Speaking as a physicist, whatever point you're trying to make about the skepticism of physicists seems misguided. You say that physicists are traditionally "critical and skeptical", "until AGW". Are you implying that climate modelers, who are also physicists, are not critical or skeptical of their models? It is not hard to find climate modeling papers which discuss the limitations of climate models. In fact, you cite a well known textbook written by climate modelers to buttress your point that climate models have limitations. (Or were you comparing skeptical physicists to uncritical non-scientists? But then why do you say that the skepticism of physicists has "changed" with the advent of AGW?)

    Anyway, back to your discernible point that models have flaws: it is not news that models aren't perfect. No model, in any field, is perfect. That doesn't mean that they don't have anything useful to say. (Box's famous quote comes to mind.)

    No one has succeeded in substantially reducing the uncertainty in climate sensitivity coming from climate models past what was known circa 1980. (See, e.g., last year's Science comment by Roe and Baker for physical reasons why this may be the case.) And observationally the range has not yet been reduced much, although it is only in recent years that people have started considering observational constraints other than global surface temperature.

    However, that doesn't mean that we can't say anything useful about climate sensitivity. We can say, for instance, that both modeled climate physics and observational data heavily disfavor an unamplified climate sensitivity of ~1 K, which is what one needs to claim that AGW is less significant than is concluded by the IPCC. How significant it is (e.g., a climate sensitivity of 2 K vs. 4 K) is still open to question, which is why there are large error bars on the projections.

  22. Sean,

    "Before you call anthropogenic global warming settled, please show me the data that shows that the same factors that created the numerous cycles of climate change before humans invented the internal combustion engine are no longer at work."

    No one has claimed that they are not at work, just that they alone do not adequately explain the recent warming.

    "Call me naive, but I think the sun's output has cycles."

    It does, but the 20th/21st century trend in solar output is very small (as Cameron noted), and in fact slightly decreased during the last few decades during which the greatest warming has been experienced. Solar output may have contributed noticeably to some of the early 20th century warming, but it is extremely difficult to explain the late 20th century warming using solar output as a major forcing.

    Speaking of political bias, the wording of your comment implies that climatologists are unaware of the existence of solar cycles, or believe that they no longer exist. That seems a curious allegation to make. Skepticism founded on legitimate physical arguments is one thing (see the discussion of cloud feedbacks above), but skepticism founded on the belief that climatologists don't know (or care) about natural drivers of climate change … sounds like bias to me.

  23. Ah yes; the "I must investigate what weird pathology could possibly cause you to disagree with me" meme.

    I'm never quite sure whether to find it humorous, disgusting, or both.

  24. I'd like to offer a hypothesis for consideration: do college kids pursue all the various possible major courses of study (regardless of party affiliation) in the same proportion?

    I ask this because not all graduates come out equally equipped to gauge the magnitude & nature of controversy among the ranks of scientists. (Is the dissent more akin to those who demanded more evidence from Cold Fusion proponents, or is the dissent more like that which led to refinements of theories that are still true in-the-large: relativism versus newtonian mechanics, for example. Or modern evolutionary biology versus Darwin's first-order theory?)

    I would not expect students of business, finance, accounting, public-relations, etc. to be as equipped in this regard as students of physics, ecology, psychology, computer-science (etc).

    So could it be that republican students favor profit-driven courses of study that much more?

  25. I think it less about ideology than about change. Liberals are more open to opinion, argument, and evidence. The weakness would be they are easily swayed. Conservatives are more closed and skeptical of anything other than what reinforces their own prior beliefs. The weakness would be they fall for anything that reinforces those beliefs uncritically. If this is true, belief in intelligent design should also be higher among educated conservatives.

    I agree with your description here on climate change and intelligent design, but I think if we changed the topic the roles would reverse.

    For example, I know conservatives who used to accept the "consensus" that Joe McCarthy was a bad guy, but then read Ann Coulter and now think otherwise. And who used to think disparities in black/white outcomes were due to social factors, legacy of racism, etc., but now think it could be genetic.

    In contrast, no matter how much "evidence" (and I'm putting that in quotes because obviously Ann Coulter isn't the same as IPCC) they read on these issues, I think many liberals would continue to be "deniers." And certainly a large portion of that resistance I think would come from the political implications of those views, namely that it might encourage more witch hunts and less support for programs to help black Americans.

    Just to reiterate, I don't endorse the two books I linked to. I'm just saying I don't think it's true that conservatives won't change their mind after reading something, while liberals will. It depends on the topic.

    Also, to anticipate another possible objection, I know one might say, "Oh give me a break, conservatives are leaning towards Commie witch hunts and racism anyway, and these 'thinkers' just gave specific form to their prejudices."

    Right, but I think a lot of liberals are predisposed to think the excessive consumerism is bad and that humans are hurting the planet, and so are predisposed to accept the IPCC reports.

  26. Could it possibly be that college educated Republicans have experienced first hand the tendency common to professional academics of using impressive looking tools to dress up their opinions as fact?

    I know of a history professor at a top university who teaches his students that FDR knowingly permitted the Japanese to bomb Pearl Harbor. If someone is a liberal, they may be perfectly inclined to accept such an argument. But think of the Republicans who are in this class: These students know that the professor is incorrect and may even attempt to voice their concerns. Who will win the ensuing argument? Most certainly the professor – despite the fact that history is not actually on his side.

    Experiences such as these will make it far easier for Republicans to be skeptical of – or even to dismiss – academic findings.

  27. "You can measure the radiative forcings (GHGs, solar irradiance, aerosols, dust, etc.) and the observed climate response, and calculate how large a total feedback you need for the observed forcing to produce the observed response"

    Huh? Please specify exactly what you are talking about. I assume you are basically talking about a dynamical model or time series model of the form

    y(t) = alpha*f(t) + e(t)

    where y(t) is some average surface temp measure in Kelvin, alpha is the sensitivity in K/(Wm^-2), f(t) is the total forcing in W/m-2 and e(t) is the residual. t is discrete and in years. How good is the fit for a model like this? What about a model which exponentially averages the forcing over the last 1000 years.

  28. assman,

    You describe a simple linear model. The simplest model one should consider using is an energy balance model (EBM),

    dH/dt = C dT/dt = F – lambda T

    where H is the heat content of the Earth system, F is the radiative forcing, C is the heat capacity, and lambda is a feedback parameter. That will give a response which, under constant forcing, exponentially approaches a new equilibrium temperature with a characteristic response time related to the heat capacity of the system. You can think of it as a smoothed version of your model (because of the response time).

    In reality, at a minimum you would really want to use an energy balance atmosphere coupled to a 1D upwelling-diffusion ocean. That's to try to represent the ocean response more correctly, since it has such a large influence on the response rate. (If there's a high climate sensitivity, it can be masked in the observations by a slowly responding ocean.)

    In recent years people have used Earth system models of intermediate complexity, which are 2D/3D models which can sometimes do real circulation, but have the feedbacks in a highly parameterized form suitable for estimation.

    The fit from an EBM is decent for global average surface temperatures (which is about all they're good for, plus maybe ocean heat). They don't match the decadal ups and downs, but the main point of an EBM is mostly to explain the total warming between pre-industrial times and now. They're not hugely better than your instantaneous linear response model, but because they have the feedback and response time parameters in there in a physical way, they can give you a first cut ballpark estimate of those values. Basically, they just keep track of how much heat you're adding to the system and throw in an amplifying factor and a lag time to make it a little more physical.

    See Figure 10 in this document. It's about 10 years old but gives you a rough idea of what EBMs can do.

    I don't know what you mean by "exponentially averaging the forcing". As for the last 1000 years, there is a longer time period for observational (proxy) data, but we are less certain of both the forcing and the response. You still get something roughly around the IPCC canonical range … modern instrumental observations, millennial scale proxy data, glacial cycle and deep time proxy data, and GCM physics all tend to hover around the 1.5-4.5 K range for climate sensitivity to 2xCO2. (For reference, no feedback gives about 1.1-1.2 K, so this implies a net positive feedback, possibly substantial.)

  29. Liberals are not driven by belief but by truth. Does a problem exist? What can be done about it? What implications does it carry? Scientists are their natural authority figures.

    They are more likely open to changing their opinions and conservatives realize this which is why they encourage argument, denial, and doubt. But liberals also see the biased position conservatives start from which undermines their arguments. They realize conservatives are driven by belief, not evidence, so such arguments become the tedious disputations of the uninformed. The good news for conservatives is they only have to convince climatologists. The bad news is they are not in a position to do so. The bad news for liberals is it is unlikely any argument will be successful. About the best liberals can do is appeal to the authorities of conservatives such as religious figures.

    While field of study is not irrelevant, it is also probably not that significant. Most of what people pick up in college is not beliefs or facts but exposure to other opinions and tools of argument and reasoning which are then applied to their worldview.

  30. Did the study hold for what Democrats and Republicans study in college? In my experience Republicans tend to study the 'harder' subjects –engineering, finance, medicine–while Democrats studied the softer stuff-history, poli sci, gender/race studies, education, etc.

    College graduates with serious degrees may be better at critical reasoning.

    And Mr./Ms. Lord, that is the most arrogant comment I've read in a long time. The scary thing is that I can tell you really believe that.

  31. We found the same results at the Chicago Council on Global Affairs in 2006-2007 Global Views Survey, though it was focused on other issues and comparative foreign policy opinion, but we thougth that doing some partisan differences might be interesting. Here is a quote from page 37 of the 2006 full report:

    "Global warming is perceived as a critical threat by a 32-percentage-point higher proportion of Democrats (62%) than Republicans (30%). This is the largest partisan difference in our data, and it has widened considerably since the 19-point difference of 2004."

    the release is here: http://www.thechicagocouncil.org/dynamic_page.php

Comments are closed.