Guest post: The Elephant in the Room

This is a guest post by Lawrence Hamilton, Senior Fellow at the Carsey School of Public Policy, University of New Hampshire.

This will not surprise you, but surveys see an elephant.

Several colleagues and I have a new paper describing results from 35 surveys that asked the same climate-change question. The question is present-tense and neutrally worded. It does not mention policy or the future. One response choice corresponds to the scientific consensus:

Which of the following three statements do you personally believe?
– Climate change is happening now, caused mainly by human activities.
– Climate change is happening now, but caused mainly by natural forces.
– Climate change is NOT happening now.

Figure 1

Respondents can say they don’t know, or decline to answer, but few do. The first three charts in Figure 1 show results from representative US surveys conducted by different researchers, with different methods, in different years (2011, 2012, 2014). In stark contrast to the consensus among active scientists, each of these general-public surveys found just 52 or 53% agreement on the reality of anthropogenic climate change (ACC). The fourth chart pools data from 21 surveys in the northeastern US state of New Hampshire (2010 to 2015). New Hampshire runs a few points higher (55%) than the nation but might not be a bad proxy.

So who agrees with scientists that human activities are changing the climate? Figure 2 breaks down the percentage of now/human responses on New Hampshire surveys by gender, age, education, political party—the usual suspects in survey research—and also by ideology and frequency of watching Fox news.

Figure 2

I often write about political divisions while wishing that I could talk instead about differences among age cohorts or the effects of weather or something else. But politics is the elephant in the room. Whereas evidence for the reality of ACC has brought overwhelming agreement among scientists, in these general-public data liberals and conservatives stand 62 points apart. Only the divisions by other political indicators come close, such as Democrat/Tea Party (58 points) or no-Fox/daily-Fox (49 points); the gender, age and education divisions on climate change are each below 20 points. Moreover, while the reality of ACC is among the most polarized survey questions, other science or environmental topics, from concern about local beach pollution to trust in scientists for information about vaccines, show divisions going in the same direction.

Figure 3

After politics, education is the second-strongest predictor of views on climate. But politics can neutralize or even reverse the effects of education. College-educated respondents more actively assimilate information in accord with their prejudices, whether these prejudices incline them toward scientific or ideological sources. Figure 3 depicts the probability of a now/human response as a function of education and politics (details here). The pattern is reproduced with remarkable consistency across 34 surveys. Among Democrats and Independents, agreement with the scientific consensus rises with education. Among Republicans, agreement with the scientific consensus does not rise with education, and sometimes even falls. This fall becomes steeper if we separate Tea Party supporters into their own group.

Environmental sociologists Riley Dunlap and Robert Brulle, under the auspices of the American Sociological Association, recently edited a book called Climate Change and Society: Sociological Perspectives. Sociological perspectives lie outside most scientific reports on climate change, even those from interdisciplinary groups with a “human dimensions” component. The reason partly involves that elephant in the room, unseen in physical data and hard for physical scientists to address—yet unmissable in society and social data. As Dunlap and Brulle summarize,

The problem is that despite climate change having never come close to being a consensual, post-political issue in the United States—with fossil fuels and many other corporations and now nearly all of the conservative movement challenging its reality and opposed to acting to limit it—key actors concerned about the issue seem to act as if it were. Thus, government agencies sponsoring and funding climate science and response efforts (epitomized by the USGCRP), foundations funding “climate work,” most national environmental organizations, many scientific organizations, many climate scientists, and individual citizens acting to “save the climate” all tend to downplay if not ignore the extreme degree of conflict and polarization surrounding climate change in the U.S.

Postscript: Same elephant, different day

Last week (9/17 to 9/23) a routine political poll asked more than 700 New Hampshire residents who they favored in the coming presidential primaries, also posing some hypothetical questions like this:

Suppose the 2016 presidential election was being held today and the candidates were Donald Trump, the Republican, and Joe Biden, the Democrat, who would you vote for?

Coincidentally, the same poll also carried several science questions like this one:
Which of the following three statements do you think is more accurate? Scientific measurements have confirmed that in recent decades, the concentration of CO2 or carbon dioxide in the Earth’s atmosphere is increasing, decreasing, or staying about the same?

The upper chart in this graphic plots the race. The lower chart compares supporters of the two candidates and finds a 34-point gap in what they believe about CO2.