Poll questions shift public views on global warming

• U.S. majorities have agreed on global warming being real for three decades%2C poll analysis says
• Poll question wording has skewed survey answers%2C in some cases
• After a dip in 2009%2C public agreement on climate change has resumed climbing toward 75%25

The Arctic keeps melting, the atmosphere keeps warming and polls keep bouncing around on what people think about global warming.

After every hurricane, heat wave or hullabaloo over climate science, the polling results come, watched breathlessly for some shift in U.S. public opinion about global warming.

All the hand-wringing may not have amounted to anything, suggests one long-term look at all those polls. The majority of the public pretty much understands that global warming is happening, and has for a long time, the authors say. Some of what looks like confusion about what folks think may result more from the poll questions themselves, rather than from the people answering the questions.

"Belief that global warming is happening has been mostly stable and increasing for the last thirty years," says social scientist Orie Kristel of The Strategy Team, an applied social science company based in Columbus, Ohio. The agreement has approached 75%, and although it dipped in recent years, that consensus has since resumed its upward march, according to a just-released report sponsored by the Skoll Global Threats Fund, a foundation founded by eBay billionaire Jeff Skoll that looks for solutions to global problems such as pandemics, nuclear proliferation and environmental challenges. In it, Kristel and his colleagues weigh together public opinion polls dating back to 1986, from more than 150 nationwide questionnaires in all.

He and his colleagues report that wording of poll questions may have created some of the appearance of shifts in public opinion about global warming.

"Do you think the greenhouse effect really exists or not?" a poll first asked U.S. respondents in 1986. About 73% answered "yes" in that year, setting a pattern. When pollsters asked folks whether they believed climate change was happening in some sense, most said they did. When they asked folks, "Is there solid evidence the average temperature on Earth has been getting warmer over the past four decades, or not," as a Brookings Institution survey asked in 2007, the responses were "consistently lower," the analysis finds. More polls have been asking the question this way in recent years.

In other words, ask people what they believe and they will mostly say they believe global warming is happening. If you pile on top of that question the additional task of asking people to assess what they know of the evidence (which may be very little), they become more doubtful in their answers. In that case, more than half of people say they believe in global warming, but the level of agreement drops below 70%, Kristel says.

Scientists might point out that the greenhouse effect indisputably exists, a finding reinforced by a 2010 report from the U.S. National Academy of Sciences, and that greenhouse gases such as carbon dioxide do heat the atmosphere no matter what people tell pollsters. They might add that more than 97% or more of climate scientists agree that it is "very likely" that man-made greenhouse gases added to the air have caused "most" of the "unequivocal" warming of the Earth's average global temperature in the second half of the 20th century, according to a 2010 survey in the Proceedings of the National Academy of Sciences journal.

Regardless, weighing all the polls together over the past 30 years, Kristel and colleagues conclude that people increasingly agreed on global warming until around 2008. At that point it dipped to below 70% until 2010, when agreement on global warming resumed its climb above that mark.

A Stanford University-led poll last year suggested that hardening of opinion against global warming among the most conservative of voters largely accounted for the dip, with the rest of the public not changing its opinion too much. Edward Maibach, a climate public opinion expert at George Mason University in Fairfax, Va., echoes this point, saying, "The overall trend on agreement about climate change masks the divergence between liberals and Democrats increasingly agreeing that climate change is occurring while conservatives and Republicans increasingly disagree."

Experts have proposed a run of cold winters followed by recent drought-stricken summers and other extreme weather as the reason for the dip and rise. They also attribute it to economics, , the great recession or the shift among Republican politicians such as Mitt Romney, away from support for ways to address climate change.

Maibach suggests that a drop in news coverage after 2007 about climate followed by people getting hit by extreme weather events personally may account for the dip and resurgence, as well.

At the same time, more poll questions asked people to "recall what you have heard or read" about global warming before answering whether they believed it was occurring, which Kristel suggests may have piled more noise into their thought processes before they responded and led to less agreement.

"This may illustrate an interesting psychological insight," Kristel says. Questions that ask people to do more than simply say what they believe, straight up or straight down, might introduce more opportunities for doubts to creep into their answers, changing the results, he says.

Maibach agrees that "questions do matter," but cautions that simply noting broad agreement on the matter of climate change might miss the political tenor underlying the divisions.

Another example of the way question wording changes poll answers comes from the second part of the Strategy Team report. Folks have been asked whether they view global warming as a "serious problem" at least as far back as 1980, almost 180 times, in fact. How were they asked?

In three ways, it turns out: About 50 times they were asked whether global warming was a serious problem in general, without specifying whom or what it might affect. Another 50 times or so, they were asked if global warming threatened them, their family or town. Finally, they were asked about it threatening the nation, the world or future generations.

Once again, how you asked the question mattered. The closer the threat was to themselves, the less often that people, less than half since 1987, viewed global warming as a serious problem. Similarly, seeing global warming as a threat to their family or town has garnered fewer affirmative answers since 1991, down below half now.

When you ask them to evaluate the risk threat to people they don't know, more and more people rated global warming as a threat, with more than 75% of people viewing it as a serious problem for future generations. This number hopped even higher when pollsters worded the question to ask folks to identify the most serious problem that will face the world "if nothing is done to stop it."

"Overall, these results suggest Americans have a nuanced perspective regarding global warming as a serious problem," the report says.

"It is an idea that you almost couldn't come up with on your own if you tried to better raise the suspicions of people who are already nervous about the environmental movement," said David Goldston of the Natural Resources Defense Council, speaking at the recent American Association for the Advancement of Science meeting in Boston.

What might people take from this the next time they read about a poll on climate and public opinion?

"The bottom line is that I would tell people reading about a poll to first read who was actually asked the question," Kristel says, noting that surveys that polled only registered voters or people in one part of the country may have different results than a national survey. "And then they should see what was actually asked in the poll, how the question was worded. Be a careful reader."