America Isn't That Divided on the Environment
After the election, Bernie Sanders spoke to an enthusiastic crowd of students at my university and delivered a message rarely heard in recent months: America is not divided, he said. Rather, the overwhelming majority of citizens agree on core economic goals.
Might that be true, as well, of environmental goals?
Judging from politics, the divide seems real and difficult to surmount. With the exception of an update this year to the Toxic Substances Control Act, Congress has not passed any major environmental statute or amended any important environmental law since the Clean Air Act Amendments of 1990. And this inaction stems from lawmakers’ perception that public opinion in this area divides along party lines.
Polls seem to back up the assumption: For more than 20 years, Gallup has asked Americans whether they “personally worry about” environmental problems -- including water and air pollution, species loss and climate change. Since at least 2000, respondents have given drastically different responses according to their political party affiliation. Democrats worry about environmental problems; Republicans do not. Other surveys have had similar results.
But this evidence from surveys and focus groups is a narrow and often inaccurate base on which to develop environmental policy, because while it reflects what people say when asked their opinion, it says nothing about how those same people behave. And solving environmental problems is fundamentally about changing human behavior.
Three recent randomized field experiments examined the environmental actions of tens of thousands of people. And none of them identified any partisan divide.
The first looked at people’s willingness to conserve water, and found that the more frequently a household voted, the greater was its commitment to water reduction during droughts -- regardless of party affiliation. Voters cut back on water use more than people who don’t vote in party primaries and people who are not registered to vote. Comparing the voters of the two major parties, the researchers found no greater difference than what would happen by chance.
Another large-scale randomized experiment examined energy conservation, and found that whether people identified as liberal or conservative made little difference in how much they would voluntarily conserve electricity.
A third study, conducted by the U.S. Department of Agriculture, examined efforts to motivate farmers in the Great Lakes region to take actions to mitigate climate change. Farmers generally tend to be conservative and vote Republican, and the USDA wanted to see if they would respond better to messages that avoided the use of words that might be politically provocative -- terms such as “greenhouse gas emissions.” As it turned out, that language had no effect. Among the 10,000 farmers contacted in the study, interest in learning more about how to better manage carbon-rich soils was not affected by references to climate change -- interest was the same whether the messages that farmers received spoke of “soil health” and “local water quality” or of “climate change” and “greenhouse gas emissions.” So the farmers’ stated opinions of climate change turned out not to be good predictors of their willingness to act to protect the climate.
Indeed, in all three experiments, the behavioral partisan divide turned out to be small.
Behavioral scientists would explain these findings by pointing out that all three experiments motivated people’s actions by describing a variety of private and public benefits. They enabled people to make their decisions privately, and they avoided hot-button cultural-identity questions and trigger words, such as “green.”
Together, the experiments provide evidence that partisan polarization should not be an insurmountable obstacle to taking action to protect the environment. They also suggest that if polling is to be useful in setting environmental policy, it should focus less on beliefs and attitudes and more on actions. Better yet, it should work to identify agreement on acceptable actions across partisan lines. Ideally, we’d go one step further and complement polling data with more field experiment data, so that politicians and program administrators could place less weight on evidence from surveys and more on evidence derived from real behavior.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the author of this story:
Paul J. Ferraro at firstname.lastname@example.org
To contact the editor responsible for this story:
Mary Duenwald at email@example.com